Anda di halaman 1dari 472

Medical Illness


Second Edition

This page intentionally left blank

Medical Illness

Second Edition

Edited by

Jonathan M. Meyer, M.D.

Henry A. Nasrallah, M.D.

Washington, DC
London, England

Note: The authors have worked to ensure that all information in this book is
accurate at the time of publication and consistent with general psychiatric and
medical standards, and that information concerning drug dosages, schedules,
and routes of administration is accurate at the time of publication and consistent with standards set by the U.S. Food and Drug Administration and the general medical community. As medical research and practice continue to advance,
however, therapeutic standards may change. Moreover, specific situations may
require a specific therapeutic response not included in this book. For these reasons and because human and mechanical errors sometimes occur, we recommend that readers follow the advice of physicians directly involved in their care
or the care of a member of their family.
Books published by American Psychiatric Publishing, Inc., represent the views
and opinions of the individual authors and do not necessarily represent the policies and opinions of APPI or the American Psychiatric Association.
If you would like to buy between 25 and 99 copies of this or any other APPI title,
you are eligible for a 20% discount; please contact APPI Customer Service at or 800-368-5777. If you wish to buy 100 or more copies of the
same title, please e-mail us at for a price quote.
Copyright 2009 American Psychiatric Publishing, Inc.
Manufactured in the United States of America on acid-free paper
13 12 11 10 09
5 4 3 2 1
Second Edition
Typeset in Adobes Book Antiqua and Albertus
American Psychiatric Publishing, Inc.
1000 Wilson Boulevard
Arlington, VA 22209-3901
Library of Congress Cataloging-in-Publication Data
Medical illness and schizophrenia / edited by Jonathan M. Meyer, Henry A.
Nasrallah. 2nd ed.
p. ; cm.
Includes bibliographical references and index.
ISBN 978-1-58562-346-4 (alk. paper)
1. SchizophrenicsDiseases. 2. SchizophrenicsMedical care. 3. Mentally
illDiseases. 4. Mentally illMedical care. I. Meyer, Jonathan M., 1962 II.
Nasrallah, Henry A.
{DNLM: 1. Schizophreniacomplications. 2. Schizophreniaepidemiology.
3. Comorbidity. 4. Patient Care. WM 203 M489 2009]
RC514.M425 2009
British Library Cataloguing in Publication Data
A CIP record is available from the British Library.

Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ix
Disclosure of Interests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

Part I
Public Health Issues for
Schizophrenia Patients

Chapter 1
Improving Physical Health Care for Patients With Serious
Mental Illness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
David Folsom, M.D., M.P.H.

Chapter 2
Excessive Mortality and Morbidity Associated With
Schizophrenia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17
Daniel E. Casey, M.D.
Thomas E. Hansen, M.D.

Chapter 3
Medical Outcomes From
the CATIE Schizophrenia Study . . . . . . . . . . . . . . . . . . . . . 37
Henry A. Nasrallah, M.D.

Part II
Metabolic Disease, Heart Disease,
and Related Conditions

Chapter 4
Obesity and Schizophrenia . . . . . . . . . . . . . . . . . . . . . . . . . .61
Tony Cohn, M.B.Ch.B., M.Sc., F.R.C.P.C.

Chapter 5
Glucose Intolerance and Diabetes in Patients With
Schizophrenia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .91
David C. Henderson, M.D.
Kathleen Miley, B.S.

Chapter 6
Effects of Antipsychotics on Serum Lipids . . . . . . . . . . . . . 117
Jonathan M. Meyer, M.D.

Chapter 7
The Spectrum of Cardiovascular Disease in
Patients With Schizophrenia . . . . . . . . . . . . . . . . . . . . . . . .169
Jimmi Nielsen, M.D.
Egon Toft, M.D., F.E.S.C.

Chapter 8
Behavioral Treatments for Weight Management of
Patients With Schizophrenia. . . . . . . . . . . . . . . . . . . . . . . . 203
Rohan Ganguli, M.D., F.R.C.P.C.
Tony Cohn, M.B.Ch.B., M.Sc., F.R.C.P.C.
Guy Faulkner, B.Ed., M.Sc., Ph.D.

Chapter 9
Nicotine and Tobacco Use in Patients With
Schizophrenia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
Andrea H. Weinberger, Ph.D.
Tony P. George, M.D., F.R.C.P.C.

Part III
Special Topics and Populations

Chapter 10
HIV and Hepatitis C in Patients With Schizophrenia . . . . . 247
Milton L. Wainberg, M.D.
Francine Cournos, M.D.
Karen McKinnon, M.A.
Alan Berkman, M.D.
Mark Drew Crosland Guimares, M.D., D.Sc., M.P.H.

Chapter 11
Substance Abuse and Schizophrenia . . . . . . . . . . . . . . . . . 275
Peter F. Buckley, M.D.
Jonathan M. Meyer, M.D.

Chapter 12
Sexual Dysfunction and Schizophrenia . . . . . . . . . . . . . . . . 303
Heidi J. Wehring, Pharm.D., B.C.P.P.
Deanna L. Kelly, Pharm.D., B.C.P.P.

Chapter 13
Managing the Health Outcomes of
Schizophrenia Treatment in Children and Adolescents . . . 343
Christoph U. Correll, M.D.

Chapter 14
Medical Health in Aging Persons With Schizophrenia . . . . 377
Samantha Brenner, M.P.H.
Carl I. Cohen, M.D.

Chapter 15
Managing Health Outcomes of Women With
Schizophrenia During Pregnancy and Breastfeeding . . . . . . .415
Adele C. Viguera, M.D., M.P.H.
Mackenzie Varkula, D.O.
Katherine Donovan, B.A.
Ross J. Baldessarini, M.D.

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435

Ross J. Baldessarini, M.D.
Professor of Psychiatry, Harvard Medical School and Massachusetts
General Hospital, Boston, Massachusetts; Director, International Consortium for Psychotic Disorders Research, McLean Hospital, Belmont,
Alan Berkman, M.D.
Associate Professor of Epidemiology, Mailman School of Public Health,
Columbia University, New York, New York
Samantha Brenner, M.P.H.
Medical Student, Department of Psychiatry, State University of New
York Health Science Center at Brooklyn, Brooklyn, New York
Peter F. Buckley, M.D.
Professor and Chairman, Department of Psychiatry, Medical College of
Georgia, Augusta, Georgia
Daniel E. Casey, M.D.
Professor of Psychiatry and Neurology, Oregon Health and Science
University, Portland, Oregon
Carl I. Cohen, M.D.
Professor and Director, Division of Geriatric Psychiatry, Department
of Geriatric Psychiatry, State University of New York Health Science
Center at Brooklyn, Brooklyn, New York
Tony Cohn, M.B.Ch.B., M.Sc., F.R.C.P.C.
Director, Mental Health and Metabolism Clinic, Centre for Addiction
and Mental Health; Assistant Professor, Departments of Psychiatry
and Nutritional Sciences, University of Toronto, Toronto, Ontario,


Medical Illness and Schizophrenia

Christoph U. Correll, M.D.

Assistant Professor of Psychiatry and Behavioral Sciences, Albert Einstein College of Medicine; Medical Director, Director, Adverse Events
Assessment and Prevention Unit, The Zucker Hillside Hospital North
Shore, Long Island Jewish Health System, Glen Oaks, New York
Francine Cournos, M.D.
Professor of Clinical Psychiatry, Columbia College of Physicians and
Surgeons; Director, Washington Heights Community Service, New
York State Psychiatric Institute, New York, New York
Mark Drew Crosland Guimares, M.D., D.Sc., M.P.H.
Professor of Epidemiology, Department of Preventive and Social Medicine, Faculty of Medicine, Federal University of Minas Gerais, Belo
Horizonte, Minas Gerais, Brazil
Katherine Donovan, B.A.
Clinical Research Coordinator, Massachusetts General Hospital Center
for Womens Health, Boston, Massachusetts
Guy Faulkner, B.Ed., M.Sc., Ph.D.
Professor, Faculty of Physical Education and Health, University of
Toronto, Toronto, Ontario, Canada
David Folsom, M.D., M.P.H.
Assistant Professor of Psychiatry, University of California, San Diego,
La Jolla, California
Rohan Ganguli, M.D., F.R.C.P.C.
Professor of Psychiatry and Executive Vice-President of Clinical
Programs, Centre for Addiction and Mental Health, Toronto, Ontario,
Tony P. George, M.D., F.R.C.P.C.
Head, Addiction Psychiatry Program, Department of Psychiatry, University of Toronto, and Schizophrenia Program, Centre for Addiction
and Mental Health, Toronto, Ontario, Canada
Thomas E. Hansen, M.D.
Associate Professor of Psychiatry, Oregon Health and Science University; Staff Psychiatrist, Portland Oregon State Hospital, Portland,



David C. Henderson, M.D.

Director, Schizophrenia, Diabetes, and Weight Reduction Research Program, Massachusetts General Hospital; Associate Professor of Psychiatry, Harvard Medical School, Boston, Massachusetts
Deanna L. Kelly, Pharm.D., B.C.P.P.
Associate Professor of Psychiatry and Acting Director, Treatment Research Program, Maryland Psychiatric Research Center, University of
Maryland School of Medicine, Baltimore, Maryland
Karen McKinnon, M.A.
Research Scientist, New York State Psychiatric Institute; Project Director,
Columbia University HIV Mental Health Training Project, Columbia
University College of Physicians and Surgeons, New York, New York
Jonathan M. Meyer, M.D.
Assistant Professor of Psychiatry, University of California, San Diego;
Staff Psychiatrist, San Diego VA Medical Center, La Jolla, California
Kathleen Miley, B.S.
Research Assistant, Massachusetts General Hospital, Boston, Massachusetts
Henry A. Nasrallah, M.D.
Professor of Psychiatry, Neurology and Neuroscience, University of
Cincinnati College of Medicine, Cincinnati, Ohio
Jimmi Nielsen, M.D.
Research Psychiatrist, Unit for Psychiatric Research, Aarhus University
Hospital and Aalborg Psychiatric Hospital, Aalborg, Denmark
Egon Toft, M.D., F.E.S.C.
Professor, Department of Health Science and Technology, Aalborg University, Aalborg, Denmark
Mackenzie Varkula, D.O.
Resident Physician, Department of Psychiatry and Psychology, Cleveland Clinic Foundation, Cleveland, Ohio
Adele C. Viguera, M.D., M.P.H.
Associate Professor of Psychiatry, Department of Psychiatry and Psychology, Cleveland Clinic Foundation, Cleveland, Ohio


Medical Illness and Schizophrenia

Milton L. Wainberg, M.D.

Associate Clinical Professor of Psychiatry, Columbia University
College of Physicians and Surgeons; Director of Medical Education,
Columbia University HIV Mental Health Training Project, New York,
New York
Heidi J. Wehring, Pharm.D., B.C.P.P.
Fellow, Maryland Psychiatric Research Center, University of Maryland
School of Medicine, Baltimore, Maryland
Andrea H. Weinberger, Ph.D.
Assistant Professor, Program for Research in Smokers with Mental
Illness, Division of Substance Abuse, Department of Psychiatry, Yale
University School of Medicine, The Connecticut Mental Health Center,
New Haven, Connecticut

Disclosure of Interests
The following contributors to this book have indicated a financial interest in or other
affiliation with a commercial supporter, a manufacturer of a commercial product, a provider of a commercial service, a nongovernmental organization, and/or a government
agency, as listed below:
Daniel E. Casey, M.D.Consultant: Abbott Laboratories, Bristol-Myers Squibb,
Dainippon Sumitomo Pharma, Janssen Pharmaceutica, NuPathe, Pfizer,
Solvay Pharmaceuticals, and Wyeth Pharmaceuticals; Speakers bureau: Abbott Laboratories, Bristol-Myers Squibb, Janssen, and Pfizer.
Tony Cohn, M.B.Ch.B., M.Sc., F.R.C.P.C.Research grant funding: Pfizer
Canada; Speakers fees: Pfizer Canada.
Christoph U. Correll, M.D.Grant support: Einstein Institute for Medical Research, National Institute of Mental Health (NIMH), and Stanley Foundation; Consultant: AstraZeneca, Bristol-Myers Squibb, Eli Lilly, Janssen,
Otsuka, Pfizer, Supernus, and Vanda; Speaker's bureau: AstraZeneca, Bristol-Myers Squibb, and Otsuka.
Rohan Ganguli, M.D., F.R.C.P.C.Consultant: Janssen; Speakers honoraria:
Bristol-Myers Squibb.
Tony P. George, M.D., F.R.C.P.C.Grant support: Donaghue Medical Research
Foundation, National Alliance for Research on Schizophrenia and Depression (NARSAD), National Institute on Drug Abuse (NIDA), Sanofi-Aventis,
and Sepracor; Advisory board/consultant: Eli Lilly, Evotec, and Pfizer.
David C. Henderson, M.D.Research grants: NIMH, Solvay, and Takeda; Honoraria: Bristol-Myers Squibb, Covance, Janssen, Pfizer, Primedia, Reed
Medical Education, and Solvay Pharmaceuticals.
Deanna L. Kelly, Pharm.D., B.C.P.P.Advisory board: Bristol-Myers Squibb
and Solvay.
Jonathan M. Meyer, M.D.Research support: Bristol-Myers Squibb, NIMH, and
Pfizer; Speaking/advising fees: AstraZeneca, Bristol-Myers Squibb, Organon,
Pfizer, Vanda Pharmaceuticals, and Wyeth.
Henry A. Nasrallah, M.D.Research support: AstraZeneca, Forest Pharmaceuticals, GlaxoSmithKline, Janssen, NIMH, Pfizer, Roche, and Sanofi; Consultant: AstraZeneca, Cephalon, Janssen, Pfizer, and Vanda; Speakers
bureau: Abbott, AstraZeneca, Janssen, Pfizer, Solvay, and Vanda.



Medical Illness and Schizophrenia

Adele C. Viguera, M.D., M.P.H.Grant/research support: AstraZeneca (Claffin

Award), Berlex Laboratories, Eli Lilly, Epilepsy Foundation*, Forest,
GlaxoSmithKline, Beecham Pharmaceuticals, Harvard Medical School's
Scholars in Medicine Fellowship, Janssen, NARSAD, NIMH, Pfizer, Sepracor, Stanley Medical Research Institute, and Wyeth-Ayerst; Speaker's
bureau/honoraria: AstraZeneca*, Eli Lilly, Forest*, GlaxoSmithKline, Novartis Pharmaceuticals, Wyeth-Ayerst; Advisory board: GlaxoSmithKline*,
Novartis, and Wyeth-Ayerst*.
*Funding in the last 12 months.
Andrea H. Weinberger, Ph.D.Grant support: NARSAD and Sepracor.
The following authors affirmed that they have no competing interests:
Alan Berkman, M.D.
Samantha Brenner, M.P.H.
Carl I. Cohen, M.D.
Francine Cournos, M.D.
Mark Drew Crosland Guimares, M.D., D.Sc., M.P.H.
Guy Faulkner, B.Ed., M.Sc., Ph.D.
David Folsom, M.D., M.PH.
Thomas E. Hansen, M.D.
Karen McKinnon, M.A.
Jimmi Nielsen, M.D.
Egon Toft, M.D., F.E.S.C.
Milton L. Wainberg, M.D.
Heidi J. Wehring, Pharm.D., B.C.P.P.

American Psychiatric Publishing, Inc. (APPI)Subsequent to peer review and

acceptance for publication of this volume, APPI entered into an agreement for
Dainippon Sumitomo Pharma America, Inc., to underwrite and distribute a
separate printing of Medical Illness and Schizophrenia, 2nd Edition.

Since the publication of the first edition of this volume in 2003,
there has been a dramatic increase in the literature on optimizing medical outcomes for patients with schizophrenia. Sitting before me is the
August 2008 issue of Psychiatric Services, which includes no less than
eight research articles devoted to medical issues of relevance to those
who care for individuals with severe mental illness. Despite the growing awareness of the unique medical needs in this patient population,
natural causes of death remain the primary source of excess mortality,
with recent data indicating a widening mortality gap between patients
with schizophrenia and the general population (Saha et al. 2007). The
recognition that antipsychotic treatment can be associated with deleterious metabolic and other health effects emphasizes the need to address
medical aspects of schizophrenia treatment, yet the mental health community as a whole still has a great deal to accomplish toward achieving
parity in physical health monitoring and treatment between the severely mentally ill and the rest of society. Barriers remain, but they are
not insurmountable.
We have expanded this second edition with the aim of providing a
comprehensive resource for those who wish to be informed about the
various medical issues that impinge on schizophrenia treatment. Our
intent is to create a practical manual that can serve as a reference for the
medical management of patients with severe mental illness across the
age spectrum, with a focus on those areas of significant importance to
schizophrenia patients, such as cardiometabolic disorders and their relationship to antipsychotic therapy, smoking, substance use, and infection with human immunodeficiency virus (HIV) and hepatitis C. We
have also added a section on the important but neglected topic of sexual
function in patients with schizophrenia and have devoted an entire
chapter to the behavioral management of weight gain and obesity. The
intent, as always, is to stimulate those who work in mental health settings to take a broader view of schizophrenia care, recognizing that this
is a systemic disease with multiple manifestations that go beyond the
obvious psychiatric symptoms.
Jonathan M. Meyer, M.D.


Medical Illness and Schizophrenia

Saha S, Chant D, McGrath J: A systematic review of mortality in schizophrenia:
is the differential mortality gap worsening over time? Arch Gen Psychiatry
64:11231131, 2007

Public Health
Issues for

This page intentionally left blank

Improving Physical
Health Care for
Patients With Serious
Mental Illness
David Folsom, M.D., M.P.H.

Much has been written about the need to integrate mental

health care into the primary care setting. Psychiatrists and other mental
health care providers have often criticized primary care physicians for
underrecognizing and undertreating depression, substance abuse, and
other common mental health conditions (Wells et al. 2000). Meanwhile,
the challenge of providing quality medical care to patients with serious
mental illnesses represents a similar challenge for psychiatrists.
Atypical antipsychotic medications are the mainstay of contemporary pharmacological treatment for psychotic disorders (Harrington et
al. 2000; IMS Health 2002; Jin et al. 2004). However, in the past decade,
concern and awareness have been growing that these medications may
cause or exacerbate medical conditions including weight gain and obesity (Daumit et al. 2003; Lieberman et al. 2005), elevated blood sugar
and diabetes (Henderson et al. 2005; Leslie and Rosenheck 2004; Lieberman et al. 2005), and cerebrovascular events (Schneider et al. 2005). The
U.S. Food and Drug Administration currently mandates that the package inserts for all six atypical antipsychoticsaripiprazole (Abilify

Medical Illness and Schizophrenia

2008), clozapine (Clozaril 2008), olanzapine (Zyprexa 2008), quetiapine

(Seroquel 2008), risperidone (Risperdal 2008), and ziprasidone (Geodon
2008)include black-box warnings regarding the increased risk of
mortality in elderly patients with dementia and a warning about the
risk of hyperglycemia and diabetes. These warnings and the emerging
data on the medical side effects of antipsychotics have raised psychiatrists concern and awareness about comorbid medical conditions in
their patients and have led to guidelines for the medical monitoring of
patients taking these medications (American Diabetes Association et al.
2004; Marder et al. 2002). However, evidence that predates the atypical
antipsychotics suggested that patients with severe mental illness (SMI)
are at higher risk for several medical conditions and often receive inferior medical care, both of which likely contribute to their elevated mortality. In this chapter, I briefly review the data on the prevalence of
comorbid medical conditions and the lack of quality medical care for
patients with SMI, examine prior studies that attempt to improve the
health status of these patients, and consider policies that might improve
the integration of medical and mental health care for patients with SMI.

Medical Comorbidity and Access to

Quality Medical Care for Patients With
Serious Mental Illness
Patients with schizophrenia and other serious mental illnesses constitute a particularly vulnerable group, for whom access to quality medical care is problematic (Druss et al. 2001a; Harvey et al. 2005). Druss et
al. (2001a) reported that following myocardial infarction, patients with
schizophrenia had a 30% greater 1-year mortality than patients who
were not mentally ill. Approximately half of this increased mortality
was due to a lack of quality medical treatment after the myocardial infarction. Folsom et al. (2002) found that in homeless patients with mental illness, those with schizophrenia were less likely to receive primary
and preventive care than were patients with major depression. Furthermore, investigators have reported elevated cardiovascular risk (McCreadie 2003), as well as inferior medical care and elevated mortality in
persons with schizophrenia compared with the general population (Koran et al. 1989; Phelan et al. 2001). For example, the rates of reported hypertension were 40% lower in persons with schizophrenia than in the
general population, but the rates of admission for end-stage complications of hypertension, including cardiomyopathy and pulmonary

Improving Physical Health Care

edema, were 1.8 and 1.5 times greater, respectively, in the patients with
schizophrenia (Munk-Jrgensen et al. 2000). These findings suggest
that patients with schizophrenia were less likely to receive treatment in
the early phase of a disease and instead were receiving care later, when
the disease became more severe and required hospitalization.
Some encouraging data have also been published regarding health
care for patients with schizophrenia. An investigation of patients
treated in community-based settings found that, compared with the
general population, a greater percentage of patients with schizophrenia
had received primary care treatment in the prior year (Dickerson et al.
2003). At the same time, the growing concerns about the risk of diabetes, myocardial infarction, and stroke in patients taking atypical antipsychotics have increased the awareness of the importance of diagnosis
and treatment of comorbid medical conditions in patients with schizophrenia (Folsom and Okereke 2005; Jin et al. 2004).

Models for Improving Medical Care for

Patients With Serious Mental Illness
Although the literature on interventions and service delivery models to
improve the health outcomes of patients with SMI is not as extensive as
that of treating depression in primary care, four basic approaches have
been examined: 1) health care skills training for patients, 2) training of
psychiatrists to directly provide primary care, 3) facilitated referral to
primary care, and 4) co-location or integration of primary care with
mental health care (Druss 2007).

Health Care Skills Training for Patients

Several interventions that have been developed and tested focus on
providing education and skills training for patients with chronic mental
illnesses who also have a comorbid medical condition. Two skills training interventions specifically related to health care outcomes are particularly noteworthy. One intervention has been described by McKibbin
et al. (2006), who enrolled 60 patients over age 40 with schizophrenia
and comorbid diabetes into their diabetes awareness and rehabilitation
training (DART) program. These authors tested whether patients who
received an intervention based on social cognitive theory and incorporating standard diabetes self-management approaches experienced
greater reductions in weight and hemoglobin A1c than did a comparison group that received only diabetes education. The intervention

Medical Illness and Schizophrenia

lasted 24 weeks, with patients randomly assigned to one of the two

groups. At the end of the study, patients in the DART intervention lost
an average of 2.3 kg, compared with a 2.7-kg weight gain in the comparison group; however, there were no significant differences in hemoglobin A1c between the two groups.
The second skills training intervention is exemplified by Wu et al.s
(2008) study, which included patients who had experienced at least a
10% weight gain from atypical antipsychotics. A 2 2 factorial design
was used to compare the effects of a lifestyle intervention, metformin (a
diabetes medication that inhibits hepatic glucose production), both interventions, and placebo. This study, conducted in China, enrolled 128
patients and examined changes in weight after 12 weeks of treatment.
Over the 12-week period, patients in the placebo group gained an average of 3.1 kg, whereas those receiving the lifestyle intervention lost
1.4 kg, those taking metformin lost 3.2 kg, and those in the combined
lifestyle-metformin group lost 4.7 kg. Similar results were reported by
Weber and Wyne (2006), who also tested a cognitive-behavioral intervention for overweight patients taking an atypical antipsychotic. The
mean body mass index of patients enrolled in this study was 33, and at
the end of the investigation, the mean weight loss for those in the intervention group was 2.4 kg, compared with 0.6 kg for those in the comparison group.
These results demonstrate that lifestyle interventions modified for
patients with schizophrenia and other serious mental illnesses can result in improved health outcomes. However, aside from weight gain
and a single report on patients with diabetes, data are lacking on the effects of patient education interventions on other health outcomes for
patients with SMI. Ongoing studies funded by the National Institute of
Mental Health (NIMH) are examining health care outcomes, and in the
coming years, information from these studies should help provide data
on improving the health of patients with SMI and other health problems.

Training of Psychiatrists to Provide Primary Care

A second approach to improving the health of patients with SMI is to
have some aspects of primary health care provided by psychiatrists.
McIntyre and Romanos (1977) article Is There a Stethoscope in the
House (and Is It Used)? highlighted the lack of practicing psychiatrists comfort with conducting physical examinations of their patients.
Shore (1996) proposed that psychiatrists need to be able to provide

Improving Physical Health Care

basic primary care to their patients with SMI, arguing that psychiatrists
are frequently the only physicians these patients see. Golomb et al.
(2000) assembled a consensus panel of Veterans Affairs (VA) psychiatrists, internists, hospital administrators, and nurses to identify health
conditions that would be appropriately treated by a psychiatrist who
has primary care training and on-site supervision by an internist. All
panel members agreed that a psychiatrist with primary care training
could provide care for 17 of 38 preventive screening and counseling interventions but could perform only 10 of 157 evaluations of medical
conditions and provide treatment for only 14 of 125 medical conditions.
Examples of common medical conditions that the consensus panel
thought primary care psychiatrists could treat included cough, headache, and nasal congestion. Examples of common conditions that the
panel did not think that primary care psychiatrists could treat included
chest pain and ear infection.
A few psychiatry residencies have begun to train their residents to
provide primary care to patients with serious mental illness. These programs include the University of California San Diego and the Oregon
Health and Science University in Portland. In a series of reports, Dobscha and colleagues at the Portland VA describe the outcomes of providing this type of training in an integrated psychiatry primary medical
care (PPMC) program (Dobscha and Ganzini 2001; Dobscha et al. 2005;
Snyder et al. 2008). In one of these studies (Dobscha et al. 2005), nine
graduates of the PPMC program and 34 other residency graduates were
asked how comfortable they were with general medical conditions and
how frequently they screened for or provided treatment for medical
conditions. Responses indicated that, overall, the residents who
worked in the PPMC program were more comfortable with general
medical conditions than were their fellow residents who did not work
in this program. However, few of the PPMC-trained residents provided
direct medical care to their patients after graduation.
In summary, despite some recommendations that training psychiatrists should provide a limited range of primary care to their patients
with a serious mental illness, this model has been adopted in only a few
residency training programs. In addition, although the data from these
programs are limited, the psychiatrists whose training included providing primary care for some of their patients appeared to be more comfortable with the medical problems of their patients than were
psychiatrists without such training, but the primary caretrained psychiatrists did not provide much direct medical care after leaving the
training program.

Medical Illness and Schizophrenia

Facilitated Referral to Primary Care

The third model for improving the health care of patients with SMI is to
improve the ability of mental health centers to refer their patients to primary care providers and medical clinics. This can be done either by providing training, time, and resources to all of the mental health clinicians
in a particular clinic or by having a specific care manager provide this
linkage and care coordination. Some empirical evidence supports this
model, and findings indicate that patients who receive specific help
with accessing primary care treatment receive more primary and preventive health care than patients who do not receive such help.
In a small pilot study, Bartels et al. (2004) compared the primary care
received by 24 older patients with a serious mental illness who worked
with a nurse care manager whose job was to help patients access primary and preventive health care. Using a chart review, the researchers
found that, prior to the intervention, 17% of patients did not have a designated primary care provider and 29% had not received any preventive health care in the past year, whereas after working with the nurse
case manager, all the patients had a designated primary care provider
and had completed a preventive health care visit.
In a larger study of 470 patients with a substance abuse disorder entering a detoxification program, Samet et al. (2003) randomly assigned
half to a one-time comprehensive assessment plus working with a social worker to arrange for follow-up primary medical care at a clinic,
and the other half of the patients received usual care. The patients in the
clinic intervention group were more likely to have successfully seen a
primary care physician at least once in the 12 months after completing
the detoxification program than were patients who received usual care
(69% vs. 53%).
Currently, at least two ongoing NIMH-funded studies are examining
whether this model would result in improved health outcomes for patients with serious mental illness (S. Bartels, personal communication,
July 2008; Druss 2007).

Co-Location or Integration of
Mental Health Treatment and Primary Care
The fourth approach to improving the health care of patients with SMI
is to co-locate primary care and mental health care within the same
clinic, ideally in an integrated manner that has psychiatrists and primary care providers working together to care for patients with SMI and
chronic medical conditions. These clinics would be staffed by a combination of psychiatry and primary care physicians, or by physicians du-

Improving Physical Health Care

ally trained in psychiatry and a primary care specialty (family medicine

or internal medicine).
Druss et al. (2001b) compared patient outcomes in a clinic that provided integrated medical and psychiatric care with those of a clinic providing a usual care model. In this VA-based study, 59 patients were
randomly assigned to receive care in an integrated care clinic based in
a VA mental health clinic that was also staffed by a full-time nurse practitioner, a part-time family physician, and a nurse case manager. The
comparison group included 61 veterans who received their care in a
typical VA outpatient medical clinic. Approximately three-fourths of
the patients had a serious mental illness, with posttraumatic stress disorder, substance use disorders, and schizophrenia being the most common diagnoses. In addition, more than half of the patients had a
previously unknown medical diagnosis. Compared to the patients receiving usual care, those in the integrated care clinic were more likely
to use primary care services and less likely to use the emergency room.
In addition, patients in the integrated clinic were more likely to receive
preventive health care, including screening for colon cancer and diabetes, immunizations, and health education. Patients in the integrated
clinic also rated their satisfaction with care more highly than did those
in the comparison group, and demonstrated a small but significant improvement in their physical health-related quality of life.
In an interesting qualitative assessment of this integrated clinic,
Miller et al. (2003) used focus groups to identify three aspects of integrated care that were particularly important to people with a serious
mental illness. First, patients reported that they had previously faced
difficulty obtaining medical care in more traditional settings. For example, patients reported that in the past they had been turned away from
medical clinics, were not treated with respect, or had their medical concerns dismissed. Second, the integrated clinic had greater flexibility and
availability of resources, including a nurse case manager and a smaller
patient panel size for the primary care providers. Finally, both patients
and clinic staff noted that the communication between the patients
psychiatrist and primary care provider was greater than what they had
previously experienced, and primary care and mental health clinicians
reported that they benefited from regular contact with each other.
Another option for integrated care, training physicians in both psychiatry and a primary care specialty, has been less well studied. To date,
two studies have been published describing the type of practice setting
for dually trained physicians. A survey of physicians trained in both
family medicine and psychiatry found that most (60%) practice both
family medicine and psychiatry (McCahill and Palinkas 1997). In con-


Medical Illness and Schizophrenia

trast, psychiatrists also trained in internal medicine reported that only

15% practiced any type of medicine and 75% identified themselves as
psychiatrists (Stiebel and Schwartz 2001). However, neither of these reports examined the patient outcomes of care provided by these dually
trained physicians compared with usual care. Also, both of these studies are based on data that are more than a decade old, and more training
programs are available now than when these data were published.
Dual training programs in the United States are limited. Only seven
training programs are offered in family medicine and psychiatry and 11
in internal medicine and psychiatry, and most of these programs take
only two residents per year. In 2007, a total of 11 graduating medical
students matched into family medicine and psychiatry residencies and
22 matched into internal medicine and psychiatry residencies (National
Resident Matching Program 2008).Therefore, although this model of
training physicians in both psychiatry and a primary care specialty offers the possibility of truly integrated medical and psychiatric care for
patients with serious mental illness, the capacity of the current training
programs is not large enough to provide this kind of care on a large
In summary, two models have been proposed for providing integrated medical and psychiatric care to people with serious mental illness: 1) co-locating primary care physicians in mental health clinics and
2) staffing these clinics with physicians trained in both psychiatry and
a primary care specialty. Druss (2007) suggested that these types of clinics may be best suited for VA system and staff model health maintenance organizations, which use shared funding streams and medical

Challenges to Providing Medical Care for

Patients With Serious Mental Illness
Patient Level
Researchers have offered several explanations as to why people with
schizophrenia receive less health care than the general population. People with schizophrenia may have problems explaining their medical
symptoms to primary care physicians. Physicians may be uncomfortable treating persons with schizophrenia (Lawrie et al. 1996), possibly
reflecting a stigmatization of people with schizophrenia. Similarly, psychiatrists may not feel comfortable providing primary and preventive
health care for their patients (McIntyre and Romano 1977). Fang and

Improving Physical Health Care


Rizzo (2007) used data from a national survey of physicians to examine

whether psychiatrists had less access to medical services for their patients than did other medical specialists. This study compared psychiatrists reports of their ability to obtain medically necessary medical care
for their patients with those of other specialist physicians. The study
found that psychiatrists rated their ability to obtain most types of medical care dramatically lower than did other specialists, in part because
many patients did not have insurance and those who did had barriers
from their health plan.

Policy Level
According to Horvitz-Lennon et al. (2006), challenges to integrating primary health care into specialty mental health treatment exist on three
levels: clinical, organizational, and financial. Clinical challenges include
the lack of training in physical health care by mental health clinicians.
It is important to note that although most psychiatrists do not feel comfortable providing primary medical care, the knowledge of and comfort
with medical problems is even lower among other mental health care
providers, who rarely have formal medical training. Financial challenges include the high rates of uninsurance among persons with serious mental illness and the fact that financing for public mental health
programs is typically separate from that of other public health programs. For example, in the San Diego County public mental health system, approximately 50% of patients with a serious mental illness are
uninsured (Folsom et al. 2005). Furthermore, the application process for
and financial restrictions of the public mental health system are separate and different from those of the safety-net health care program, resulting in a large number of patients receiving mental health treatment
who do not have coverage for medical care.

The growing recognition that people with schizophrenia and other serious mental illnesses have elevated rates of many medical conditions
and often do not receive adequate medical care challenges psychiatrists
and the mental health field to come up with practical approaches to improving their care. The four potential approaches examined in this
chapterhealth care skills training for patients, training psychiatrists
to provide limited primary care, facilitated referral to primary care, and
integration of primary care into mental health clinicsare all supported by some data. However, none of these approaches currently is

Medical Illness and Schizophrenia


widespread, and no single approach is likely to provide a solution. In

an ideal world, psychiatrists would help ensure that their patients receive high-quality physical health care; staff at mental health clinics
would help patients who need health care; mental health clinics would
offer onsite primary care; and the financial barriers to providing integrated physical and mental health care would be removed.

Key Clinical Points

Patients with schizophrenia and other serious mental illnesses receive

less primary and preventive care than needed.

Guidelines alone are unlikely to result in improved health care for patients with serious mental illness.

Prior studies have examined four ways of improving health care for
people with serious mental illness: health care skills training for patients, training psychiatrists to provide limited primary care, facilitated
referral to primary care, and providing onsite primary care in mental
health clinics.

Integrating primary care into mental health clinics may require significant changes in how mental health care is funded and delivered.

Abilify (package insert). Princeton, NJ, Bristol-Myers Squibb, 2008
American Diabetes Association, American Psychiatric Association, American
Association of Clinical Endocrinologists, et al: Consensus Development
Conference on Antipsychotic Drugs and Obesity and Diabetes. Diabetes
Care 27:596601, 2004
Bartels SJ, Forester B, Mueser KT, et al: Enhanced skills training and health care
management for older persons with severe mental illness. Community
Ment Health J 40:7590, 2004
Clozaril (package insert). East Hanover, NJ, Novartis Pharmaceuticals Corp., 2008
Daumit GL, Crum RM, Guallar E, et al: Outpatient prescriptions for atypical antipsychotics for African Americans, Hispanics, and whites in the United
States. Arch Gen Psychiatry 60:121128, 2003
Dickerson FB, McNary SW, Brown CH, et al: Somatic healthcare utilization
among adults with serious mental illness who are receiving community
psychiatric services. Med Care 41:560570, 2003

Improving Physical Health Care


Dobscha SK, Ganzini L: A program for teaching psychiatric residents to provide

integrated psychiatric and primary medical care. Psychiatr Serv 52:1651
1653, 2001
Dobscha SK, Snyder KM, Corson K, et al: Psychiatry resident graduate comfort
with general medical issues: impact of an integrated psychiatry-primary
medical care training track. Acad Psychiatry 29:448451, 2005
Druss BG: Improving medical care for persons with serious mental illness: challenges and solutions. J Clin Psychiatry 68:4044, 2007
Druss BG, Bradford WB, Rosenheck RA, et al: Quality of medical care and excess mortality in older patients with mental disorders. Arch Gen Psychiatry
58:565572, 2001a
Druss BG, Rohrbaugh RM, Levinson CM, et al: Integrated medical care for patients with serious psychiatric illness. Arch Gen Psychiatry 58:861868,
Fang H, Rizzo JA: Do psychiatrists have less access to medical services for their
patients? J Ment Health Policy Econ 10:6371, 2007
Folsom DP, Okereke OI: Medical comorbidity in geriatric psychiatry. Am J
Geriatr Psychiatry 13:1, 2005
Folsom D, McCahill M, Bartels S, et al: Medical comorbidity and receipt of medical care by older homeless people with schizophrenia or depression. Psychiatr Serv 53:14561460, 2002
Folsom DP, Hawthorne W, Lindamer L, et al: Prevalence and risk factors for
homelessness and utilization of mental health services among 10,340 patients with serious mental illness in a large public mental health system.
Am J Psychiatry 162:370376, 2005
Geodon (package insert). New York, NY, Pfizer, Inc., 2008
Golomb BA, Pyne JM, Wright B, et al: The role of psychiatrists in primary care
of patients with severe mental illness. Psychiatr Serv 51:766773, 2000
Harrington C, Gregorian R, Gemmen E, et al; for The Lewin Group: Access and utilization of new antidepressant and antipsychotic medications. January 2000.
Available at:
index.htm#TOC. Accessed October 8, 2008.
Harvey SB, Newton A, Moye GA: Physical health monitoring in schizophrenia:
the use of an invitational letter in a primary care setting. Primary Care and
Community Psychiatry 10:7174, 2005
Henderson DC, Cagliero E, Copeland PM, et al: Glucose metabolism in patients
with schizophrenia treated with atypical antipsychotic agents: a frequently
sampled intravenous glucose tolerance test and minimal model analysis.
Arch Gen Psychiatry 62:1928, 2005
Horvitz-Lennon M, Kilbourne AM, Pincus HA: From silos to bridges: meeting
the general health care needs of adults with severe mental illnesses. Health
Aff (Millwood) 25:659669, 2006
IMS Health: Atypical antipsychoticsgenerating evidence to inform policy and
practice. 2002. Available at:
research_schizophrenia.htm. Accessed May 15, 2006.


Medical Illness and Schizophrenia

Jin H, Meyer JM, Jeste DV: Atypical antipsychotics and glucose dysregulation:
a systematic review. Schizophr Res 71:195212, 2004
Koran LM, Sox HC, Marton KI, et al: Medical evaluation of psychiatric patients:
results in a state mental health system. Arch Gen Psychiatry 46:733740,
Lawrie SM, Parsons C, Patrick J, et al: A controlled trial of general practitioners
attitudes to patients with schizophrenia. Health Bull 54:201203, 1996
Leslie DL, Rosenheck RA: Incidence of newly diagnosed diabetes attributable
to atypical antipsychotic medications. Am J Psychiatry 161:17091711, 2004
Lieberman JA, Stroup TS, McEvoy JP, et al: Effectiveness of antipsychotic drugs
in patients with chronic schizophrenia. N Engl J Med 353:12091223, 2005
Marder SR, Essock SM, Miller AL, et al: The Mount Sinai Conference on the
Pharmacotherapy of Schizophrenia. Schizophr Bull 28:516, 2002
McCahill ME, Palinkas LA: Physicians who are certified in family practice and
psychiatry: who are they and how do they use their combined skills? J Am
Board Fam Pract 10:111115, 1997
McCreadie RG; Scottish Schizophrenia Lifestyle Group: Diet, smoking and cardiovascular risk in people with schizophrenia: descriptive study. Br J Psychiatry 183:534539, 2003
McIntyre JS, Romano J: Is there a stethoscope in the house (and is it used)? Arch
Gen Psychiatry 34:11471151, 1977
McKibbin CL, Patterson TL, Norman G, et al: A lifestyle intervention for older
schizophrenia patients with diabetes mellitus: a randomized controlled
trial. Schizophr Res 86:3644, 2006
Miller CL, Druss BG, Rohrbaugh RM: Using qualitative methods to distill the
active ingredients of a multifaceted intervention. Psychiatr Serv 54:568
571, 2003
Munk-Jrgensen P, Mors O, Mortensen PB, et al: The schizophrenic patient in
the somatic hospital. Acta Psychiatr Scand Suppl 407:9699, 2000
National Resident Matching Program. Available at:
data/resultsanddata2007.pdf. Accessed August 12, 2008.
Phelan M, Stradins L, Morrison S: Physical health of people with severe mental
illnesses. BMJ 322:443444, 2001
Risperdal (package insert). Titusville, NJ, Janssen Pharmaceutica, Inc., 2008
Samet JH, Larson MJ, Horton NJ, et al: Linking alcohol- and drug-dependent
adults to primary medical care: a randomized controlled trial of a multidisciplinary health intervention in a detoxification unit. Addiction 98:509
516, 2003
Schneider LS, Dagerman KS, Insel P: Risk of death with atypical antipsychotic
drug treatment for dementia: meta-analysis of randomized placebocontrolled trials. JAMA 294:19341943, 2005
Seroquel (package insert). Wilmington, DE, AstraZeneca Pharmaceuticals, LP,
Shore JH: Psychiatry at a crossroad: our role in primary care. Am J Psychiatry
153:13981403, 1996

Improving Physical Health Care


Snyder K, Dobscha SK, Ganzini L, et al: Clinical outcomes of integrated psychiatric and general medical care. Community Ment Health J 44:147154, 2008
Stiebel V, Schwartz CE: Physicians at the medicine/psychiatric interface: what
do internist/psychiatrists do? Psychosomatics 42:377381, 2001
Weber M, Wyne K: A cognitive/behavioral group intervention for weight loss
in patients treated with atypical antipsychotics. Schizophr Res 83:95101,
Wells KB, Sherbourne C, Schoenbaum M, et al: Impacts of disseminating quality improvement programs for depression in managed primary care: a randomized controlled trial. JAMA 283:3204, 2000
Wu RR, Zhao JP, Jin H, et al: Lifestyle intervention and metformin for treatment
of antipsychotic-induced weight gain: a randomized controlled trial.
JAMA 299:185193, 2008
Zyprexa (package insert). Indianapolis, IN, Eli Lilly, Inc., 2008

This page intentionally left blank

Excessive Mortality
and Morbidity
Associated With
Daniel E. Casey, M.D.
Thomas E. Hansen, M.D.

Psychiatric conditions have long been recognized to be associated with premature mortality. In his review, Brown (1997) noted that
lunatics (perhaps 25% having schizophrenia) in a study from 1841 experienced mortality at three to 14 times the rate seen in the general population. According to Brown, Eugen Bleuler reported that mortality in
patients with schizophrenia was 1.4 times the expected rate, caused by
accidents, suicide, refusing food, infectious disease, and other diseases
from poor sanitation. Brown stated that although Emil Kraepelin considered mortality to be only slightly increased in dementia praecox, he
reported that suicide, negativism, diet, and poor cooperation with medical care contributed to increased mortality. Studies from the first half
of the twentieth century found mortality rates in psychiatric patients to
be two and four times higher than in age- and gender-matched populations (Brown 1997). Hazards of institutional care (such as tuberculosis
and gastrointestinal disease) added to risks associated with the underlying psychiatric illness. Unnatural causes (suicide and accidental


Medical Illness and Schizophrenia

death) became an increasingly important cause of premature death

with the shift from inpatient to outpatient care for patients with schizophrenia (Black et al. 1985). More recently, cardiovascular disease and
diabetes may be increasing in importance as causes of death and morbidity in these patients (Jeste et al. 1996).
This chapter updates our earlier review (Casey and Hansen 2003) of
mortality and morbidity associated with schizophrenia. Data clearly demonstrate that the life span for patients with schizophrenia is shortened.
Many conditions leading to premature death are evident during patients
lifetimes and account for the majority of the excess morbidity seen in patients with schizophrenia. Both patient behaviors and societal aspects of
care for schizophrenia can contribute to this increase in morbidity.

Life Span and Mortality in Patients

With Schizophrenia
Mortality Rates
Reports on mortality rates commonly use record linkage methodology:
databases of all deaths in a region are linked with population registers
of psychiatric patients (community or hospital derived) so that the
number of deaths in a group of patients with a specific diagnosis can be
determined. This observed number is contrasted to the number expected based on life expectancy data for people of the same age and
gender applied to the time period studied. The ratio of observed to expected cases is the standardized mortality ratio (SMR), sometimes reported as the ratio multiplied by 100. The 95% confidence interval (CI)
values indicate statistical significance, with lower CI values above 1.0
(or 100, when the SMR is multiplied by 100) supporting significance for
elevated SMR, and upper CI values below 1.0 (or 100) supporting significance for decreased SMR (Harris and Barraclough 1998). In this
chapter, for ease of comparison, we report SMR and CI data without
multiplying by 100, even if so reported in cited references.
Reviews of mortality rates consistently find increased death rates in
patients with schizophrenia. This consistency is remarkable when one
considers variation in methods across individual studies and reviews.
For instance, individual studies from earlier periods may be overly influenced by inclusion of hospitalized patients (lower risk of suicide,
higher risk of diseases related to institutionalization), whereas later
studies are likely to include outpatients (greater risk of suicide) and to
be more precise in diagnosis.

Excessive Mortality and Morbidity


Three reviews that summarized results from studies conducted between 1942 and 1996 found the mortality rate in patients with schizophrenia to be between 1.5 and 2 times the expected rate. Allebeck (1989)
reviewed nine studies published between 1942 and 1985 in which patients had been followed at least 5 years and found mortality about
twice the expected rate. He noted difficulty in ascertaining whether
rates had changed because diagnostic and study criteria had varied
over time. Brown (1997) examined studies published between 1986 and
1996 with schizophrenia patient populations of at least 100 individuals
recruited after 1952 that included observed and expected death rates
and that had follow-up periods of at least 2 years, with losses to followup of less than 15%. He found an aggregate SMR of 1.51 (95% CI 1.48
1.54). The SMR for suicide was elevated (8.38, 95% CI 7.848.94), as were
the SMR for accidents (2.16, 95% CI 1.962.36) and the SMR for natural
causes of death (1.1, 95% CI 1.051.15). Brown interpreted data about
specific natural causes of death conservatively, stating that patients
with schizophrenia die from the same diseases as the rest of the population. He noted that peptic ulcer (perhaps related to alcohol use) and
pneumonia (probably associated with elderly institutionalized patients) caused deaths at higher than expected rates. Harris and Barraclough (1998) reviewed mortality rates in cohorts of patients with
various mental disorders between 1966 and 1995 (20 studies, nine countries, almost 36,000 patients) and found a death risk in patients with
schizophrenia that was 1.6 times expected. The death risk from unnatural causes (suicide and violence) was 4.3 times that expected, whereas
death from natural causes was 1.4 times that expected, but natural
causes accounted for 68% of the excess in expected deaths.
In our earlier review (Casey and Hansen 2003), we described three
studies published in 20002001, with data on patients studied between
1973 and 1995, in which mortality rates were reported to be somewhat
higher than in the earlier reviews (significantly increased relative risk
or SMR of 2.43.8, depending on patient gender). A meta-analysis by
Saha et al. (2007) confirms that SMRs for schizophrenia are increasing;
the analysis reviewed 37 articles published between 1980 and 2006 from
25 countries and found the overall SMR to be 2.58. When studies were
grouped by midpoint of the follow-up period, SMR increased from 1.84
in the 1970s, to 2.98 in the 1980s, and to 3.2 in the 1990s. A study not included in the meta-analysis from rural China (Ran et al. 2007) followed
510 patients with schizophrenia for the 10-year period of 19942004 and
further demonstrated higher mortality in more recent studies (overall
SMR of 4.0, 95% CI 2.45.8).


Medical Illness and Schizophrenia

Life expectancy and survival curve data offer further confirmation of

increased mortality associated with schizophrenia. In our earlier review
(Casey and Hansen 2003), we noted that in three studies based on 5- to
10-year follow-up of patients, authors projected decreased life expectancy of 816 years in general and of 34 years for patients ages 60 years
and older. Three more recent studies also utilize survival data. Patients
with first admissions for psychosis in Suffolk County, New York, between 1989 and 1995 were evaluated for survival rates at 5 and 10 years
using survival curves; survival rates were 96.9% at 5 years and 94.5% at
10 years for the 235 patients with schizophrenia or schizoaffective disorder, with 60% of deaths resulting from unnatural causes (Craig et al.
2006). In a 10-year (19811991) mortality study of 255 patients with
schizophrenia from Uppsala, Sweden, 23% of patients died compared
with 11% of the 1,275 controls (P=0.003) (Fors et al. 2007). Patients with
schizophrenia (n =319) seen at the Mayo Clinic from 19501980 were followed up to 2005 (mean 23.5 years); 44% died during this time, a rate
higher (P <0.001) than for a comparable U.S. population (the specific
rate was not reported, but appears to be 34% from the survival curve
comparison) (Capasso et al. 2008).
Thus, data from the last 50 years demonstrate that the increase in
mortality seen for patients with schizophrenia in earlier time periods
persists. Furthermore, the disparity in death rates between patients
with schizophrenia and the general population appears to be widening.

Suicide and Other Unnatural Causes of Mortality

Reviews consistently find that patients with schizophrenia die from
suicide at increased rates compared with the general population (Allebeck 1989; Harris and Barraclough 1998), although underreporting is a
possibility because suicides can be incorrectly reported as resulting
from accidents or undetermined causes (Allebeck 1989). In this section,
we focus on suicide as an unnatural cause of death given the lower mortality rates from accidents. For instance, one review reported SMRs of
8.38 for suicide and 2.16 for accidents (Brown 1997). In a Danish register
study of patients admitted between 1973 and 1993 who died by 1993,
Hiroeh et al. (2001) reported for males an SMR of 10.7 for suicide compared with 2.1 for accidents, and for females an SMR of 10.8 for suicide
compared with 2.9 for accidents.
In our earlier review (Casey and Hansen 2003), we summarized data
from multiple studies; SMR values generally ranged from 12 to 20, except
SMR of 1.32 in one study (Mortensen and Juel 1990) that included mostly
elderly institutionalized patients from 1957 to 1986. Several recent stud-

Excessive Mortality and Morbidity


ies have found suicide SMRs in the higher range. The rate of suicide in
the general population and in patients with schizophrenia in Denmark
declined between 1981 and 1997, but the incidence ratio remained 20
times higher for patients with schizophrenia (Nordentoft et al. 2004).
Even correcting for social and demographic differences, the rate remained high at 12.26. A French 10-year follow-up study starting in 1993
(Limosin et al. 2007) found the suicide SMR for males to be 15.8 (95% CI
13.718.2) and for females to be 17.7 (95% CI 13.623.1). In both of these
studies, suicide rates were highest within the first year of follow-up.
The suicide mortality data support several observations. The increase in risk from suicide for patients with schizophrenia was much
lower many years ago when patients were commonly kept in hospitals
for long time periods. The risk is greatest early in the course of illness
and close to the hospital discharge, so that mortality rates vary between
studies depending in part on study inclusion criteria. Finally, the risk of
suicide appears to be increasing, perhaps related to greater reliance on
less intensive treatment emphasizing outpatient care.
A national psychological autopsy study from Finland provides important details about the nature of suicide by patients with schizophrenia (Heila et al. 1997). All suicides occurring during a 1-year period
during 1987 and 1988 were identified. Providers and next of kin were
interviewed, and adequacy of medication was examined. This sample
had a somewhat older age of suicide (mean 40 years old) than hospitaltreated or initially treated patient groups. The mean illness duration
was 15.5 years, and 75% of patients had active symptoms. About 9%
had current suicide-command hallucinations. Twenty-five (27%) were
hospitalized at the time of suicide, and 29 (32%) had been discharged
within 3 months. A depressive syndrome had been present during either a residual or an active phase of psychosis in 61 patients (66%), with
depressive symptoms at the time of suicide in 59 (64%). Drug overdose
was the most common single cause of death (34 patients, 37%), although various violent means were used in 37 cases (40%). Of 34 drug
overdoses, 27 (79%) were with neuroleptic medication and 25 (74%)
specifically involved low-potency neuroleptics. The authors suspected
that with the shift toward outpatient management, the availability of
drugs to use in overdoses has increased. Additional data on risk factors
and adequacy of treatment were reviewed for the 88 patients who were
in treatment at the time of suicide (Heila et al. 1999). Data on compliance and treatment responsiveness were available for about 72 patients.
For the entire group of patients who committed suicide (either with active psychotic symptoms or during a residual phase of illness), only 34
(47%) appeared to be compliant and to be receiving an adequate dosage


Medical Illness and Schizophrenia

of antipsychotic medication. Data were similar when the patients in active phase of illness were examined separately. Only 26 (46%) of the 57
patients experiencing active symptoms at the time of suicide were receiving an adequate dosage (defined as 700 mg daily in chlorpromazine
equivalents) and were compliant in taking medication. These two studies suggest the importance of effective medication management and
provision of appropriate intensity of treatment (if not inpatient, then
close outpatient observation) in this era of managed care. The French
study cited above also notes the importance of substance abuse (Limosin et al. 2007).
Estimates of the lifetime risk of suicide, defined as the proportion of
all deaths to which suicide contributes, provide an additional perspective on the importance of suicide in schizophrenia-associated mortality.
Data from a large meta-analysis (Harris and Barraclough 1998) were analyzed by plotting the percentage of patients who had died against the
percentage of patients who had died by suicide. The curve that best
modeled the data was extrapolated to the point at which all members of
the cohort would be deceased; the percentage of deaths by suicide was
estimated to be 4% of all deaths, which is lower than the frequently
cited value of 10% lifetime risk for suicide in patients with schizophrenia (Inskip et al. 1998). This finding was confirmed in a more recent
meta-analysis (Palmer et al. 2005). Using both proportionate mortality
and case fatality statistics from 61 studies to maximize accuracy in estimating suicide risk, the authors concluded that 4.9% of patients with
schizophrenia will commit suicide, usually near the onset of illness.
An emphasis on suicide rates may be misleading in terms of absolute
risk, because suicide is rare and natural causes of death are far more
common in the general population. Roughly two-thirds of excess
deaths in patients with schizophrenia are by causes other than suicide,
illustrating that although many patients with schizophrenia die from
suicide, far more die from other causes. For example, suicide accounted
for only 38% of the excess mortality in one review (Harris and Barraclough 1998) and 35% of excess mortality in another (Brown 1997). In a
Southampton, England, cohort treated in 19811982 with 13 years of
follow-up, suicide and other unnatural causes accounted for 33% of the
deaths (Brown et al. 2000). Another way to understand this issue is to
note that although the relative risk for suicide in patients with schizophrenia compared with the general population is higher than the relative risk for natural causes of death, the absolute risk for natural causes
of death is far higher than for suicide in patients with schizophrenia.
Thus, excess mortality in schizophrenia derives primarily from natural
causes rather than from suicide. (See Table 21.)

Excessive Mortality and Morbidity



Evidence for increased mortality associated with


Increased overall mortality rates in patients with schizophrenia:

1.54 times expected
Decrease in longevity from survival curves: 816 years
Increased mortality rates for suicide: 12 to more than 20 times
Lifetime death from suicide: about 5%
Increased standardized mortality ratio from natural causes: at least
2 times expected (varies by disorder)

Natural Causes of Death in Patients With Schizophrenia

An autopsy study from the time when phenothiazine medications were
new, roughly 50 years ago, provided a view of death by natural causes
(Hussar 1966). Autopsy data from 1,275 chronically hospitalized patients with schizophrenia who died at age 40 or older found that heart
disease and cancer were the most common causes of death (similar to
findings in the general population age 40+ years of that time). Pneumonia was somewhat overrepresented as a cause of death. Undetermined
causes and aspiration of food were in the top eight causes of death in
the patients with schizophrenia but not in the general population. The
authors speculated that phenothiazines could be responsible, although
they noted that other investigators at the time found no association between the new medications and death rates.
More recent studies quantify increased mortality for various natural
causes in patients with schizophrenia. Harris and Barraclough (1998)
reviewed mortality rates in cohorts of patients with various mental disorders between 1966 and 1995. Cohorts of patients with schizophrenia
had an SMR of 1.37 for natural causes, with significant increases in expected deaths from a variety of medical conditions (infectious, endocrine, circulatory, respiratory, digestive, and genitourinary disorders).
In our previous review (Casey and Hansen 2003), we discussed several
subsequent studies that reported somewhat higher values for diseasespecific SMR (1.92.3).
Reviewing individual studies for increased risk of specific disorders
presents challenges. Many studies report categories of diseases from
the International Classification of Diseases (ICD) rather than specific disorders, not all studies report on the same categories of illness, and some
illnesses could fall into more than one category. Also, findings vary between studies, perhaps because the SMR is greatly influenced by a


Medical Illness and Schizophrenia

small number of observed cases when the expected rate is low. Still, examining individual studies may be helpful in understanding risks for
more recent periods, because reviews are likely to include older studies
that reflect risks that are no longer as common (e.g., institutionalization
and infectious diseases). Two recent Danish register studies of patients
with schizophrenia covering overlapping time periods between 1973
and 2001 found significant increases in SMR (with some variation by
gender) on the order of 1.442.98 for cardiovascular disease, cancer, respiratory diseases, endocrine conditions, gastrointestinal and genitourinary diseases, nervous system illnesses, and infectious diseases (Hiroeh
et al. 2008; Laursen et al. 2007). (See Table 22.)
Cancer, cardiovascular disease, and diabetes warrant additional discussion. The failure to demonstrate increased risk of cancer in patients
with schizophrenia when all causes of cancer are aggregated has led to
much speculation, including the possibility that having schizophrenia
may convey protection from cancer. Consistent and increasing findings
of both numbers of deaths from, and risk for, disorders such as cardiovascular disease and diabetes are of current interest.
Cancer incidence and mortality rates reported for patients with
schizophrenia range from decreased to increased, varying by type
of cancer, gender, and country, as well as across studies. In a U.S. population report (Cohen et al. 2002) using a mortality follow-back database that sampled 1% of all deaths from 1986, cancer was the cause of
25 of the 130 deaths in patients with schizophrenia. The unadjusted
odds ratio indicated a decreased risk for cancer (0.62, 95% CI 0.40.96).
Overall, the patients with schizophrenia died at a younger age (55.6 vs.
63.7 years, P< 0.001); adjusting for age and other factors yielded an odds
ratio of 0.59 (95% CI 0.380.93). A Danish register study (Dalton et al.
2005) of patients admitted with schizophrenia between 1969 and 1995
and followed through 1995 found no difference in overall cancer risk,
but differences were seen based on gender and location of cancer. Men
had decreased risks of prostate, rectal, and nonmelanoma skin cancer,
and a nonsignificant decrease in lung cancer, whereas women had an
increased risk of breast cancer. Other authors have noted offsets in mortality risk by location. A decreased rate of lung cancer in males was
noted to offset an increased rate of breast cancer in females (Mortensen
and Juel 1990) and digestive cancers (Newman and Bland 1991). In the
latter study, the SMR for digestive disorders was 1.7 (P<0.05) and for
lung cancer was 0.7 (NS). Limited access to cigarettes for institutionalized patients has been raised as a possible explanation for reduced rates
of lung cancer (Harris and Barraclough 1998).

Excessive Mortality and Morbidity


TABLE 22. Conditions associated with increased mortality from


Cardiovascular disease
Respiratory disorders
Gastrointestinal diseases
Central nervous system disorders
Genitourinary diseases
Infectious diseases
Cancer (controversial; breast cancer may be increased and lung
cancer decreased)

Two incidence studies on cancer in patients with schizophrenia illustrate varying results between geographic locations for different types of
cancer. The authors of the World Health Organizations three-cohort
study (Gulbinat et al. 1992) concluded that patients with schizophrenia
do not demonstrate any consistent increase or decrease in risk for cancer. In contrast, the authors of a records linkage study using the Finnish
cancer and psychiatric illness registries found a modest but significantly elevated standardized incidence ratio (SIR) for cancer (1.17,
95% CI 1.091.25) in patients with schizophrenia (Lichtermann et al.
2001), with the incidence of lung cancer twice as high in patients with
schizophrenia (SIR 2.17, 95% CI 1.782.60). They also noted that the incidence of gallbladder cancer was elevated (SIR 2.07, 95% CI 1.033.70)
and wondered about the role of obesity and poor diet in patients with
schizophrenia, given that these are risk factors for gallstones, and hence
possibly for gallbladder cancer.
An investigation of possible lower rates of lung cancer death in patients
with schizophrenia (Masterson and OShea 1984) led to speculation that
phenothiazine medication might have antitumor activity. However, an alternative explanation may be that patients with schizophrenia die from
other causes, such as cardiovascular disease, before they reach the expected age of death from lung cancer. Arguing against the hypothesis that
phenothiazines have antitumor activity, the authors found that breast
cancer caused proportionately more deaths in the patients with schizophrenia (8.3% vs. 3.4% in the general population, P<0.04). The authors
also speculated that being nulliparous and having elevated prolactin,
both associated with schizophrenia illness and treatment, while practicing inadequate breast care (insufficient self-examinations and incomplete
reporting of lumps) could have caused the increase in breast cancer.


Medical Illness and Schizophrenia

Whether patients with schizophrenia have an increased risk of breast

cancer remains unresolved. Two reports using similar Danish register
data analysis methods and time periods came to different conclusions
about breast cancer risk (Dalton et al. 2003, 2005). As part of a larger
study of all types of cancer in men and women, female patients
(N= 9,743) admitted for schizophrenia between 1969 and 1993 had a
breast cancer SIR of 1.2 (95% CI 1.051.38) (Dalton et al. 2005). In another study of women who developed breast cancer between 1970 and
1997, patients with schizophrenia were identified (N=7,541) and found
to have decreased risk (relative risk 0.97, 95% CI 0.761.20). When controlled for fertility issues, the relative risk fell to 0.90 (95% CI 0.711.12).
The authors noted that this finding is not surprising because women
with schizophrenia are more likely to be nulliparous, a risk factor for
breast cancer. They also speculated that alcohol use, another risk factor
for breast cancer (not controlled for in their study), may contribute to
the apparent increase in risk for breast cancer reported in other studies
(Dalton et al. 2003).
Researchers who study disease-related mortality rates often report
on endocrine and metabolic conditions without specifically mentioning
diabetes, although one might expect it to be the most common endocrine condition. For instance, in a British study of patients followed
from 1981 to 1994 (Brown et al. 2000), the SMR for all endocrine disorders was elevated at 11.66 (95% CI 3.7927.21) and the diabetes SMR
was 9.96 (95% CI 2.0529.11), indicating that most of the endocrine mortality was accounted for by diabetes. Two recent studies found lower
but significant elevations in mortality from all endocrine disorders. In a
Danish register study of patients between 1973 and 1993 (Hiroeh et al.
2008), the endocrine SMR was 1.99 (95% CI 1.372.91) for females and
2.29 (95% CI 1.583.32) for males. In another Danish study (Laursen et
al. 2007) covering 19732001, the rate was higher for males (SMR 3.56,
95% CI 2.684.73) and lower for females (SMR 1.69, 95% CI 1.142.5).
Given the accepted increased risk of diabetes associated with some
atypical antipsychotic medications (American Diabetes Association et
al. 2004), one should expect increasing mortality from diabetes in the future. However, some of the increase may instead appear as increased
cardiovascular mortality.
Cardiovascular disease causes a large number of deaths in both the
general population and patients with schizophrenia. In the review by
Harris and Barraclough (1998), circulatory diseases caused 2,313 deaths
out of the 5,591 deaths from natural causes, leading to an SMR in males
of 1.1 (95% CI 1.041.16) and in females of 1.02 (NS, 95% CI 0.961.08).
Although the increase in risk is relatively small, the number of excess

Excessive Mortality and Morbidity


deaths is substantial (230 more than expected compared with a total of

excess deaths from natural causes of 1,250). Subsequent studies found
higher cardiovascular mortality rates. In one study (Brown et al. 2000),
circulatory disease had an SMR of 2.49 (95% CI 1.643.63); more specifically, cardiovascular disease had an SMR of 1.87 (95% CI 1.022.98).
The Danish register studies noted comparable rates of 1.62.1. In their
study involving patients from 1973 to 1993, Hiroeh et al. (2008) found
the circulatory disorder SMR in females to be 1.61 (95% CI 1.471.76)
and in males to be 1.91 (95% CI 1.752.08), whereas Laursen et al. (2007),
in their 19732001 study, found the cardiovascular disease SMR in
females to be 1.72 (95% CI 1.531.93) and in males to be 2.07 (95% CI
With rising mortality rates related to natural causes, especially diabetes and cardiovascular disease, increased attention to morbidity and
health behaviors in patients with schizophrenia is warranted. In this
chapter, we touch on these topics only briefly because they are covered
in detail in subsequent chapters of this volume (Chapter 4, Obesity
and Schizophrenia; Chapter 5, Glucose Intolerance and Diabetes in
Patients With Schizophrenia; Chapter 7, The Spectrum of Cardiovascular Disease in Patients With Schizophrenia; and Chapter 9, Nicotine and Tobacco Use in Patients With Schizophrenia).

Morbidity in Schizophrenia
Medical Conditions
The increased rate of mortality from natural causes associated with
schizophrenia should be reflected in higher rates of morbidity. In our
chapter (Casey and Hansen 2003) in the previous edition of this book,
we discussed reviews and studies that found medical conditions in
35%70% of patients with schizophrenia. The disorders included diabetes, hyponatremia, thyroid disorder, urinary tract infection, bladder
dysfunction, hypertension, liver disease, seizures, and visual problems,
among others.
Results from the interview survey study of patients with schizophrenia from the Patient Outcomes Research Team program (Dixon et al.
1999) illustrate increased medical morbidity. Of 719 patients, 469 (65%)
reported having at least one lifetime medical condition, 256 (36%) had
more than one condition, and 343 (48%) said the medical condition was
active. The authors did not have a comparison control group but noted
that the rates for diabetes, hypertension, and cardiac diseases were all
higher than rates reported for these conditions in the National Health


Medical Illness and Schizophrenia

Interview Survey. Not surprisingly, the number of medical comorbidities correlated significantly with patient self-rating of physical health.
Jeste et al. (1996) reviewed reports on medical comorbidity in schizophrenia; for unclear reasons, rheumatoid arthritis seems to occur only
rarely in patients with schizophrenia, whereas rates of cardiovascular
disease and diabetes mellitus appear to be increased. The authors reported that studies vary in finding increased rates of other specific illnesses and that the impression of decreased risk of cancer does not
appear to be valid. In the first of two morbidity studies of middle-aged
to elderly patients, the authors compared patients with schizophrenia
with a group of patients with depression and another group with
Alzheimers disease; lower rates of physical illnesses were found in the
patients with schizophrenia. In the second study, 45 patients with
schizophrenia were compared with 38 normal control subjects; patients
in both groups were older than age 45 years. Indications of morbidity
were comparable, but the control group was about 10 years older than
the patients with schizophrenia, suggesting that the schizophrenia
groups rate of morbidity was comparable to that of an older population. The authors noted that patients with schizophrenia endorsed
more medical concerns when questioned in a structured manner.
The expected increase in morbidity in patients with schizophrenia
was also seen in a case-control study of hospitalizations in Stockholm,
Sweden (Dalmau et al. 1997). The authors gathered data regarding
treatment of somatic conditions in 775 schizophrenia patients and 775
case-matched control subjects regardless of whether the treatment occurred in a psychiatric or a general medical ward. The schizophrenia
group had 523 (67%) admissions compared with 373 (48%) for the control group over the 15-year study period (McNemar test, P=0.000). An
increased odds ratio was found for almost all medical disorders occurring in patients with schizophrenia, with the exception of genitourinary, musculoskeletal, and sensory organ diseases. Tumors (combined
malignant and benign) were found twice as often in patients with
In summary, data demonstrate increased morbidity from medical
conditions in patients with schizophrenia. Although almost all disorders can occur at higher than expected rates, cardiovascular disease and
diabetes are particular concerns.

Substance Abuse
Patients with schizophrenia have increased rates of all types of
substance abuse (Dixon 1999), as is covered in detail in Chapter 11,

Excessive Mortality and Morbidity


Substance Abuse and Schizophrenia. The adverse psychiatric effects

(Dixon 1999) combined with medical problems caused by substance
abuse lead to the expectation of increased morbidity and mortality in
patients with schizophrenia who abuse drugs or alcohol. Not surprisingly, Dalmau et al. (1997) found that inclusion of schizophrenia patients with substance abuse in their study cohort increased the number
of somatic conditions that occurred more frequently in schizophrenia
patients than in a nonpsychiatric control population. Also, results from
a study of diabetes mortality in patients with co-occurring psychotic
and substance use disorders suggested that diabetes may be increased
in this group and be associated with a high mortality rate (Jackson et al.

Health Maintenance in
Patients With Schizophrenia
Treatment Concerns
Treatment during the first half of the twentieth century undoubtedly
contributed to morbidity and mortality in schizophrenia. Physical interventions such as leucotomy, insulin coma therapy, and cardiazolinduced convulsive therapy all carried substantial risk (Brown 1997).
In the past 50 years, morbidity and mortality concerns related to treatment have shifted to medications, with recent attention focused on
extrapyramidal side effects, cardiac conduction, and metabolic side effects (American Diabetes Association et al. 2004; Fontaine et al. 2001). In
a 10-year follow-up study of mortality and medication use in patients
with schizophrenia, Waddington et al. (1998) found that antipsychotic
polypharmacy and absence of cotreatment with anticholinergic medication significantly increased mortality. No clear explanation related to
other factors that might covary with polypharmacy seemed plausible,
leaving the authors to wonder whether concurrent use of multiple antipsychotic medications causes some adverse biological consequence,
or whether unrecognized subclinical drug-induced parkinsonism could
contribute to mortality. A 19781980 Finnish study also found that
increased mortality correlated with the number of prescribed typical
antipsychotic medications (Joukamaa et al. 2006). Certainly, extrapyramidal side effects, including tardive dyskinesia, cause substantial morbidity in patients with schizophrenia (Casey 1993; Hansen et al. 1997).
For further reading on the contribution of atypical antipsychotic medications to various metabolic problems, please see other chapters in this


Medical Illness and Schizophrenia

volume: Chapter 4, Obesity and Schizophrenia; Chapter 5, Glucose

Intolerance and Diabetes in Patients With Schizophrenia; Chapter 7,
The Spectrum of Cardiovascular Disease in Patients With Schizophrenia; and Chapter 9, Nicotine and Tobacco Use in Patients With

Patient Health Habits and Related Concerns

Behaviors such as abuse of drugs and alcohol, smoking, lack of exercise,
and dietary indiscretion appear to contribute to mortality in the general
population, and thus are likely to do so in patients with schizophrenia
(Brown et al. 1999). Brown et al. (2000) noted that the SMR related to
natural causes of death was significantly elevated in smokers (3.6, 95%
CI 2.74.71) but not in nonsmokers (1.78, 95% CI 0.853.28). The authors
reviewed apparently avoidable natural causes of death in these patients; causes included failure to recognize medical disease by the patient or care provider (38 cases), missed diagnoses (3 cases), poor
treatment compliance (unable to quantify), treatment refusal (2 cases),
and inadequate social support (1 case).
Brown et al. (1999) directly investigated lifestyle concerns in 102 patients with schizophrenia. Patients with schizophrenia ate a diet significantly higher in fat and lower in fiber (significant for males, trend for
females) than the reference population. Roughly one-third of the patients with schizophrenia reported doing no exercise, and no patient reported doing strenuous exercise (comparison rates not provided). In
their sample, female patients demonstrated a trend toward obesity
(P= 0.09). The rate of smoking was higher in patients with schizophrenia than in the comparison sample (68% vs. 28% in males, 57% vs. 25%
in females); alcohol consumption was decreased in males and unexceptional in females.
A study of 22 outpatients with schizophrenia conducted in 1992
1993 found similar results (Holmberg and Kane 1999): patients with
schizophrenia were less likely to practice health-promoting behaviors
than were nonpsychiatric populations. In a study of smoking in Irish inpatients with schizophrenia (Masterson and OShea 1984), the rate of
smoking was higher than for the general population (males 84% vs.
41%, females 82% vs. 36%), as were daily consumption of cigarettes,
preference for medium- to high-tar cigarettes (59% vs. 1%), and duration of smoking (longer than 16 years, 80% vs. 56%).
The sum of the evidence indicates clearly that patients with schizophrenia engage in behaviors that can be expected to increase morbidity
and mortality.

Excessive Mortality and Morbidity


Access to Health Care

One can easily imagine that decreased access to either medical or psychiatric care could exacerbate morbidity and contribute to increased
mortality in patients with schizophrenia, and a number of reports
suggest that access to medical care may be limited for patients with
schizophrenia. Multiple factors could contribute to decreased access, including limitations in the communication of symptoms by patients,
poor cooperation by psychotic patients, stigma toward patients with
schizophrenia, and insufficient attention to medical problems by mental
health providers (Druss and Rosenheck 1997; Goldman 1999). Dixon et
al. (1999) found that less than 70% of the patients with medical problems
were receiving treatment for their medical conditions in a study of 719
patients with schizophrenia from the Patient Outcomes Research Team
project. Masterson and OShea (1984) speculated that inadequate breast
examinations and incomplete reporting of breast lumps could have contributed to the increased rate of death from breast cancer in their study.
The diagnosis of schizophrenia was significantly negatively associated
with timeliness, access, and intensity of postdischarge medical care in a
study of U.S. veterans (Druss and Rosenheck 1997). Finally, two related
studies have demonstrated disparity in care for patients with schizophrenia following myocardial infarction. Patients with schizophrenia
were less likely to have cardiac catheterization, more likely to die in the
year after discharge, and less likely to receive interventions that represent quality of care following acute myocardial infarction (i.e., reperfusion therapy, or at discharge, aspirin, beta-blockers, and angiotensinconverting enzyme inhibitors) (Druss et al. 2000, 2001).
Among the homeless, having a diagnosis of schizophrenia may be
associated with better care and reduced mortality compared with the
general homeless population. The age-adjusted mortality rate for a
New York City homeless sample (follow-up period 19871994) was
about four times that in a comparable reference group (Barrow et al.
1999). Having a mental illness had a significant protective effect, possibly explained by provision of relatively greater services and better
housing for homeless people who were mentally ill than for other
homeless people. Likewise, having schizophrenia appeared to convey
protection from mortality in a Boston homeless population (Hwang et
al. 1998). The benefit of additional care most likely creates only an impression of protection, however, considering that the comparison
groups are highly ill and impoverished. For example, mortality was
studied in homeless individuals who had been referred to a clinic in
Sydney, Australia, between 1988 and 1991 (Babidge et al. 2001). The

Medical Illness and Schizophrenia


SMR for the nonschizophrenia group was 4.41 (95% CI 2.026.19) compared with 2.52 (95% CI 1.83.43) for the schizophrenia group. Thus, although the SMR was much lower in the schizophrenia group, it still
represented a mortality rate more than twice that expected in the general population.

The literature conclusively establishes that patients with schizophrenia
die at higher rates and earlier than other people. The causes of this excess mortality have varied over time, reflecting issues related to
changes in the care of patients with schizophrenia and the unhealthy
lifestyles that result from a combination of their illness and the type of
care available. At this time, interventions must address risks from suicide and from medical illnesses. Effective medications with side-effect
profiles that promote adherence to treatment plans must be provided.
Avoiding adverse metabolic side effects, or at least monitoring and attempting to correct them as they occur, is also critical in reversing the
trend toward higher mortality in patients with schizophrenia. Finally,
the mental health system must work to ensure that adequate social services, medical treatment, and psychiatric care are provided despite societys drive to reduce health care costs.

Key Clinical Points

Schizophrenia carries increased mortality, as demonstrated by increased mortality rates and shortened life span.

Differences in mortality rates between patients with schizophrenia and

the general population are increasing.

Suicide rates are very high for patients with schizophrenia, especially
early in the course of illness.

Increased rates of death from natural causes account for many of the
excess deaths seen in patients with schizophrenia, although the increases
in mortality rates associated with natural causes are not as great as
those associated with suicide.

The natural causes of death for patients with schizophrenia are similar
to those that occur in the general population, with cardiovascular disease and diabetes being notably problematic.

Excessive Mortality and Morbidity


Allebeck P: Schizophrenia: a life-shortening disease. Schizophr Bull 15:8189, 1989
American Diabetes Association, American Psychiatric Association, American
Association of Clinical Endocrinologists, et al: Consensus Development
Conference on Antipsychotic Drugs and Obesity in Diabetes. Diabetes
Care 27:596601, 2004
Babidge NC, Buhrich N, Butler T: Mortality among homeless people with
schizophrenia in Sydney, Australia: a 10-year follow-up. Acta Psychiatr
Scand 103:105110, 2001
Barrow SM, Herman DB, Cordova P, et al: Mortality among homeless shelter
residents in New York City. Am J Public Health 89:529534, 1999
Black DW, Warrack G, Winokur G: The Iowa record-linkage study, I: suicides
and accidental deaths among psychiatric patients. Arch Gen Psychiatry
42:7175, 1985
Brown S: Excess mortality of schizophrenia: a meta-analysis. Br J Psychiatry
171:502508, 1997
Brown S, Birtwistle J, Roe L, et al: The unhealthy lifestyle of people with schizophrenia. Psychol Med 29:697701, 1999
Brown S, Inskip H, Barraclough B: Causes of the excess mortality of schizophrenia. Br J Psychiatry 177:212217, 2000
Capasso RM, Lineberry TW, Bostwick JM, et al: Mortality in schizophrenia and
schizoaffective disorder: an Olmsted County, Minnesota cohort: 1950
2005. Schizophr Res 98:287294, 2008
Casey DE: Neuroleptic-induced acute extrapyramidal syndromes and tardive
dyskinesia. Psychiatr Clin North Am 16:589610, 1993
Casey DE, Hansen TE: Excessive morbidity and mortality in schizophrenia, in
Medical Illness and Schizophrenia. Edited by Meyer JM, Nasrallah HA.
Washington, DC, American Psychiatric Publishing, 2003, pp 1334
Cohen ME, Dembling B, Schorling JB: The association between schizophrenia and
cancer: a population-based mortality study. Schizophr Res 57:139146, 2002
Craig TJ, Ye Q, Bromet EJ: Mortality among first-admission patients with psychosis. Compr Psychiatry 47:246251, 2006
Dalmau A, Bergman B, Brismar B: Somatic morbidity in schizophrenia: a case
control study. Public Health 111:393397, 1997
Dalton SO, Laursen TM, Mellemkjaer L, et al: Schizophrenia and the risk for
breast cancer. Schizophr Res 62:8992, 2003
Dalton SO, Mellemkjaer L, Thomassen L, et al: Risk for cancer in a cohort of patients hospitalized for schizophrenia in Denmark, 19691993. Schizophr
Res 75:315324, 2005
Dixon L: Dual diagnosis of substance abuse in schizophrenia: prevalence and
impact on outcomes. Schizophr Res 35(suppl):S93S100, 1999
Dixon L, Postrado L, Delahanty J, et al: The association of medical comorbidity
in schizophrenia with poor physical and mental health. J Nerv Ment Dis
187:496502, 1999


Medical Illness and Schizophrenia

Druss BG, Rosenheck RA: Use of medical services by veterans with mental disorders. Psychosomatics 38:451458, 1997
Druss BG, Bradford DW, Rosenheck RA, et al: Mental disorders and use of cardiovascular procedures after myocardial infarction. JAMA 283:506511,
Druss BG, Bradford WD, Rosenheck RA, et al: Quality of medical care and excess mortality in older patients with mental disorders. Arch Gen Psychiatry
58:565572, 2001
Fontaine KR, Heo M, Harrigan EP, et al: Estimating the consequences of antipsychotic induced weight gain on health and mortality rate. Psychiatry Res
101:277288, 2001
Fors BM, Isacson D, Bingefors K, et al: Mortality among persons with schizophrenia in Sweden: an epidemiological study. Nord J Psychiatry 61:252
259, 2007
Goldman LS: Medical illness in patients with schizophrenia. J Clin Psychiatry
60 (suppl 21):1015, 1999
Gulbinat W, Dupont A, Jablensky A, et al: Cancer incidence of schizophrenic
patients: results of record linkage studies in three countries. Br J Psychiatry
Suppl 7583, 1992
Hansen TE, Casey DE, Hoffman WF: Neuroleptic intolerance. Schizophr Bull
23:567582, 1997
Harris EC, Barraclough B: Excess mortality of mental disorder. Br J Psychiatry
173:1153, 1998
Heila H, Isometsa ET, Henriksson MM, et al: Suicide and schizophrenia: a nationwide psychological autopsy study on age- and sex-specific clinical
characteristics of 92 suicide victims with schizophrenia. Am J Psychiatry
154:12351242, 1997
Heila H, Isometsa ET, Henriksson MM, et al: Suicide victims with schizophrenia in different treatment phases and adequacy of antipsychotic medication. J Clin Psychiatry 60:200208, 1999
Hiroeh U, Appleby L, Mortensen PB, et al: Death by homicide, suicide, and
other unnatural causes in people with mental illness: a population-based
study. Lancet 358:21102112, 2001
Hiroeh U, Kapur N, Webb R, et al: Deaths from natural causes in people with
mental illness: a cohort study. J Psychosom Res 64:275283, 2008
Holmberg SK, Kane C: Health and self-care practices of persons with schizophrenia. Psychiatr Serv 50:827829, 1999
Hussar AE: Leading causes of death in institutionalized chronic schizophrenic
patients: a study of 1,275 autopsy protocols. J Nerv Ment Dis 142:4557,
Hwang SW, Lebow JM, Bierer MF, et al: Risk factors for death in homeless
adults in Boston. Arch Intern Med 158:14541460, 1998
Inskip HM, Harris EC, Barraclough B: Lifetime risk of suicide for affective disorder, alcoholism and schizophrenia. Br J Psychiatry 172:3537, 1998

Excessive Mortality and Morbidity


Jackson CT, Covell NH, Drake RE, et al: Relationship between diabetes and
mortality among persons with co-occurring psychotic and substance use
disorders. Psychiatr Serv 58:270272, 2007
Jeste DV, Gladsjo JA, Lindamer LA, et al: Medical comorbidity in schizophrenia. Schizophr Bull 22:413430, 1996
Joukamaa M, Helivaara M, Knekt P, et al: Schizophrenia, neuroleptic medication and mortality. Br J Psychiatry 188:122127, 2006
Laursen TM, Munk-Olsen, T, Nordentoft M, et al: Increased mortality among
patients admitted with major psychiatric disorders: a register-based study
comparing mortality in unipolar depressive disorder, bipolar affective disorder, schizoaffective disorder, and schizophrenia. J Clin Psychiatry
68:899907, 2007
Lichtermann D, Ekelund J, Pukkala E, et al: Incidence of cancer among persons
with schizophrenia and their relatives. Arch Gen Psychiatry 58:573578,
Limosin F, Loze J-Y, Philippe A: Ten-year prospective follow-up study of the
mortality by suicide in schizophrenic patients. Schizophr Res 94:2328,
Masterson E, OShea B: Smoking and malignancy in schizophrenia. Br J Psychiatry 145:429432, 1984
Mortensen PB, Juel K: Mortality and causes of death in schizophrenic patients
in Denmark. Acta Psychiatr Scand 81:372377, 1990
Newman SC, Bland RC: Mortality in a cohort of patients with schizophrenia: a
record linkage study. Can J Psychiatry 36:239245, 1991
Nordentoft M, Laursen TM, Agerbo E, et al: Change in suicide rates for patients
with schizophrenia in Denmark, 198197: nested case-control study. BMJ
329:261, 2004
Palmer BA, Pankratz VS, Bostwick JM: The lifetime risk of suicide in schizophrenia: a reexamination. Arch Gen Psychiatry 62:247253, 2005
Ran MS, Chen E, Conwell Y, et al: Mortality in people with schizophrenia in rural China: 10-year cohort study. Br J Psychiatry 190:237242, 2007
Saha S, Chant D, McGrath J: A systematic review of mortality in schizophrenia:
is the differential mortality gap worsening over time? Arch Gen Psychiatry
64:11231131, 2007
Waddington JL, Youssef HA, Kinsella A: Mortality in schizophrenia: antipsychotic polypharmacy and absence of adjunctive anticholinergics over the
course of a 10-year prospective study. Br J Psychiatry 173:325329, 1998

This page intentionally left blank

Medical Outcomes
From the CATIE
Schizophrenia Study
Henry A. Nasrallah, M.D.

When the Clinical Antipsychotic Trials for Intervention Effectiveness (CATIE) schizophrenia study was designed and submitted to the
National Institute of Mental Health (NIMH) for funding in 1999, the
main purpose of the study was to establish whether second-generation
antipsychotics (SGAs) were more effective than first-generation antipsychotics (FGAs). Specifically, effectiveness was defined as discontinuation due to any of four reasons: lack of efficacy, lack of tolerability,
emergence of safety problems (i.e., potential medical health threat), and
patient decision (i.e., nonadherence and dropping out of the study).
Measures related to physical health were considered a secondary outcome of this large study.
By the time the initial findings of the CATIE trial were published in
2005 (Lieberman et al. 2005), the predominant focus in the field of
schizophrenia pharmacotherapy had shifted from claims over relative
efficacy of various SGAs to the medical complications of the SGA class,
specifically metabolic disorders such as obesity, diabetes, hyperlipidemia, and hypertension (Nasrallah et al. 2006; Newcomer 2005). Fortunately, the CATIE study included several laboratory measures of
metabolic function, including all five metabolic syndrome criteria:


Medical Illness and Schizophrenia

waist circumference, fasting triglycerides, fasting high-density lipoprotein, blood pressure, and fasting serum glucose (Grundy et al. 2004). A
total of 1,460 persons with schizophrenia consented to participate in the
CATIE study, and several analyses of metabolic parameters at CATIE
baseline and again after administration of the SGA and FGA medications used in phases 1 and 2 have been published and are discussed in
this review. Interestingly, although the effectiveness results of CATIE
(i.e., the differential discontinuation of the four SGA agents and one
FGA agent during phase 1) were subjected to intense critiques due to
methodological issues, the metabolic data were widely accepted and
were regarded as objectively valid and uninfluenced by design stipulations (Kasper and Winkler 2006; Meyer 2007).
In addition to analyzing the CATIE metabolic data (with pre hoc
hypotheses based on the totality of the literature and evidence that
emerged while CATIE was being conducted 19992004), the researchers involved in the CATIE study also accumulated data on other healthrelated issues such as substance abuse and neurological movement
disorders, both of which are discussed in this overview of the medical
findings of the CATIE study.

The Buildup to the CATIE

Awareness of the adverse metabolic effects from certain SGAs accelerated during the 5 years that it took to conduct the CATIE study (Nasrallah 2003; Newcomer et al. 2004). A key review by Allison et al. (1999)
raised awareness of the seriousness of weight gain secondary to antipsychotic treatment, whether from FGAs or SGAs. Allison et al.s literature review of comparative weight gain across old and new
antipsychotics was a turning point in that it reframed the discussion
about weight gain and its serious metabolic consequences as an important threat to the health of patients receiving antipsychotics. Awareness
of the serious (and occasionally fatal) metabolic effects posed by some
SGAs essentially replaced concerns over tardive dyskinesia, which had
reigned supreme as the greatest health issue related to FGA exposure.
Interestingly, Allison et al.s review also implied that weight gain was
a major problem with certain low-potency FGA agents (thioridazine
and chlorpromazine), a fact that was often overlooked due to the preoccupation with tardive dyskinesia and the increasing use of highpotency FGAs in the years prior to widespread SGA use. Allison et al.s
analysis influenced the CATIE study design to the extent that on the basis of Allison et al.s data, a weight-neutral SGA, ziprasidone, was
planned as a treatment arm more than 2 years before final U.S. Food

Medical Outcomes From the CATIE Study


and Drug Administration (FDA) approval in 2001. Because of its benign

metabolic profile, ziprasidone was included not only as one of the SGA
options in phase 1 but also as a key element of the tolerability pathway
in phase 2 of the CATIE study (Stroup et al. 2006). Aripiprazole, another
metabolically benign SGA, came too late to be included in phase 1 or 2
but was included in phase 3, the open-label phase of CATIE in which
investigators could choose any approved antipsychotic for subjects
who had dropped out of phase 2 (Stroup et al., in press).
Although weight gain secondary to clozapine was widely recognized in the 1990s, prior to the CATIE study (Umbricht et al. 1994), it
was regarded as the unavoidable cost for treating refractory schizophrenia patients for whom there were no options; however, after several first-line SGA drugs were introduced (risperidone in 1993,
olanzapine in 1996, quetiapine in 1997), the incidence of weight gain
and diabetes increased markedly due to the drugs widespread use. By
the turn of the millennium, while patients were being actively recruited
for the CATIE study, metabolic disorders related to SGA use moved
into the spotlight for both clinicians and researchers. From 2002
through 2004, the FDA MedWatch data from cases of diabetes and diabetic ketoacidosis (including deaths) were published in a series of papers (Koller and Doraiswamy 2002; Koller et al. 2003, 2004). The
European Union and Japan subsequently added a diabetes warning to
olanzapines product label in 2002, and in August 2003, the FDA imposed a class warning for diabetes on all SGA drugs, despite the absence of data implicating ziprasidone and aripiprazole.
Three months after the FDA class warning, the American Diabetes
Association, along with the American Psychiatric Association, American Association of Clinical Endocrinologists, and North American Association for the Study of Obesity, held a meeting in Philadelphia on
November 19, 2003. The purpose of this joint conference was to generate a consensus statement (American Diabetes Association et al. 2004)
on the association between SGA exposure and weight gain, lipid
changes, and diabetes (predominantly level 2, 3, and 4 evidence) after
reviewing the pre-CATIE world literature. The consensus statement,
which was published jointly in the February 2004 issues of Diabetes Care
and the Journal of Clinical Psychiatry, concluded that the members of the
SGA class were associated with differential risks of weight gain, diabetes, and hyperlipidemia. Clozapine and olanzapine were found to have
the highest risk, risperidone and quetiapine to have intermediate risk,
and aripiprazole and ziprasidone to have the lowest risk. As will be discussed in the next section, the prospective, randomized, double-blind
CATIE trial generated level 1 evidence about metabolic complications


Medical Illness and Schizophrenia

of SGAs that professionals were waiting for, and confirmed the conclusions of the consensus statement regarding obesity and diabetes. Moreover, the CATIE data refined the knowledge of the relative metabolic
effects of quetiapine and indicated that risperidone might be somewhat
less of a metabolic offender than quetiapine, particularly with regard to
effects on serum triglyceride levels (Meyer et al. 2008a, 2008b).

CATIE Baseline Data on

Medical Illness in Schizophrenia
Several health-related findings were reported in the CATIE schizophrenia sample, which comprised 1,460 outpatients from 57 public, academic, and private settings around the United States. The sample was
75% male, with a mean age of 41 years and a mean duration of illness of
over 14 years. About 60% of the sample was white, 35% black, and 5%
other races. Twelve percent of the sample was of Hispanic or Latino ethnicity. The mean educational level was 12.2 years, about 60% had never
married, and 85% were unemployed. CATIE employed a broad enrollment strategy in an attempt to replicate the patient population seen in
real-world clinical settings, as opposed to the type of patient often recruited for pharmaceutical trials, which excluded patients with medical
comorbidities or substance use. As a result, the data provide a representative, cross-sectional view of the health of chronic schizophrenia patients in the United States.

Diabetes and Metabolic Syndrome

At CATIE study baseline, 11% of enrolled subjects had diabetes type 1
or 2, 14% had hyperlipidemia, and 20% had hypertension. Table 31
shows the mean number of metabolic syndrome criteria met at CATIE
study baseline in subjects with fasting blood samples available, along
with overall metabolic syndrome prevalence (McEvoy et al. 2005).
Metabolic syndrome prevalence in the CATIE sample, in both males
and females, was twice the rate seen in a sample from the general population that was matched for age, gender, race, and ethnicity (using the
National Health and Nutrition Examination Survey [NHANES] III
study population as the source for comparator matching) (Figure 31).
These were the first detailed data documenting an unusually high prevalence of metabolic syndrome in the schizophrenia outpatient population in the United States during the period when subjects were being
recruited into the CATIE study.

Medical Outcomes From the CATIE Study


Cardiovascular Risk
A comparison of 10-year cardiac risk estimates in the CATIE schizophrenia sample compared to matched controls from the NHANES III
was also conducted using the baseline CATIE data (Goff et al. 2005).
This analysis showed that the 10-year coronary heart disease (CHD)
risk (using the Framingham CHD risk algorithm) was significantly elevated in the CATIE sample versus the matched general population
(P= 0.0001) in males (9.4% vs. 7.0%) and females (6.3% vs. 4.2%). The patients with schizophrenia also had significantly higher rates of smoking
(68% vs. 35%), diabetes (13% vs. 3%), and hypertension (27% vs. 17%),
and lower HDL cholesterol (43.7 vs. 49.3 mg/dL) compared to the
NHANES III control group. Goff et al. concluded that the CATIE data
showing greater 10-year CHD risk are consistent with the published
studies of increased cardiac mortality in schizophrenia.

Lack of Medical Treatment

Percentage of patients with metabolic syndrome

In light of the high prevalence of metabolic syndrome in the CATIE

schizophrenia sample, one would expect that the subjects with metabolic disorders such as diabetes, hyperlipidemia, and hypertension
would be receiving the appropriate standard medical treatments for


CATIE (N = 689)

NHANES (N = 689)




FIGURE 31. Prevalence of metabolic syndrome in CATIE schizophrenia

study participants at baseline versus the general adult population (National
Health and Nutrition Examination Survey [NHANES] III data).
*P= 0.0001 CATIE versus NHANES.



Metabolic syndrome and criteria prevalences among CATIE fasting subjects

Subject cohort

Mean age (years)

Criteria met

Mean number of
criteria met

MS prevalence,
AHA criteria
Criteria met


Black male

Hispanic male

White female

Black female

Hispanic female



38.0 12.3

44.2 10.7

44.0 9.8




29.0 6.6

32.5 8.2

33.8 8.1


40.9% (n=342)

22.7% (n=141)

31.8% (n=66)

56.5% (n=92)

43.1% (n=72)

73.3% (n=15)

(n =343)
2.11 1.32

(n =141)
1.59 1.35

(n =67)
2.05 1.31

(n =93)

(n= 73)
2.26 1.19

(n= 15)
2.87 1.25

44.4% (n=342)

23.4% (n=141)

40.9% (n=66)

58.1% (n=93)

47.2% (n=72)

73.3% (n=15)

(n =343)

(n =141)

(n =67)

(n =93)

(n= 73)

(n= 15)

Medical Illness and Schizophrenia

Mean BMI (kg/m )

MS prevalence,
ATPIII criteria

White male


Metabolic syndrome and criteria prevalences among CATIE fasting subjects (continued)
Subject cohort


Mean number of
criteria met

Black male

Hispanic male

White female

Black female

Hispanic female

2.25 1.38

1.62 1.35

2.21 1.38


2.37 1.23

2.93 1.28

Criteria prevalence: all subjects

Waist circumference 41.2% (n= 663)
Blood pressure
46.9% (n= 674)
54.3% (n= 671)

28.5% (n= 351)

55.3% (n= 358)
35.6% (n= 357)

37.8% (n =127)
38.8% (n =129)
55.0% (n =129)

70.8% (n =195)
38.4% (n =198)
69.1% (n =194)

76.5% (n =153)
48.1% (n =154)
52.6% (n =154)

77.0% (n= 40)

47.5% (n= 40)
64.1% (n= 39)

Criteria prevalence: fasting subjects

Waist circumference 39.0% (n= 341)
Blood pressure
45.9% (n= 342)
58.3% (n= 343)
55.1% (n= 343)
Glucose 100
26.5% (n= 343)
Glucose 110
13.4% (n =343)

26.2% (n= 141)

57.1% (n= 141)
31.9% (n= 141)
34.8% (n= 141)
17.7% (n= 141)
14.9% (n= 141)

38.5% (n =65)
40.9% (n =66)
61.2% (n =67)
53.7% (n =66)
28.4% (n =67)
11.9% (n =67)

76.1% (n =92)
41.9% (n =93)
53.8% (n =93)
72.0% (n =93)
22.6% (n =93)
16.1% (n =93)

74.7% (n= 71)

50.0% (n= 72)
28.8% (n= 73)
52.1% (n= 73)
34.3% (n= 73)
23.3% (n= 73)

86.7% (n= 15)

60.0% (n= 15)
46.7% (n= 15)
66.7% (n= 15)
33.3% (n= 15)
26.7% (n= 15)


Note. AHA=American Heart Association; ATPIII= National Cholesterol Education Program Adult Treatment Panel III;
BMI= body mass index; HDL=high-density lipoprotein; MS =metabolic syndrome.

Medical Outcomes From the CATIE Study


White male


Medical Illness and Schizophrenia

these serious medical disorders. However, Nasrallah et al. (2006) found

that a high proportion of subjects were not receiving treatment for their
serious medical diseases at the time of their enrollment into the CATIE
trial. This serious lack of care means that persons with schizophrenia
face a double jeopardy for early cardiovascular mortality: high rate of
metabolic disorders and diminished quality of medical treatment. This
combination of high medical risk and poor medical care may be one of
the reasons that persons with schizophrenia lose an average of more
than 25 potential years of life, according to research from several U.S.
states (Colton and Manderscheid 2006; B.J. Miller et al. 2006). Metabolic
disorders are also highly associated with depression; when CATIE subjects with metabolic syndrome were compared to CATIE subjects without metabolic syndrome, the only significant difference found was a
higher rate of depression and somatic concern among the group with
metabolic syndrome (Meyer et al. 2005).

Other Medical Risk Factors

Of the CATIE subjects, 37% met criteria for substance abuse or dependence, 23% met criteria specifically for substance use, and 40% were abstinent (Swartz et al. 2008). Given the health risks of oral and
intravenous substance use, abuse, and dependence, this is an additional
medical risk factor for individuals with schizophrenia, a fact discussed
at length in Chapter 11, Substance Abuse and Schizophrenia. Another medical risk factor in the CATIE sample was tardive dyskinesia,
a serious form of neurotoxicity. Of the 1,460 subjects, 231 (15.8%) had
tardive dyskinesia at the time of enrollment in CATIE (D.D. Miller et al.
2005). Of interest are the findings that the tardive dyskinesia subgroup
had 1.5 times higher prevalence of substance abuse and 4.5 times higher
odds of more severe psychopathology compared to those without tardive dyskinesia.

CATIE Phase 1 Medical Findings

Several metabolic parameters were differentially influenced by the five
antipsychotic agentsperphenazine (an FGA), risperidone, quetiapine,
olanzapine, and ziprasidone (added after 40% of the sample had been
recruited)selected for phase 1 of the CATIE trial (Lieberman et al.
2005). Figure 32 shows the mean monthly weight change, with the
olanzapine group showing over fourfold greater weight gain compared
with the other antipsychotics during phase 1, and subjects randomized

Medical Outcomes From the CATIE Study


to ziprasidone showing a mean decrease in weight. Figure 33 shows

the percentage of subjects who gained a significant amount of weight
(> 7%) with each of the five antipsychotics. Again, the olanzapine cohort
had the greatest problems with weight gain, with 30% gaining more
than 7% of their baseline weight, and ziprasidone was the least offending medication, causing weight gain in only 6% of subjects. No statistically significant between-group changes in mean fasting glucose levels
occurred across the various antipsychotics, although olanzapine had
the highest numeric increase and ziprasidone the lowest. However, results indicated significant between-drug differences for change in glycosylated hemoglobin, a long-term measure of glycemic control that
reflects values over the prior 2 months. As seen in Figure 34, the greatest increase in glycosylated hemoglobin was in the olanzapine group,
and the ziprasidone group had an overall decrease. Similar divergences
were seen with cholesterol and triglycerides (Figures 35 and 36) and
are discussed in detail in Chapter 6, Effects of Antipsychotics on Serum Lipids (see Tables 62 and 63).
Each patients corrected QT interval on the electrocardiogram (QTc)
was measured every 3 months while participating in the CATIE study.
Figure 37 shows that the QTc was minimally increased in all antipsychotic groups during CATIE phase 1. These negligible increases suggest
that, contrary to the perception at the launch of the CATIE study, antipsychotic-related QTc changes are not a major medical threat to persons receiving atypical antipsychotics. The effects of antipsychotics on
QTc are discussed at length in Chapter 7, The Spectrum of Cardiovascular Disease in Patients With Schizophrenia, but data from CATIE
phase 1 revealed that in routine clinical practice, there are not worrisome QT changes with any SGA, including ziprasidone, which had
been a source of concern. The past several years of clinical experience
with millions of patients confirms the CATIE data in that unlike metabolic complications (e.g., metabolic syndrome and diabetes), which
have increased markedly in populations receiving SGAs, morbidity or
mortality from QTc changes is relatively minuscule.
The development of cataracts was a concern only for patients taking
quetiapine, based on preclinical findings from animal models. In phase
1, no difference was found among the antipsychotics, with quetiapine
in particular showing minimal risk for cataracts, confirming that cataract development is a phenomenon encountered for the preclinical species studied (beagle dogs) but not humans.










Medical Illness and Schizophrenia

Mean change from baseline (lbs)






Phase 1

Phase 2T

Phase 2E

FIGURE 32. Monthly weight gain (pounds/month) by subjects during CATIE phase 1, phase 2 tolerability arm (2T), and phase
2 efficacy arm (2E).









Medical Outcomes From the CATIE Study






Phase 1

Phase 2 E

Percentage of subjects with weight gain >7% in CATIE phase 1, phase 2 tolerability arm (2T), and phase 2 efficacy

arm (2E).

Phase 2 T

Medical Illness and Schizophrenia








Mean change from baseline (%)








Phase 1 changes in glycosylated hemoglobin.

Change in Metabolic Syndrome Status

With Treatment
Detailed analysis of changes in metabolic syndrome criteria in CATIE
phase 1 were published in 2008 (Meyer et al. 2008a) and showed that after
3 months of exposure, the prevalence of metabolic syndrome increased in
the olanzapine group from 34.8% to 43.9%, but decreased in the ziprasidone group from 37.7% to 29.9% (olanzapine vs. ziprasidone P=0.001)
(Table 32). Subjects in the olanzapine and quetiapine groups also had an
average waist circumference increase of 1.8 cm, compared with a 1-cm
increase in the risperidone group and no change in the ziprasidone
group, whereas subjects in the perphenazine group experienced a loss of
1 cm. The greatest increase in triglycerides occurred with olanzapine
(+21.5 mg/dL). Although these findings are consistent with metabolic
studies published before release of the CATIE results, they indicated
greater effects of quetiapine on adiposity than previously indicated by
the American Diabetes Association et al. (2004) consensus paper.

Nonfasting Triglyceride Levels

Although the metabolic syndrome criterion for triglycerides is a fasting
level of 150 mg/dL or greater, recent studies indicate that nonfasting

Medical Outcomes From the CATIE Study


TABLE 32. Proportion of subjects at baseline and 3 months meeting

the criteria for metabolic syndrome in phase 1 of the
CATIE schizophrenia trialall classifiable subjects
Metabolic syndrome

Overall treatment difference


3 months



Change from baseline to 3 months in proportion of subjects meeting

criteria for metabolic syndrome is greater for olanzapine than for
ziprasidone (P=0.001) among all classifiable subjects.
triglyceride levels may have a stronger association with cardiovascular
risk than do fasting triglycerides. An analysis of the phase 1 CATIE data
after 3 months of exposure (Meyer et al. 2008b) showed that the highest
mean and median increases were in subjects randomly assigned to quetiapine (mean +54.7 mg/dL, median +26 mg/dL) and olanzapine (mean
+54.7 mg/dL, median +26.5 mg/dL), whereas ziprasidone was neutral
(mean +0.0 mg/dL, median +8 mg/dL) and decreases were seen in both
risperidone (mean 18.4 mg/dL, median 6.5 mg/dL) and perphenazine (mean 1.3 mg/dL, median 22 mg/dL). Pairwise comparison indicated a significant between-group difference for perphenazine versus
olanzapine and a trend for perphenazine versus quetiapine.

Inflammatory Markers
Several biomarkers of cardiometabolic risk were collected in phase 1
of the CATIE trial and are still being analyzed at the time of writing
of this chapter. However, data for one important inflammatory marker,
C-reactive protein (CRP) have been analyzed (J.M. Meyer et al., in preparation) to examine the comparative effects of various antipsychotics after 3 months of exposure. Olanzapine and quetiapine had the greatest
numerical increases, and significant treatment differences were observed between olanzapine versus perphenazine (P< 0.001) and risperidone (P = 0.001) in subjects with lower baseline cardiovascular risk
(defined as CRP <1 mg/L). After 12 months of exposure, the difference









Medical Illness and Schizophrenia

Change from baseline (mg/dL)





Ph ase 1


Ph ase 2 T

Ph ase 2 E

Changes in serum triglycerides across CATIE phase 1, phase 2 tolerability arm (2T), and phase 2 efficacy arm (2E).








Medical Outcomes From the CATIE Study

Change from baseline (mg/dL)




Ph ase 1


Ph ase 2T

Ph ase 2E

Changes in total cholesterol across CATIE phase 1, phase 2 tolerability arm (2T), and phase 2 efficacy arm (2E).







Medical Illness and Schizophrenia

Mean change from baseline (msec)

Phase 1


Phase 2T

Mean changes in QT interval in CATIE phase 1 and phase 2 tolerability arm (2T).

Medical Outcomes From the CATIE Study


between olanzapine and perphenazine groups remained significant,

and a significant difference (P= 0.003) was observed between olanzapine
and ziprasidone. These data confirm that agents associated with greater
cardiometabolic risk are associated with adverse changes in other measures that correlate with deleterious cardiovascular outcomes.

Phase 2 Findings
Subjects who discontinued their antipsychotics in phase 1 of the CATIE
study due to lack of efficacy were randomized either to clozapine or to
one of three atypical antipsychotics: olanzapine, quetiapine, or risperidone. Those who discontinued due to tolerability or safety reasons (or
who refused participation in the efficacy arm) were randomized to either
ziprasidone or to one of the three atypical antipsychotics. In both arms of
phase 2, the tolerability arm and the efficacy arm, all subjects were randomly assigned to receive agents that they did not take in phase 1. Metabolic parameters (weight, waist circumference, glucose, triglycerides,
and cholesterol) were measured as in phase 1. (See Figures 35 and 36.)
Figure 32 shows the mean weight gain per month in both pathways
of phase 2 compared to phase 1, and Figure 33 shows the percentage of
subjects with weight gain >7% for both phase 2 pathways. Figure 38
also shows the weight changes in phase 2 in subjects who had a significant amount of weight gain (> 7%) in phase 1. Only those randomized to
ziprasidone lost much of the weight they had gained from the phase 1
antipsychotic. Differences did emerge between the phase 1 results and
those seen in the phase 2 efficacy and tolerability pathways, but a consistent pattern remains: olanzapine is associated with the greatest deleterious changes in metabolic parameters, whereas ziprasidone is
associated with the most benign changes.

Phase 3 Findings
The open-label phase 3 included subjects who had discontinued phases 1
and 2. They were given a wide range of FGAs and/or SGAs. Aripiprazole, the only atypical antipsychotic that was not included in the first two
phases of the CATIE because the drug was launched too late, was used in
some patients in phase 3. The only surprising data regarding the metabolic changes in phase 3 were that glucose levels increased by 13 mg/dL
in patients taking aripiprazole, which was a larger increase than with any
other antipsychotic in this phase (Stroup et al., in press). This finding was
unexpected, because aripiprazole has demonstrated a benign metabolic
profile in FDA trials and in previous published studies. All other metabolic findings were consistent with phase 1 and phase 2 profiles.








Phase 2 tolerability arm weight change in CATIE subjects with weight gain >7% in phase 1.

Medical Illness and Schizophrenia

Mean change from baseline (lbs)

Medical Outcomes From the CATIE Study


The 5-year, double-blind, NIMH-sponsored effectiveness study known
as the CATIE schizophrenia trial provided highly informative data
about the comparative metabolic profiles of various antipsychotics and
revealed the high risk for metabolic syndrome and CHD in a large sample (N= 1,460) of patients with chronic schizophrenia in the United
States. These metabolic findings are likely to influence clinical practice
research studies, prevention initiatives, and public policy in the foreseeable future. Although the primary measure outcome of the CATIE
schizophrenia trial was effectiveness (i.e., all-cause discontinuation), a
valuable body of metabolic data was obtained from the study sample,
one that points to significant health problems seen in patients with

Key Clinical Points

The prevalence of metabolic syndrome in the CATIE schizophrenia

subjects at the time of enrollment was 43%, about twice the prevalence
in the general population.

Olanzapine was associated with a fourfold higher monthly weight gain

in phase 1 compared to the other antipsychotics.

Fasting blood glucose, cholesterol, and triglyceride levels were most elevated with olanzapine and were not changed or declined with ziprasidone.

Quetiapine resulted in higher lipid increases than risperidone, which

had no deleterious effect on lipids.

The overall metabolic profile of second-generation antipsychotics in

phase 2 (both the efficacy and tolerability pathways) were the same as
seen in phase 1 for olanzapine (highest increases) and ziprasidone (lowest increases or actual decreases), but were somewhat variable for quetiapine and risperidone.

Subjects with >7% body weight increase in phase 1 lost weight only if
they were assigned to ziprasidone in phase 2.

QTc was minimally affected by all second-generation antipsychotics

used in the CATIE study.

Levels of the inflammatory marker C-reactive protein were increased

with olanzapine or quetiapine but not with other antipsychotics.

Medical Illness and Schizophrenia


Aripiprazole was associated with the highest mean increase in fasting

glucose among first-generation or second-generation antipsychotics in
phase 3, the opposite of what was expected.

Allison DB, Mentore JL, Heo M, et al: Antipsychotic-induced weight gain: a
comprehensive research synthesis. Am J Psychiatry 156:16861696, 1999
American Diabetes Association, American Psychiatric Association, American
Association of Clinical Endocrinologists, et al: Consensus Development
Conference on Antipsychotic Drugs and Obesity and Diabetes. Diabetes
Care 27:596601, 2004
Colton CW, Manderscheid RW: Congruencies in increased mortality rates,
years of potential life lost, and causes of death among public mental health
clients in eight states. Prev Chronic Dis 3:114, 2006
Goff DC, Sullivan L, McEvoy JP, et al: A comparison of ten-year cardiac risk estimates in schizophrenia patients from the CATIE study and matched controls. Schizophr Res 80:4553, 2005
Grundy SM, Brewer B, Cleeman JI, et al: Definition of metabolic syndrome: report of the National Heart, Lung, and Blood Institute/American Heart Association conference on scientific issues related to definition. Circulation
109:433438, 2004
Kasper S, Winkler D: Addressing the limitations of the CATIE study. World J
Biol Psychiatry 7:126127, 2006
Koller EA, Doraiswamy PM: Olanzapine-associated diabetes mellitus. Pharmacotherapy 22:841852, 2002
Koller EA, Cross JT, Doraiswamy PM, et al: Risperidone-associated diabetes
mellitus: a pharmacovigilance study. Pharmacotherapy 23:735744, 2003
Koller EA, Weber J, Doraiswamy PM, et al: A survey of reports of quetiapineassociated hyperglycemia and diabetes mellitus. J Clin Psychiatry 65:857
863, 2004
Lieberman JA, Stroup TS, McEvoy JP, et al: Effectiveness of antipsychotic drugs
in patients with chronic schizophrenia. N Engl J Med 353:12091223, 2005
McEvoy JP, Meyer JM, Goff DC, et al: Prevalence of the metabolic syndrome in
patients with schizophrenia: baseline results from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) schizophrenia trial and
comparison with national estimates from NHANES III. Schizophr Res
80:1932, 2005
Meyer JM: Strategies for the long-term treatment of schizophrenia: real-world
lessons from the CATIE trial. J Clin Psychiatry 68 (suppl 1):2833, 2007

Medical Outcomes From the CATIE Study


Meyer JM, Nasrallah HA, McEvoy JP, et al: The Clinical Antipsychotic Trials of
Intervention Effectiveness (CATIE) schizophrenia trial: clinical comparison of subgroups with and without the metabolic syndrome. Schizophr Res
80:918, 2005
Meyer JM, Davis VG, Goff DC, et al: Change in metabolic syndrome parameters
with antipsychotic treatment in the CATIE schizophrenia trial: prospective
data from phase 1. Schizophr Res 101:273286, 2008a
Meyer JM, Davis VG, McEvoy JP, et al: Impact of antipsychotic treatment on
nonfasting triglycerides in the CATIE schizophrenia trial. Schizophr Res
103:104109, 2008b
Miller BJ, Paschall CB, Svendsen DP: Mortality and medical comorbidity
among patients with serious mental illness. Psychiatr Serv 57:14821487,
Miller DD, McEvoy JP, Davis SM, et al: Clinical correlates of tardive dyskinesia
in schizophrenia: baseline data from the CATIE schizophrenia trial.
Schizophr Res 80:3343, 2005
Nasrallah H: A review of the effect of atypical antipsychotics on weight. Psychopharmacology 28:8396, 2003
Nasrallah HA, Meyer JM, Goff DC, et al: Low rates of treatment for hypertension, dyslipidemia, and diabetes in schizophrenia: data from the CATIE
schizophrenia trial sample at baseline. Schizophr Res 86:1522, 2006
Newcomer JW: Second-generation (atypical) antipsychotics and metabolic effects: a comprehensive literature review. CNS Drugs 19 (suppl 1):193,
Newcomer JW, Nasrallah HA, Loebel AD: The Atypical Antipsychotic Therapy
and Metabolic Issues National Survey: practice patterns and knowledge of
psychiatrists. J Clin Psychopharmacol 5 (suppl 1):S1S6, 2004
Stroup TS, Lieberman JA, McEvoy JP, et al: Effectiveness of olanzapine, quetiapine, risperidone, and ziprasidone in patients with chronic schizophrenia following discontinuation of a previous atypical antipsychotic. Am J
Psychiatry 163:611622, 2006
Stroup TS, Lieberman JA, McEvoy JP, et al: Results of phase 3 of the CATIE
schizophrenia trial. Schizophr Res (in press)
Swartz MS, Wagner HR, Swanson JW, et al: The effectiveness of antipsychotic
medications in patients who use or avoid illicit substances: results from the
CATIE study. Schizophr Res 100:3952, 2008
Umbricht DS, Pollack S, Kane JM: Clozapine and weight gain. J Clin Psychiatry
55 (suppl B):157160, 1994

This page intentionally left blank

Metabolic Disease,
Heart Disease,
and Related

This page intentionally left blank

Obesity and
Tony Cohn, M.B.Ch.B., M.Sc., F.R.C.P.C.

Obesity is a common and burgeoning health problem that has

an impact on both life expectancy and quality of life. Epidemic proportions of individuals are overweight and obese worldwide (James et al.
2001; Ogden et al. 2007). The prevalence of obesity is even greater
among individuals with schizophrenia, who are known to have increased morbidity and mortality from obesity-related diseases (McEvoy et al. 2005; Osby et al. 2000), as well as poor access to primary and
preventive health care (Druss et al. 2001; Meyer 2007). Individuals with
schizophrenia may be predisposed to obesity because of biological and
lifestyle factors but also require long-term treatment with antipsychotic
medications that can promote weight gain (Green et al. 2000; Newcomer 2005). With the growth in prescription of second-generation antipsychotic (SGA) medications over the past 10 years, antipsychoticassociated weight gain has emerged as an iatrogenic issue of major public health concern. The problem is compounded further by the fact that
commonly coprescribed psychiatric medications, antidepressants, and
mood stabilizers also contribute to weight gain (Aronne and Segal 2003;
Fava 2000). In addition, patients with schizophrenia are challenged by
cognitive and functional deficits that need to be considered in clinical
evaluation and in the delivery of obesity interventions. Although obesity medicine is now being recognized as a subspecialty in its own right,


Medical Illness and Schizophrenia

its application for individuals with serious mental illness, such as

schizophrenia, is in its infancy. The onus for management is falling to
mental health professionals. Clinicians of today who treat patients with
schizophrenia are challenged to become knowledgeable in the assessment, prevention, and treatment of obesity and its complications.
In this chapter, I review the clinical definition of obesity and its medical consequences; the epidemiology and determinants of obesity in the
general population and in individuals with schizophrenia; and the assessment, monitoring, and management of obesity for patients with
schizophrenia within a mental health context.

Definition and Measurement

Obesity is considered a disease by the World Health Organization
(2004), which defines obesity as a condition of excess body fat to the extent that health is impaired. However, measuring body fat accurately is
difficult and requires highly specialized procedures (Pi-Sunyer 2000).
For clinical purposes, obesity is usually defined by the body mass index
(BMI), a measure of weight that takes height into account.
In addition to BMI, a measure of the distribution of body fat is important, because the central accumulation of fat within the abdominal region (central, visceral, or abdominal adiposity) is a major factor that
influences the effect of increasing weight on health. Central adipose tissue is now seen as an endocrine organ that releases substances (adipokines) that have systemic effects. Adipokines include leptin, tumor
necrosis factor-alpha, adiponectin, C-reactive protein, interleukin-6,
and plasminogen activator inhibitor type 1.
Central adiposity is most accurately measured by computer-assisted
tomography or magnetic resonance imaging, which enable a distinction
to be made between subcutaneous and intra-abdominal fat. However, for
practical purposes, the simple waist circumference is the common clinical
proxy for central adiposity (Lemieux et al. 1996). Waist-to-hip ratio has
also been used for this purpose, but this measure is less favored because
racial and ethnic differences in hip size complicate interpretation (Ogden
et al. 2007). Both BMI and waist circumference are independently correlated with health outcomes, but waist circumference has been shown to
be an even better predictor of obesity-related disease.
BMI and waist circumference measures have limitations. Although
body fat is a continuous variable with an incremental health risk, BMI
and waist circumference are often used as categorical variables with cutoff values that reflect risk. Also, these measures do not adequately take



into account individual, gender, and racial or ethnic differences in body

type and body composition. BMI is calculated by dividing weight (measured in kilograms) by height (measured in meters) squared (i.e.,
BMI=kg/m2). Consensus exists among international health agencies, including the National Heart, Lung, and Blood Institute and the World
Heath Organization, regarding BMI cutoffs. A BMI of 2529.9 kg/m2
is considered overweight, and a BMI of 30 kg/m 2 is deemed obese.
Obesity is further classified as class 1 (BMI 3034.9 kg/m2), class 2 (BMI
3539.9 kg/m2), and class 3 or severe obesity (BMI 40 kg/m2). BMI is a
less accurate correlate for body fat when used for muscular individuals,
who may have a high BMI and normal or low body fat, and for very
short individuals (Pi-Sunyer 2000). In addition, because of differences in
body composition, for a given BMI, women tend to have a higher percentage of body fat than men, and older individuals have a higher percentage of body fat than younger individuals (Ogden et al. 2007).
Waist circumference is measured in the standing position with a simple tape measure placed around the waist. The waist is defined as midway between the top of the pelvis and the lowest rib. Measurement is
taken after a modest expiration. For individuals with a pendulous abdomen, the waist circumference can be measured using the maximum
girth, with the patient lying on his or her side on an examination table.
Cutoff points are 88 cm for women and 102 cm for men. Racial or ethnic
differences in body habitus can also influence the interpretation of
waist circumference; therefore, the International Diabetes Federation
has defined ethnic-specific cutoffs for waist circumference for the metabolic syndrome (Alberti et al. 2006).
Metabolic syndrome, also known as syndrome X and insulin-resistance
syndrome, is a clustering of risk markers reflecting a state of insulin resistance and compensatory hyperinsulinemia that has wide-ranging
medical consequences. According to the revised Adult Treatment Panel
III/American Heart Association definition, three out of the five following component criteria are required to diagnose metabolic syndrome
(Grundy et al. 2005):
1. Abdominal obesity (waist circumference) 88 cm in women and
102 cm in men
2. Blood pressure 130/85 mmHg or antihypertensive treatment
3. Fasting glucose 5.6 mmol/L (100 mg/dL) or antidiabetic treatment
4. Fasting triglycerides 1.8 mmol/L (150 mg/dL) or treatment
5. Fasting high-density lipoprotein (HDL) <1.03 mmol/L (40 mg/dL)
in men and <1.29 mmol/L (50 mg/dL) in women


Medical Illness and Schizophrenia

By the International Diabetes Federation definition of metabolic syndrome (Alberti et al. 2006), abdominal obesity is a necessary criterion,
and two additional criteria (as listed above) are required for the diagnosis. Waist circumference cutoffs vary by racial or ethnic group as
Europids or Middle Eastern: men >94 cm; women >80 cm
South Asian, Chinese: men > 90 cm; women >80 cm
Japanese: men >85 cm; women > 90 cm

Health Consequences of Obesity

Central adiposity is specifically associated with the development of insulin resistance and, consequently, type 2 diabetes, hyperlipidemia (especially raised triglycerides and low HDL), coronary heart disease,
hypertension, and polycystic ovarian disease. With obesity in general,
rates of gallbladder disease; respiratory disease, including obstructive
sleep apnea; certain cancers (colorectal, prostate, endometrial, gallbladder, cervical, ovarian, and breast); gout; and arthritis are also increased
(Pi-Sunyer 1993).
Obesity reduces life expectancy. In examining deaths attributable to
being overweight and obese, investigators found that more than 80% of
these deaths occurred in individuals with BMIs of at least 30 kg/m2 (Allison et al. 1999b). Years of life are lost primarily because of coronary
heart disease (Dorn et al. 1997). Shortened life span has been specifically
noted in patients with serious mental illness, such as schizophrenia
(Osby et al. 2000), with a recent report indicating as much as 2530 years
of potential life lost (Colton and Manderscheid 2006).
Obesity also has a negative impact on the quality of life, and this has
also been explored in samples of schizophrenia patients. Based on results from the Psychological Well-Being Index, Allison et al. (2003) reported that weight gain with antipsychotic treatment was related to
poorer quality of life as well as reduced well-being and vitality. Strassnig et al. (2003a) found that the physical component score of the Medical Outcomes Study 36-Item Short-Form Health Status Survey, a
measure of quality of life that assesses both physical and mental health,
was inversely correlated with the physical component score of patients
with schizophrenia, suggesting that the burden from obesity is primarily experienced as a physical problem, a finding that was confirmed by
Faulkner et al. (2007a) who found that, in addition, waist circumference
was a better predictor of reduced quality of life than BMI. Those with



obesity describe also being stigmatized. In the general population it has

been demonstrated that this phenomenon also occurs with health services, causing obese individuals to avoid visiting health care providers
(Drury and Louis 2002). Patients with schizophrenia face the additional
stigma of having a mental illness.
The economic burden of obesity is immense (Thompson and Wolf
2001). Direct health care costs are estimated to account for 5.5%7.0% of
national health expenditure in the United States and 2.0%3.5% in other
countries. When considering comorbidities, the three largest contributors to cost are hypertension, type 2 diabetes, and coronary artery disease. The cost of lost productivity due to obesity is almost as great as the
direct medical costs. In the treatment of patients with schizophrenia,
lipid lowering, antihypertensive, and antidiabetic medications are now
commonly coprescribed with antipsychotic medications, adding considerable costs to health care budgets.

Epidemiology of Obesity
In the general population, rates of overweight and obesity are rising at
an alarming rate, with no plateau in sight. In the United States, almost
two-thirds of adults (66%) were overweight or obese in 2004; 32% were
obese and 5% had class 3 obesity (Wang and Beydoun 2007). The prevalence of obesity has more than doubled since 1980 (Ogden et al. 2007).
In Canada in 2004, 65% of the adult population was overweight or
obese, but the overall rate of obesity (23%) was lower than in the United
States (Tjepkema 2006). Using linear regression, Wang and Beydoun
(2007) predicted that 75% of the U.S. adult population will be overweight or obese and 41% will be obese by 2015.
Studies have demonstrated that the distribution of obesity varies by
gender, racial and ethnic group, and socioeconomic status, with minority and low-socioeconomic-status groups having higher prevalence. In
2004, more men than women were overweight and obese, but women
had higher rates of obesity, especially in non-Hispanic black and Mexican American communities, where rates of obesity in women were as
high as 49% and 38%, respectively (Wang and Beydoun 2007). Weight
appears to vary more by socioeconomic status, race and ethnicity, and
nationality for women than for men, and some authors have suggested
that weight may be associated more closely with social and cultural
roles for women than for men (Ogden et al. 2007).
In patients with schizophrenia, rates of obesity are even higher than
in the general population and are estimated to fall between 40% and


Medical Illness and Schizophrenia

60% (Catapano and Castle 2004). The average BMI in recent schizophrenia trials has often approximated or been in the obese range, as illustrated in the Clinical Antipsychotic Trials of Intervention Effectiveness
(CATIE) study, where the recruitment strategy was designed to be
broadly representative of patients with schizophrenia. In the sample of
patients (n =1,460) recruited between January 2001 and December 2004
(Lieberman et al. 2005), the mean BMI at baseline was 29.7 kg/m2 (men
28.5 kg/m 2; women 33.0 kg/m2) (McEvoy et al. 2005). In a Canadian
study of 240 adult patients with schizophrenia recruited in 2003 and
2004 and treated with antipsychotic medication, Cohn et al. (2004b)
found that the mean BMI was 28.7 kg/m2 (men 28.0 kg/m 2; women
30.1 kg/m2) and the rate of obesity was 31% in men and 43% in women.
In the same study, the rate of metabolic syndrome was two times higher
in patients with schizophrenia than in age- and gender-matched general population controls. Similarly, baseline CATIE data revealed that
men and women with schizophrenia were 138% and 251% more likely,
respectively, to have metabolic syndrome than a National Health and
Nutrition Examination Survey matched sample (McEvoy et al. 2005).
To compare rates of obesity in schizophrenia patients with the general population prior to the widespread use of atypical antipsychotics,
Allison et al. (1999a) investigated BMI in patients with and without
schizophrenia using data from the mental health supplement of the
1989 National Health Interview Schedule. Men with schizophrenia had
a BMI similar to that of men without schizophrenia (26.14 vs. 25.63), but
women with schizophrenia had a higher BMI than women without
schizophrenia (27.36 vs. 24.50, P< 0.001). A consistent finding in these
various studies has been that women with schizophrenia have a differentially higher rate of obesity and metabolic syndrome relative to the
general population than do men with schizophrenia. Although this
finding is similar to what has been found in various minority and lowincome groups, the reasons have yet to be fully explored.

Determinants and Mechanisms of Obesity

General and Evolutionary Factors
Weight maintenance is governed by a formula that is deceptively simple: energy input must equal energy output. Energy input is essentially
food consumed, whereas energy output comprises a number of elements of which physical activity is the main modifiable factor and accounts for about 30% of energy output. Other energy output factors
include basal metabolism (60% of energy output), diet-induced thermo-



genesis (10% of energy output), and adaptive thermogenesis or alteration in metabolism related to environmental, psychological, or other
influences, such as habitual body movement (Groff and Gropper 1999).
(Subtle akathisia, which was a common side effect with classic antipsychotics, probably accounted for significant energy expenditure.) Energy
balance is tightly controlled but tilted toward weight gain because of
physiological adaptations that defend against weight loss, so that a very
small positive energy balance of 100 calories a day (the equivalent of
one slice of bread) results in a gain of 4.5 kg in 1 year.
Evolutionary concepts best explain the evolving epidemic of obesity
in developed and developing countries (Ogden et al. 2007; Pi-Sunyer
2003). The human body has evolved to survive the threat of famine
from a time when food was scarce and a great deal of effort and energy
was required to procure food. In contrast, many people now have an
abundant supply of cheap, easily obtainable, and palatable food, and at
the same time less requirement and opportunity for exercise. What was
an advantage in times of food scarcity becomes a liability when the food
supply is abundant or when there is change from a traditional diet to a
more westernized diet. When food is scarce and weight is lost, the body
has a powerful biological drive to regain weight: basal metabolic rate
drops, and hunger signals increase. In contrast, when excessive weight
is gained, biological signals are muted; there is no great increase in
basal metabolism, and individuals tend to become less rather than more
active. The thrifty phenotype hypothesis, as proposed by Hales and
Barker (2001), is an extension of this view: With poor intrauterine nutritional conditions, the child efficiently conserves energy and is prepared
for survival in an environment where food resources will be scarce.
Then, when exposed to an environment of food excess, such an individual who is efficient at conserving energy would be predisposed to
obesity. This hypothesis has gained considerable favor in explaining
susceptibility to metabolic disturbance, particularly type 2 diabetes.
The thrifty phenotype hypothesis also suggests a mechanism
whereby patients with schizophrenia may be predisposed to obesity
and associated metabolic disturbance. Epidemiological research has
shown an association between intrauterine and childhood undernutrition and the lifetime risk of schizophrenia. For example, there was a
doubling of the rate of schizophrenia following both the Dutch winter
famine (1945) and the Chinese famine (19591961) (Hoek et al. 1998; St
Clair et al. 2005). In addition, studies have demonstrated a relationship
between schizophrenia and small size at birth, thinness in childhood,
and short stature (Gunnell et al. 2003; Wahlbeck et al. 2001).


Medical Illness and Schizophrenia

Genetics and Neurophysiology

Estimates of the heritability of obesity range from 30% to 70% (Martnez-Hernndez et al. 2007). Different sets of monozygotic twins that
are overfed differ in the degree to which excess calories are stored as fat,
but within each set of twins the susceptibility to gain weight in a given
environment is similar. Although a number of rare syndromes associated with obesity have been found to be caused by discrete genetic defects or chromosomal abnormalities (e.g., Prader-Willi syndrome), the
more common forms of obesity likely involve more complex interactions between the environment and a multitude of polymorphisms located in genes and candidate regions throughout the genome.
Genetic investigation has focused on genes that encode and regulate
food intake, energy expenditure, and adipogenesis. Food intake is regulated by peptides such as leptin and neuropeptide Y that bind with
receptors in the hypothalamus, as well as by a range of peptides synthesized along the gastrointestinal tract, such as ghrelin and cholecystokinin (Schwartz et al. 2000).
Research on energy expenditure has focused on the sympathetic nervous system, in particular the -adrenergic receptor gene family, and on
uncoupling proteins at the mitochondrial level (Martnez-Hernndez et
al. 2007). In adipogenesis, peroxisome proliferatoractivated receptor-
(PPAR-) has become a focus of investigation (Marti et al. 2002).

Habitual diet and level of physical activity are important factors that
contribute to energy balance and the development of obesity. Dietary
intake has been studied in schizophrenia samples, but results have been
conflicting. In a cohort of 102 middle-aged patients with schizophrenia
living in the community, English researchers reported that patients had
a diet higher in fat and lower in fiber than the general population
(Brown et al. 1999). Using a 24-hour diet recall in 146 outpatients with
schizophrenia living in Pittsburgh, Strassnig et al. (2003b) reported that
patients consumed more food than matched control subjects, but the
macronutrient composition (percentage of calories from fat, protein,
and carbohydrate) was similar in both groups. In contrast, Henderson
et al. (2006) used 4-day food diariesa more rigorous sampling strategyto evaluate the diet of 88 patients with schizophrenia attending a
community mental health clinic in Boston and made comparisons with
general population controls matched by age, gender, and ethnicity. The
patients consumed significantly fewer calories and less carbohydrate,
protein, total fat, and saturated fat than the general population controls.



Habitual diet is notoriously difficult to measure and is prone to various

degrees of underreporting (ONeil 2001), which may help explain these
discrepant findings.
Patients with schizophrenia are often financially disadvantaged, and
diet can vary based on the local availability of low-cost foods. Patients
are often housed in group homes and boarding houses, where food, if
it is provided, is cost-based and tends to be high in starch and sugar. In
a Toronto study, Cohn et al. (2004a) found that the diet of 162 outpatients with schizophrenia was similar in calorie count to that of individuals in the general population but was high in simple carbohydrates,
particularly easily available chips and sodaa diet that has been implicated in development of metabolic syndrome (McKeown et al. 2004).
Findings regarding physical inactivity in samples of patients with
schizophrenia have been more consistent. Multiple investigators have reported low levels of physical activity in patients compared with the general population (Brown et al. 1999; Cohn et al. 2004a; Daumit et al. 2005;
Ussher et al. 2007). Daumit et al. (2005) pointed out that walking tended
to be the only form of physical activity and that women and those without regular social contact were more inactive. Ussher et al. (2007) indicated that patients reported low levels of social support for exercising
and that fatigue and illness were the most common barriers to activity.

Obesogenic Environment
An important consideration is how the environment influences lifestyle
and the development of obesity. The obesogenicity of an environment is the sum of influences that the surroundings, opportunities, or
conditions of life have on promoting obesity in individuals or populations (Swinburn et al. 1999). The concept of the obesogenic environment
has been applied to the built environment, particularly in cities and
suburbs (Papas et al. 2007). Attempts have been made to analyze obesogenic factors in school systems and to promote healthier environments with access to nutritious foods and the promotion of physical
activity (Nollen et al. 2007). However, little attention has been paid to
the obesogenic environment in mental health settings. How environments are structured in psychiatric group homes, inpatient units, and
community clinics, with provision of opportunities for physical exercise and healthy eating, is an important focus for future research.

Antipsychotic-Induced Weight Gain

Antipsychotic medications are a mainstay in the treatment of schizophrenia, and in most cases long-term maintenance treatment is advised.


Medical Illness and Schizophrenia

Significant weight gain has been recognized as a common side effect,

particularly with low-potency classic antipsychotics and certain second-generation antipsychotics. This issue, which is now recognized as
a public health concern (Green et al. 2000), has garnered much attention, particularly over the past 10 years.
In a comprehensive review of the published literature, including randomized clinical trials, Newcomer (2005) identified clozapine and olanzapine as being associated with the greatest risk of clinically significant
weight gain; risperidone, quetiapine, amisulpride, and zotepine as having intermediate risk; and ziprasidone and aripiprazole as having the
lowest risk. This view is consistent with the findings of a consensus development conference among the American Diabetes Association, American
Psychiatric Association, American Association of Clinical Endocrinologists, and the North American Association for the Study of Obesity. The
report, published simultaneously in two journals in 2004, similarly assigned antipsychotics to high, intermediate, and low risk for weight gain
and attributed the associated risk for type 2 diabetes and dyslipidemia in
the same rank order (American Diabetes Association et al. 2004).
Antipsychotics are associated with both short-term (1012 weeks)
and long-term weight gain. Allison et al. (1999c) conducted a metaanalysis to compare the weight-gain liability of antipsychotics over the
short term (10 weeks). Using a random-effects model, average weight
gain was 4.5 kg with clozapine, 4.1 kg with olanzapine, 2.5 kg with
chlorpromazine, 2.1 kg with risperidone, 1.1 kg with haloperidol, and
0.05 kg with ziprasidone (Allison et al. 1999c). Other investigators have
reported the short-term weight gain to be 2.3 kg with quetiapine (Jones
et al. 1999) and 0.6 kg with aripiprazole (Marder et al. 2003).
Weight gain tends to be rapid, particularly during the first 3 months
of treatment, with significant weight increases often noted within a few
weeks of treatment (Newcomer 2005). Weight gain generally plateaus
within 9 months or 1 year, although clozapine tends to have a longer
weight-gain trajectory, with patients continuing to gain weight for up
to 5 years (Henderson et al. 2000).
Considerable interindividual variability in weight gain occurs, and
average weight gain belies the fact that individual patients can gain a
very large amount of weight (sometimes in excess of 45 kg) on a particular medication. Susceptibility to weight gain may be genetically determined, and at a future time genetic testing might be used for predicting
which patients are prone to excessive weight gain (Mller and Kennedy
Clinical experience has shown that the following variables predict
weight gain:



1. Type of antipsychotic: Antipsychotics differ considerably in weightgain liability.

2. Previous antipsychotic exposure: Patients who are antipsychotic naive
or switched from a classic antipsychotic or an SGA with low
weight-gain risk to an antipsychotic with high weight-gain risk are
more susceptible to weight gain. In contrast, those switched from an
antipsychotic with high to one with low weight-gain liability tend
to lose weight.
3. Age: Children, adolescents, or young adults are more likely to gain
weight; older patients are less susceptible to weight gain.
4. Baseline weight: Patients with a low baseline weight tend to gain
more weight.
5. Race and ethnicity: Certain groups, including those of African, South
Asian, Hispanic, and Native American descent, have higher baseline rates of metabolic disturbance and may also be more susceptible to weight gain.
6. Clinical response: Although controversial, a number of investigators
have found an association between weight gain and antipsychotic
7. Antipsychotic dose and plasma level: These are generally not correlated
with weight gain when prescribed within the recommended dose
range, although clozapine may be an exception.
Longer-term weight gain is difficult to compare across antipsychotics
because of patient attrition and the difficulties adjusting for variables
such as prior antipsychotic exposure. In a randomized, controlled trial
of 263 first-episode patients with no or limited prior exposure to antipsychotics that compared olanzapine (mean modal dose 9.1 mg/day)
and haloperidol (mean modal dose 4.4 mg/day), the weight gain after
2 years of treatment for olanzapine was 10.2 kg using a last-observationcarried-forward analysis and 14 kg for observed cases. For haloperidol,
weight gain was 4 kg using a last-observation-carried-forward analysis
and 7.5 kg for observed cases (Zipursky et al. 2005).
The neuroendocrine mechanisms that underlie antipsychoticinduced weight gain are not fully understood. However, the relationship between neurotransmitter-binding profiles of different antipsychotics and their propensity to induce weight gain has been explored
(Nasrallah 2008). Wirshing et al. (1999) astutely observed a correlation
between weight gain and antipsychotic binding affinity for the histamine-1 (H1) receptor. Kroeze et al. (2003) more recently concluded that
H 1 receptor affinity is positively correlated with weight gain among
both typical and atypical antipsychotics. Others have confirmed this


Medical Illness and Schizophrenia

(S.F. Kim et al. 2007; Matsui-Sakata et al. 2005). S.F. Kim et al. (2007), using a mouse model, found that antipsychotics that induce weight gain
selectively activate hypothalamic AMP-kinase via the H1 receptors. H1
activity is involved in appetite regulation and sedation, both relevant to
weight gain. These findings may guide antipsychotic drug development toward drugs less implicit in weight gain. Other targets for investigation include the 5-HT2C gene (De Luca et al. 2007; Reynolds et al.
2002) and the SNAP-25, -2a, leptin, and GNB3 genes (Mller and
Kennedy 2006).

Managing Obesity in Schizophrenia

The management of obesity includes an understanding and evaluation
of risk factors, prevention of weight gain, metabolic assessment, and
ongoing monitoring of the patient, as well as the application of pharmacological and nonpharmacological strategies of prevention and treatment. Clearly, psychiatrists cannot do this work alone. What is required
is to develop a system of assessment and management in which roles
and tasks are clearly defined and often delegated.

Tooling the Office or Clinic

Clinicians should think through how to set up the environment, equipment, systems, and personnel within the office or clinic to facilitate care
of patients who are obese (Kushner 2007). The clinic or office should be
accessible for these patients. Hallways and doors should not be too narrow, washrooms need to be large enough, and the waiting area and examination or interview room should have a number of sturdy, armless
chairs. Posters, artwork, and magazines should not overly promote
thinness. However, posters promoting balanced healthy eating and
physical activity are appropriate, as is information about weight-loss
programs and services.
The examination area should be equipped with appropriate equipment for assessing obese patients. Most office or hospital scales do not
measure above 300350 lb (135160 kg). Clinics should obtain a scale
with greater capacity and position it so the patient can be weighed privately and away from public view. A wall-mounted measure or an extendable height meter attached to the scale should be available for
measuring the patients height. Also, a variety of cuff sizes should be
provided for measuring blood pressure. Miscuffing is the most common error in measuring blood pressure, and undercuffing accounts for
the majority of errors (Manning et al. 1983). Tape measures should be



available for measuring waist circumference, and a calculator or chart

is needed for determining BMI.
Each office or clinic needs to establish a procedure for metabolic
monitoring. The paper or electronic health record should allow for the
recording and tracking of screening medical history questions, clinical
measures, and laboratory tests used in metabolic monitoring. Commercial tracking sheets, usually sponsored by pharmaceutical companies,
are available. The clinic where I work has developed an electronic tool
that is integrated into the medical record and the patient care plan that
allows for the tracking of metabolic parameters and performs calculations such as whether the patient has metabolic syndrome, Framingham risk score, target lipid values, and so forth (Khoury et al. 2008).
Forms or computer systems should be readily available to order needed
laboratory tests. Systems need to be in place for referring patients for
primary care or specialist consultation and for referring patients for nutrition counseling and to professionals who promote physical activity.
Larger clinics or programs should develop group-based interventions
to promote healthy lifestyle changes.
The roles and tasks of personnel should be clearly defined. Registered nurses can obtain the vital measurements, including weight,
height, waist circumference, and blood pressure, and provide health
teaching. Nurse practitioners can provide most components of primary
health care. Co-location of primary health care (primary care physician
or nurse practitioner) within psychiatric services is an excellent model
for integrating health care (Druss et al. 2001). Office secretaries and assistants can facilitate referrals, remind patients of appointments, and
distribute written material. Although the exact role and function of
each team member varies from one setting to another, what is most important is that care is coordinated and that individuals work together as
a team.

Assessment and Metabolic Monitoring of the Patient

Patients can be assessed at various pointsas a new antipsychotic medication is prescribed, after substantial weight has been gained, upon admission to hospital, at annual examinations, at each visit, and so on.
Ideally, to inform care, these various assessments should be recorded
and tracked, along with prescribed medications and intervention efforts. Electronic record systems are particularly helpful in this regard
and can also provide reminders when assessments are due.
We (Cohn and Sernyak 2006) have reviewed recommendations regarding metabolic monitoring published in various countries. Table 41


Medical Illness and Schizophrenia

provides a structure informed by these guidelines and used in our clinic

for assessment of patients who are overweight or obese and are being
treated with psychiatric medications associated with weight gain (antipsychotics, mood stabilizers, and antidepressants).
Monitoring, which consists of medical history, measurements, and
laboratory tests, is conducted at the first opportunitywhen the patient
first presents at the office of clinic or at the time of hospital admission.
Monitoring is then repeated annually. A new baseline is established
when a new antipsychotic (or mood medication) is started, with a follow-up evaluation 3 months later. Weight is measured at every visit
and graphed over time across encounters (inpatient and outpatient).
Diet, activity, motivation, and environmental factors (see General and
Lifestyle column in Table 41) are evaluated during clinic visits.
To evaluate diet and physical activity, we have found it useful in our
clinic to have the patient, with the support of family or caregivers, complete 3 days of food records, a standard food frequency questionnaire,
and a physical activity questionnaire prior to the initial appointment.
We have been able to validate the reliability of the International Physical Activity Questionnaire (Craig et al. 2003) in a sample of patients
with schizophrenia (Faulkner et al. 2006). These assessments are mailed
to the patient before the appointment, reviewed during the initial consultation, and used to inform the development of a treatment plan. We
have also found it useful to get a baseline measure of fitness by using
the 6-minute walking test (ATS Committee on Proficiency Standards
2002), a simple structured assessment of the distance an individual
walks over 6 minutes.

Strategies for the Prevention and Management of Obesity

Behavioral lifestyle interventions focused on diet and physical activity
should be combined with biological (mostly pharmacological) strategies to benefit patients who are obese (Faulkner and Cohn 2006). Figure
41 provides a recommended outline for the sequencing of interventions. Behavioral interventions are discussed in detail in Chapter 8,
Behavioral Treatments for Weight Management of Patients With
Schizophrenia. This section focuses on biological strategies.
Medication Choice. Weight-gain liability should be factored in when
choosing psychotropic medications. The choice of psychotropic medication may have the greatest influence on weight gain and associated
metabolic disturbance. Good evidence exists for a range of weight-gain
liability for antipsychotic medications (Allison et al. 1999c; Newcomer


Metabolic assessment of patients treated using antipsychotic medications

Medical history

Current antipsychotic medication(s)
Previous antipsychotic trials
Newly prescribed antipsychotic
Coprescribed mood stabilizers
Coprescribed antidepressants
Coprescribed lipid medication(s)
Coprescribed diabetes medication(s)
Coprescribed antihypertensive

Current psychiatric and medical

Income and source
Living situation: Are meals provided?
Who does the shopping and cooking?
Recreational opportunities?
Self-report of weight change over time

Physical activity
Premailed International
Physical Activity Questionnaire 1
Interest in and barriers to increasing
physical activity
6-Minute walking test 2
Premailed food frequency
questionnaire and 3-day food diary
Interest in and feasibility of making
dietary changes

Examination (measurement) Laboratory tests

Waist circumference
Blood pressure

For nondiabetic
(No known diagnosis
of diabetes)
Fasting glucosea
Measuring guide
Fasting insulin
Height: No shoes
Fasting lipid
Weight: No shoes; single layer
of clothing
(Total, HDL, and
Waist circumference: Standing LDL cholesterol
and triglycerides)
Blood pressure: Average of
two measures. Patient seated For diabetic patients
Risk factors
and resting for at least 5
Fasting glucose
minutes, >30 minutes from
Family history of diabetes
Fasting lipid
last cigarette or coffee. Arm
History of gestational diabetes
relaxed and supported at
History of coronary vascular disease
heart level. Choose
(angina, myocardial infarction, or
Random urine
appropriate cuff size as
indicated by cuff markers.
Cigarette smoking


Patients with impaired fasting glucose (5.66.9 mmol or 100125 mg/dL) have follow-up glucose tolerance test (fasting glucose and 2-hour glucose after
75-g oral glucose drink). bIf elevated, urine albumin/creatinine ratio.
Craig et al. 2003; Faulkner et al. 2006. 2ATS Committee on Proficiency Standards 2002.
Source. Adapted from Cohn and Sernyak 2006.


General and lifestyle


Medical Illness and Schizophrenia

2005). Antidepressant and mood-stabilizing medications also contribute to weight gain (Aronne and Segal 2003; Fava 2000) because these
medications are commonly coprescribed with antipsychotics. Lithium
carbonate, valproic acid, selective serotonin reuptake inhibitors, and
mirtazapine have all been implicated in weight gain.
Young, first-episode patients are more susceptible to antipsychoticinduced weight gain (Zipursky et al. 2005) and also tend to have a robust response to antipsychotic medications (Perkins et al. 2004). Initial
treatment with antipsychotic medications known to induce significant
weight gain should be avoided. Care should also be taken in the choice
of concomitant psychiatric medications. In patients who have gained
significant weight or developed metabolic complications, all medications should be reviewed from both metabolic and psychiatric perspectives (Faulkner and Cohn 2006).
Medication Switching. Switching from one antipsychotic medication
to another can have significant effects on body weight and metabolic
parameters. Reviews (Weiden 2007; Weiden and Buckley 2007) suggest
that switching to low-liability antipsychotics (ziprasidone or aripiprazole) can have a more substantial impact on reversing weight gain and
lipid disturbance than other intervention strategies, particularly if the
weight gain and metabolic derangement clearly occurred during the
course of prior treatment with an antipsychotic medication with greater
metabolic liability. The metabolic benefit from switching is therefore
dependent on the prior treatment.
Outpatients, predominantly with schizophrenia, were switched
from a previous antipsychoticolanzapine (n = 104), risperidone
(n = 58), or conventional antipsychotic (n =108)to ziprasidone (mean
dose 91 mg/day) and reevaluated after 6 weeks (Weiden et al. 2003a).
Weight and metabolic benefit depended on which drug the patients
were originally taking. Those switched from olanzapine experienced
the greatest metabolic benefit, with significant reductions in body
weight (1.76 kg), nonfasting total cholesterol, and triglycerides. Those
switched from risperidone experienced less weight loss (0.86 kg) but
similar reductions in nonfasting total cholesterol and triglycerides.
However, those switched from conventional antipsychotics had no
change in weight or lipid measures. A companion paper from the same
data set (Weiden et al. 2003b) supported the view that stable but symptomatic patients can safely be switched from a previous antipsychotic
using a variety of switching strategies (e.g., abrupt discontinuation of
previous antipsychotic or gradual cross-titration), often with benefit to
psychiatric symptomatology. An extension of this study provided long-



Consider metabolic risk in

choice of antipsychotic

Provide patient, family, and

caregiver education about
potential metabolic side effects

Conduct baseline
screening and regular
metabolic monitoring

Provide lifestyle

With significant weight gain/

metabolic disturbance: Review
all medications from a metabolic
and psychiatric perspective;
consider risk/benefit of switch

Provide referral to structured

lifestyle intervention focused
on diet, physical activity,
and behavior change

If lifestyle intervention is unsuccessful,

consider adjunctive pharmacotherapy;
for severe obesity and significant
disability, consider evaluation
and workup for obesity surgery

FIGURE 41. Prevention and management of antipsychotic-related

metabolic disturbance.

Adapted from Faulkner and Cohn 2006.

term data for 153 patients followed for 58 weeks (Weiden et al. 2008).
Average weight loss from baseline was 10.3% (9.8 kg) and 7.8% (6.9 kg)
if the previous antipsychotic was olanzapine or risperidone, respectively. Those patients switched from olanzapine and risperidone also
sustained significant reductions in nonfasting cholesterol and triglycerides, with maximum reductions occurring rapidly at 6 weeks after the
switch. Those switched from conventional antipsychotics to ziprasidone did not change in body weight or lipid measures.
Data from the CATIE study (Lieberman et al. 2005) also provide
evidence in a randomized trial of differential effects on body weight


Medical Illness and Schizophrenia

when antipsychotics are switched. Patients switched to olanzapine

gained 0.9 kg per month. Patients switched to quetiapine and risperidone also gained weight, but less (0.23 kg per month on quetiapine and
0.18 kg per month on risperidone). Those switched to ziprasidone and
perphenazine lost 0.13 and 0.09 kg per month, respectively. Patients
switched to aripiprazole (Casey et al. 2003) experience weight changes
very similar to patients who are switched to ziprasidone. Again, weight
change depended on the previous antipsychotic.
Taken together, data from these studies indicate that weight loss and
metabolic benefit can be achieved by switching medication, particularly
if these problems clearly developed while a patient was taking the prior
medication. However, switching is always hazardous because patients
can decompensate. The biggest challenge is the patient who has been
doing well psychiatrically while taking an antipsychotic but has gained
weight or developed metabolic complications. In such a case, the risks
of switching may be greater, and the patient and family need to be informed of the potential risks and benefits and then carefully consider
whether to switch medication. Discontinuing or switching antipsychotics becomes a critical issue when diabetes or diabetic ketoacidosis can
be clearly linked to an antipsychotic prescription, because resolution of
diabetes can occur (Jin et al. 2004).
Medication Addition. One strategy to prevent weight gain or to promote weight loss is to add a medication associated with weight loss to
the antipsychotic medication regimen. In the general population, controlled clinical trials have established modest efficacy for obesity drugs
in combination with lifestyle therapy (Li et al. 2005). Orlistat (lipase inhibitor) and sibutramine (serotonin/dopamine/norepinephrine reuptake inhibitor) are the only drugs currently approved for long-term
weight loss. Average weight loss in the general population (psychiatric
patients were excluded) at 12 months was 2.7 kg (95% CI 2.33.1) for individuals taking orlistat and 4.3 kg (95% CI 3.64.9) for those taking
sibutramine (Padwal et al. 2003). In the context of antipsychoticinduced weight gain, brief (616 weeks) randomized controlled trials
have been conducted in patients with schizophrenia using a number of
compounds, as summarized in Table 42. In a Cochrane review,
Faulkner et al. (2007b) focused on weight change to compare efficacy reported in studies identified using a systematic search and retrieval protocol. Unfortunately, no long-term studies have been published.
In these studies, the weight loss medication is either added when the
antipsychotic is first prescribed to prevent weight gain (e.g., reboxetine
in first-episode psychosis being treated with olanzapine) (Poyurovsky



et al. 2003, 2007) or added later when patients have already gained substantial weight (e.g., the use of metformin in children and adolescents
whose weight had increased by more than 10% in less than 1 year while
taking atypical antipsychotics) (Klein et al. 2006). Many of these studies
have been conducted in patients taking clozapine and olanzapine.
The studies have been heterogeneous, and results have been mixed.
When there is a weight benefit, it tends to be modest (approximately
24 kg over 3 months). Negative results have been seen with fluoxetine
and reboxetine addition. Sibutramine was found effective with patients
on olanzapine (Henderson et al. 2005) (but the benefits were reversed
when patients stopped taking sibutramine). Sibutramine was ineffective in clozapine-treated patients (Henderson et al. 2007). Topiramate
has been more consistently effective for weight loss in open-label as
well as double-blind studies (J.H. Kim et al. 2006; Ko et al. 2005), and
doses of 200 mg/day have been more effective than 100 mg/day (Ko et
al. 2005). However, topiramate has many side effects, both common
(paresthesia) and potentially serious (cognitive impairment, narrowangle glaucoma, and metabolic acidosis).
Metformin may be more helpful in reducing or preventing weight
gain during the early stages of treatment for psychosis or in younger individuals. Metformin was found to be effective in reducing weight gain
when coprescribed with olanzapine in patients treated for a first psychotic episode (Wu et al. 2008) and in reversing weight gain and improving insulin sensitivity in children and adolescents who had recently
gained weight during treatment with atypical antipsychotics (Klein et
al. 2006). However, metformin addition was not able to prevent weight
gain in an adult sample of schizophrenia patients with a more chronic
course who were switched to olanzapine (Baptista et al. 2006).
The release of rimonabant, a selective cannabinoid CB1 receptor antagonist, was highly anticipated because of promising phase 3 results in
weight loss, smoking cessation, and improvement of metabolic profiles
in patients with metabolic syndrome, all of which would benefit patients with schizophrenia (Cox 2005). Rimonabant is currently approved in Europe but not in the United States or Canada. Because
recent data indicate significant psychiatric adverse effects, primarily
depression and anxiety, there was a recommendation against rimonabant approval in 2007 by the FDA (Rumsfeld and Nallamothu 2008).
Given that H1 receptor affinity appears to be implicated in antipsychotic-associated weight gain, interest has been shown in the use of betahistine, an H1 receptor agonist and H3 antagonist, to prevent weight
gain. Promising results were reported in a small open-label study
(Poyurovsky et al. 2005). Placebo-controlled studies are still needed.

Antipsychotic trial duration

(H2 receptor antagonist)
(removed from market)

Olanzapine 16 weeks
(Deberdt et al. 2005)
Olanzapine 6 weeks
(Poyurovsky et al. 2004)
Typical depots 12 weeks
(Goodall et al. 1988)
Olanzapine 8 weeks
(Poyurovsky et al. 2002)
Olanzapine 16 weeks
(Bustillo et al. 2003)
Olanzapine 12 weeks
(Wu et al. 2008)
Atypical AP 16 weeks
(Klein et al. 2006)
Olanzapine 14 weeks
(Baptista et al. 2006)
Olanzapine 8 weeks
(Atmaca et al. 2003)
Olanzapine 16 weeks
(Cavazzoni et al. 2003)


(H2 receptor antagonist)

Agent dose: n for intervention group: outcome, expressed as

weight change relative to placebo [95% CI]
100300 mg: n= 35: 1.7 [3.9, 0.5] kg
40 mg OD: n= 7: no effect
30 mg OD: n= 9: 2.6 [5.5, 0.1] kg
20 mg OD: n= 15: no effect
60 mg OD: n= 15: no effect
250 mg tid: n= 18: 5.0 [7.3, 2.7] kg
500 mg OD: n= 9 (ages 1017): 4.1 [7.3, 1.0] kg
8501,750 mg: n= 20: no effect
150 mg bid: n =18: 6.8 [7.9, 5.7] kg
300 mg bid: n =58: less wt gain, 3 and 4 wk; effect lost, 16 wk
150 mg bid: n =57: no effect

Medical Illness and Schizophrenia

Added agent


TABLE 42. Randomized, controlled trials of medication addition for weight gain in schizophrenia

TABLE 42. Randomized, controlled trials of medication addition for weight gain in schizophrenia (continued)
Antipsychotic trial duration

(removed from market)

Clozapine 12 weeks
(Borovicka et al. 2002)
Olanzapine 6 weeks
(Poyurovsky et al. 2003)
Olanzapine 6 weeks
(Poyurovsky et al. 2007)
Clozapine 12 weeks
(Henderson et al. 2007)
Mixed 16 weeks
(Weiden et al. 2003c)
Olanzapine 12 weeks
(Henderson et al. 2005)
Atypical AP 12 weeks
(Ko et al. 2005)

Sibutramine (serotonin/
dopamine RI)


Agent dose: n for intervention group: outcome, expressed as

weight change relative to placebo [95% CI]
75 mg: n =8: no effect
4 mg/d: n =13: 3.0 [5.6, 0.5] kg
4 mg/d: n =31: 1.5 [2.9, 0.1] kg
15 mg: n =11: no effect
15 mg: n =29: no effect


Added agent

15 mg: n =19: 4.6 [5.2, 4.0] kg

200 mg/d: n =17: 5.1 [7.4, 2.7] kg
100 mg/d: n =16: 1.4 [4.2, 1.5] kg

Note. RI= reuptake inhibitor; SNRI=serotonin-norepinephrine reuptake inhibitor; SSRI= selective serotonin reuptake inhibitor.

Adapted from Faulkner and Cohn 2006.



Medical Illness and Schizophrenia

Insufficient data are available to suggest the routine use of medication additions (Faulkner and Cohn 2006). Each medication brings the
potential of additional side effects, effect sizes are small, and accessing
and covering the costs of these drugs are often problematic. Medication
additions should, however, be contemplated when other approaches
such as behavioral interventions have been exhausted and switching to
a different antipsychotic has been ruled out.
Obesity Surgery. Surgery is an effective intervention for treating severe
obesity (Schneider and Mun 2005; Sharma and Iacobellis 2006), resulting in more significant and longer-lasting loss of weight than other
treatments. Criteria for selecting patients for bariatric surgery include
the following: 1) BMI > 40 kg/m2 or BMI > 35 kg/m 2 with significant
obesity-related comorbidities; 2) repeated failure of other nonsurgical
interventions; 3) psychologically stable patient with realistic expectations; and 4) absence of alcoholism, other addictions, or major psychopathology (NIH Conference 1991). These criteria make it difficult for
patients with schizophrenia and other forms of severe mental illness to
access surgical intervention.
Surgical procedures can involve gastric restriction using vertical or
adjustable bands to create a small neogastric pouch, or shortening the
functional length of the intestinal surface for nutrient absorption, as in
gastric bypass or biliopancreatic diversion. The current gold standard
for bariatric surgery is the Roux-en-Y gastric bypass, which is a combination of the restrictive and bypass procedures (Sharma and Iacobellis
2006). Patients with schizophrenia require advocacy and very close
medical follow-up and support to access and derive maximum benefit
from any of these effective interventions.

Although obesity is endemic in the general population, its prevalence
and consequences are amplified for patients with schizophrenia. Responsibility for organizing health promotion and obesity interventions
lies with the providers of psychiatric care, because these services are often not otherwise accessible and because medications used to treat
schizophrenia contribute to the development of obesity. Genetic, medication, lifestyle, and environmental factors are determinants of obesity
in patients with schizophrenia, but genetic prediction of an individuals
susceptibility to gain weight taking a particular medication is not yet
a reality, and practitioners have to rely on clinical evaluation and



ongoing monitoring. Clinicians and mental health services are challenged to set up systems for the assessment, monitoring, prevention,
and treatment of obesity.

Key Clinical Points

Obesity affects life expectancy and quality of life and is more common
in patients with schizophrenia than in the general population.

Genetics, psychotropic medications, lifestyle, and the environment are

contributing factors.

Clinical assessment of the patient includes a review of medical history,

diet, and level of physical activity; physical measurements (weight,
height, waist circumference, and blood pressure); and laboratory tests.
These parameters should be tracked in an organized manner along with
prescribed psychotropic medication (metabolic monitoring).

The clinic or office should be set up to facilitate care of obese patients.

Considerations include the environment, equipment, systems, and use
of personnel.

Prevention and management of obesity in schizophrenia require consideration of choice of psychotropic medication, health counseling and
behavioral interventions, medication switching, medication addition,
and obesity surgery.

Alberti KG, Zimmet P, Shaw J: Metabolic syndromea new world-wide definition: a consensus statement from the International Diabetes Federation. Diabet Med 23:469480, 2006
Allison DB, Fontaine KR, Heo M, et al: The distribution of body mass index
among individuals with and without schizophrenia. J Clin Psychiatry
60:215220, 1999a
Allison DB, Fontaine KR, Manson JE, et al: Annual deaths attributable to obesity
in the United States. JAMA 282:15301538, 1999b
Allison DB, Mentore JL, Heo M, et al: Antipsychotic-induced weight gain: a
comprehensive research synthesis. Am J Psychiatry 156:16861696, 1999c
Allison DB, Mackell JA, McDonnell DD: The impact of weight gain on quality
of life among persons with schizophrenia. Psychiatr Serv 54:565567, 2003


Medical Illness and Schizophrenia

American Diabetes Association, American Psychiatric Association, American

Association of Clinical Endocrinologists, et al: Consensus Development
Conference on Antipsychotic Drugs and Obesity and Diabetes. J Clin Psychiatry 65:267272, 2004
Aronne LJ, Segal KR: Weight gain in the treatment of mood disorders. J Clin
Psychiatry 64 (suppl 8):2229, 2003
Atmaca M, Kuloglu M, Tezcan E, et al: Nizatidine treatment and its relationship
with leptin levels in patients with olanzapine-induced weight gain. Hum
Psychopharmacol 18:457461, 2003
ATS Committee on Proficiency Standards for Clinical Pulmonary Function Laboratories: ATS statement: guidelines for the six-minute walk test. Am J
Respir Crit Care Med 166:111117, 2002
Baptista T, Martinez J, Lacruz A, et al: Metformin for prevention of weight gain
and insulin resistance with olanzapine: a double-blind placebo-controlled
trial. Can J Psychiatry 51:192196, 2006
Borovicka MC, Fuller MA, Konicki PE, et al: Phenylpropanolamine appears not
to promote weight loss in patients with schizophrenia who have gained
weight during clozapine treatment. J Clin Psychiatry 63:345348, 2002
Brown S, Birtwistle J, Roe L, et al: The unhealthy lifestyle of people with schizophrenia. Psychol Med 29:697701, 1999
Bustillo JR, Lauriello J, Parker K, et al: Treatment of weight gain with fluoxetine
in olanzapine-treated schizophrenic outpatients. Neuropsychopharmacology 28:527529, 2003
Casey DE, Carson WH, Saha AR, et al: Switching patients to aripiprazole from
other antipsychotic agents: a multicenter randomized study. Psychopharmacology (Berl) 166:391399, 2003
Catapano L, Castle D: Obesity in schizophrenia: what can be done about it?
Australas Psychiatry 12:2325, 2004
Cavazzoni P, Tanaka Y, Roychowdhury SM, et al: Nizatidine for prevention of
weight gain with olanzapine: a double-blind placebo-controlled trial. Eur
Neuropsychopharmacol 13:8185, 2003
Cohn TA, Sernyak MJ: Metabolic monitoring for patients treated with antipsychotic medications. Can J Psychiatry 51:492501, 2006
Cohn T, Greenwood C, Remington G: Body composition, physical activity and diet
in patients with schizophrenia and related disorders under antipsychotic treatment. Paris, France, Collegium Internationale Neuro-Psychopharmacologicum
(CINP). Int J Neuropsychopharmacol 7 (suppl 2):S426, 2004a
Cohn T, Prudhomme D, Streiner D, et al: Characterizing coronary heart disease
risk in chronic schizophrenia: high prevalence of the metabolic syndrome.
Can J Psychiatry 49:753760, 2004b
Colton CW, Manderscheid RW: Congruencies in increased mortality rates,
years of potential life lost, and causes of death among public mental health
clients in eight states. Prev Chronic Dis 3:A42, 2006
Cox SL: Rimonabant hydrochloride: an investigational agent for the management of cardiovascular risk factors. Drugs Today (Barc) 41:499508, 2005



Craig CL, Marshall AL, Sjstrm M, et al: International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc 35:1381
1395, 2003
Daumit GL, Goldberg RW, Anthony C, et al: Physical activity patterns in adults
with severe mental illness. J Nerv Ment Dis 193:641646, 2005
De Luca V, Mueller DJ, de Bartolomeis A, et al: Association of the HTR2C gene
and antipsychotic induced weight gain: a meta-analysis. Int J Neuropsychopharmacol 10:697704, 2007
Deberdt W, Winokur A, Cavazzoni PA, et al: Amantadine for weight gain associated with olanzapine treatment. Eur Neuropsychopharmacol 15:1321, 2005
Dorn JM, Schisterman EF, Winkelstein W Jr, et al: Body mass index and mortality in a general population sample of men and women: the Buffalo Health
Study. Am J Epidemiol 146:919931, 1997
Drury CA, Louis M: Exploring the association between body weight, stigma of
obesity, and health care avoidance. J Am Acad Nurse Pract 14:554561,
Druss BG, Rohrbaugh RM, Levinson CM, et al: Integrated medical care for patients with serious psychiatric illness: a randomized trial. Arch Gen Psychiatry 58:861868, 2001
Faulkner G, Cohn TA: Pharmacologic and nonpharmacologic strategies for
weight gain and metabolic disturbances in patients treated with antipsychotic medications. Can J Psychiatry 51:502511, 2006
Faulkner G, Cohn T, Remington G: Validation of a physical activity assessment
tool for individuals with schizophrenia. Schizophr Res 82:225231, 2006
Faulkner G, Cohn T, Remington G, et al: Body mass index, waist circumference
and quality of life in individuals with schizophrenia. Schizophr Res 90:174
178, 2007a
Faulkner G, Cohn T, Remington G: Interventions to reduce weight gain in
schizophrenia. Cochrane Database Syst Rev, Issue 1. Art. No.: CD005148.
DOI: 10.1002/14651858.CD005148.pub2, 2007b
Fava M: Weight gain and antidepressants. J Clin Psychiatry 61 (suppl 11):3741,
Goodall E, Oxtoby C, Richards R, et al: A clinical trial of the efficacy and acceptability of D-fenfluramine in the treatment of neuroleptic-induced obesity.
Br J Psychiatry 153:208213, 1988
Green AI, Patel JK, Goisman RM, et al: Weight gain from novel antipsychotic
drugs: need for action. Gen Hosp Psychiatry 22:224235, 2000
Groff JL, Gropper SS: Advanced Nutrition and Human Metabolism, 3rd Edition. St. Paul, MN, Wadsworth, 1999
Grundy SM, Cleeman JI, Daniels SR, et al: Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and
Blood Institute scientific statement. Circulation 112:27352752, 2005
Gunnell D, Rasmussen F, Fouskakis D, et al: Patterns of fetal and childhood
growth and the development of psychosis in young males: a cohort study.
Am J Epidemiol 158:291300, 2003


Medical Illness and Schizophrenia

Hales CN, Barker DJ: The thrifty phenotype hypothesis. Br Med Bull 60:520,
Henderson DC, Cagliero E, Gray C, et al: Clozapine, diabetes mellitus, weight
gain, and lipid abnormalities: a five-year naturalistic study. Am J Psychiatry 157:975981, 2000
Henderson DC, Copeland PM, Daley TB, et al: A double-blind, placebo-controlled
trial of sibutramine for olanzapine-associated weight gain. Am J Psychiatry
162:954962, 2005
Henderson DC, Borba CP, Daley TB, et al: Dietary intake profile of patients with
schizophrenia. Ann Clin Psychiatry 18:99105, 2006
Henderson DC, Fan X, Copeland PM, et al: A double-blind, placebo-controlled
trial of sibutramine for clozapine-associated weight gain. Acta Psychiatr
Scand 115:101105, 2007
Hoek HW, Brown AS, Susser E: The Dutch famine and schizophrenia spectrum
disorders. Soc Psychiatry Psychiatr Epidemiol 33:373379, 1998
James PT, Leach R, Kalamara E, et al: The worldwide obesity epidemic. Obes
Res 9 (suppl 4):228S233S, 2001
Jin H, Meyer JM, Jeste DV: Atypical antipsychotics and glucose dysregulation:
a systematic review. Schizophr Res 71:195212, 2004
Jones A, Rak I, Raniwalla J, et al: Weight changes in patients treated with Seroquel (quetiapine). Paper presented at the annual meeting of the American
College of Neuropsychopharmacology, Acapulco, Mexico, December 1999
Khoury A, Sproule BA, Cohn TA: Development and implementation of the
Metabolic Health Monitor at the Centre for Addiction and Mental Health.
Paper Presented at BC Psychopharmacology Conference. Vancouver, British Columbia, 2008
Kim JH, Yim SJ, Nam JH: A 12-week, randomized, open-label, parallel-group
trial of topiramate in limiting weight gain during olanzapine treatment in
patients with schizophrenia. Schizophr Res 82:115117, 2006
Kim SF, Huang AS, Snowman AM, et al: Antipsychotic drug-induced weight
gain mediated by histamine H 1 receptor-linked activation of hypothalamic
AMP-kinase. Proc Natl Acad Sci U S A 104:34563459, 2007
Klein DJ, Cottingham EM, Sorter M, et al: A randomized, double-blind, placebo-controlled trial of metformin treatment of weight gain associated with
initiation of atypical antipsychotic therapy in children and adolescents. Am
J Psychiatry 163:20722079, 2006
Ko YH, Joe SH, Jung IK, et al: Topiramate as an adjuvant treatment with atypical antipsychotics in schizophrenic patients experiencing weight gain. Clin
Neuropharmacol 28:169175, 2005
Kroeze WK, Hufeisen SJ, Popadak BA, et al: H 1-histamine receptor affinity predicts short-term weight gain for typical and atypical antipsychotic drugs.
Neuropsychopharmacology 28:519526, 2003
Kushner RF: Getting the office ready for the patient, in A Clinical Guide for
Management of Overweight and Obese Children and Adults. Edited by
Apovian CM, Lenders CM. Boca Raton, FL, CRC Press, 2007, pp 113



Lemieux S, Prudhomme D, Bouchard C, et al: A single threshold value of waist

girth identifies normal-weight and overweight subjects with excess visceral adipose tissue. Am J Clin Nutr 64:685693, 1996
Li Z, Maglione M, Tu W, et al: Meta-analysis: pharmacologic treatment of obesity. Ann Intern Med 142:532546, 2005
Lieberman JA, Stroup TS, McEvoy JP, et al: Effectiveness of antipsychotic drugs
in patients with chronic schizophrenia. N Engl J Med 353:12091223, 2005
Manning DM, Kuchirka C, Kaminski J: Miscuffing: inappropriate blood pressure cuff application. Circulation 68:763766, 1983
Marder SR, McQuade RD, Stock E, et al: Aripiprazole in the treatment of schizophrenia: safety and tolerability in short-term, placebo-controlled trials.
Schizophr Res 61:123136, 2003
Marti A, Corbaln MS, Martnez-Gonzlez MA, et al: CHO intake alters obesity
risk associated with Pro12Ala polymorphism of PPAR gamma gene.
J Physiol Biochem 58:219220, 2002
Martnez-Hernndez A, Enrquez L, Moreno-Moreno MJ, et al: Genetics of obesity. Public Health Nutr 10:11381144, 2007
Matsui-Sakata A, Ohtani H, Sawada Y: Receptor occupancy-based analysis of
the contributions of various receptors to antipsychotics-induced weight
gain and diabetes mellitus. Drug Metab Pharmacokinet 20:368378, 2005
McEvoy JP, Meyer JM, Goff DC, et al: Prevalence of the metabolic syndrome in
patients with schizophrenia: baseline results from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) schizophrenia trial and
comparison with national estimates from NHANES III. Schizophr Res
80:1932, 2005
McKeown NM, Meigs JB, Liu S, et al: Carbohydrate nutrition, insulin resistance,
and the prevalence of the metabolic syndrome in the Framingham offspring cohort. Diabetes Care 27:538546, 2004
Meyer JM: Strategies for the long-term treatment of schizophrenia: real-world
lessons from the CATIE trial. J Clin Psychiatry 68 (suppl 1):2833, 2007
Mller DJ, Kennedy JL: Genetics of antipsychotic treatment emergent weight
gain in schizophrenia. Pharmacogenomics 7:863887, 2006
Nasrallah HA: Atypical antipsychotic-induced metabolic side effects: insights
from receptor-binding profiles. Mol Psychiatry 13:2735, 2008
Newcomer JW: Second-generation (atypical) antipsychotics and metabolic effects:
a comprehensive literature review. CNS Drugs 19 (suppl 1):193, 2005
NIH conference: gastrointestinal surgery for severe obesity. Consensus Development Conference Panel. Ann Intern Med 115:956961, 1991
Nollen NL, Befort CA, Snow P, et al: The school food environment and adolescent obesity: qualitative insights from high school principals and food service personnel. Int J Behav Nutr Phys Act 4:18, 2007
Ogden CL, Yanovski SZ, Carroll MD, et al: The epidemiology of obesity. Gastroenterology 132:20872102, 2007
ONeil PM: Assessing dietary intake in the management of obesity. Obes Res 9
(suppl 5):361S366S; discussion 373S374S, 2001


Medical Illness and Schizophrenia

Osby U, Correia N, Brandt L, et al: Mortality and causes of death in schizophrenia in Stockholm county, Sweden. Schizophr Res 45:2128, 2000
Padwal R, Rucker D, Li S, et al: Long-term pharmacotherapy for obesity and
overweight. Cochrane Database Syst Rev, Issue 3. Art. No.: CD004094.
DOI: 10.1002/14651858.CD004094.pub2, 2003
Papas MA, Alberg AJ, Ewing R, et al: The built environment and obesity. Epidemiol Rev 29:129143, 2007
Perkins D, Lieberman J, Gu H, et al: Predictors of antipsychotic treatment response in patients with first-episode schizophrenia, schizoaffective and
schizophreniform disorders. Br J Psychiatry 185:1824, 2004
Pi-Sunyer FX: Medical hazards of obesity. Ann Intern Med 119:655660, 1993
Pi-Sunyer FX: Obesity: criteria and classification. Proc Nutr Soc 59:505509,
Pi-Sunyer FX: A clinical view of the obesity problem. Science 299:859860, 2003
Poyurovsky M, Pashinian A, Gil-Ad I, et al: Olanzapine-induced weight gain
in patients with first-episode schizophrenia: a double-blind, placebocontrolled study of fluoxetine addition. Am J Psychiatry 159:10581060,
Poyurovsky M, Isaacs I, Fuchs C, et al: Attenuation of olanzapine-induced
weight gain with reboxetine in patients with schizophrenia: a doubleblind, placebo-controlled study. Am J Psychiatry 160:297302, 2003
Poyurovsky M, Tal V, Maayan R, et al: The effect of famotidine addition on
olanzapine-induced weight gain in first-episode schizophrenia patients: a
double-blind placebo-controlled pilot study. Eur Neuropsychopharmacol
14:332336, 2004
Poyurovsky M, Pashinian A, Levi A, et al: The effect of betahistine, a histamine
H 1 receptor agonist/H 3 antagonist, on olanzapine-induced weight gain in
first-episode schizophrenia patients. Int Clin Psychopharmacol 20:101103,
Poyurovsky M, Fuchs C, Pashinian A, et al: Attenuating effect of reboxetine on
appetite and weight gain in olanzapine-treated schizophrenia patients: a
double-blind placebo-controlled study. Psychopharmacology (Berl)
192:441448, 2007
Reynolds GP, Zhang ZJ, Zhang XB: Association of antipsychotic drug-induced
weight gain with a 5-HT2C receptor gene polymorphism. Lancet 359:2086
2087, 2002
Rumsfeld JS, Nallamothu BK: The hope and fear of rimonabant. JAMA
299:16011602, 2008
Schneider BE, Mun EC: Surgical management of morbid obesity. Diabetes Care
28:475480, 2005
Schwartz MW, Woods SC, Porte D Jr, et al: Central nervous system control of
food intake. Nature 404:661671, 2000
Sharma AM, Iacobellis G: Treatment of obesity: a challenging task. Contrib
Nephrol 151:212220, 2006



St Clair D, Xu M, Wang P, et al: Rates of adult schizophrenia following prenatal

exposure to the Chinese famine of 19591961. JAMA 294:557562, 2005
Strassnig M, Brar JS, Ganguli R: Body mass index and quality of life in community-dwelling patients with schizophrenia. Schizophr Res 62:7376, 2003a
Strassnig M, Brar JS, Ganguli R: Nutritional assessment of patients with schizophrenia: a preliminary study. Schizophr Bull 29:393397, 2003b
Swinburn B, Egger G, Raza F: Dissecting obesogenic environments: the development and application of a framework for identifying and prioritizing environmental interventions for obesity. Prev Med 29:563570, 1999
Thompson D, Wolf AM: The medical-care cost burden of obesity. Obes Rev
2:189197, 2001
Tjepkema M: Adult obesity. Health Rep 17:925, 2006
Ussher M, Stanbury L, Cheeseman V, et al: Physical activity preferences and
perceived barriers to activity among persons with severe mental illness in
the United Kingdom. Psychiatr Serv 58:405408, 2007
Wahlbeck K, Forsen T, Osmond C, et al: Association of schizophrenia with low
maternal body mass index, small size at birth, and thinness during childhood. Arch Gen Psychiatry 58:4852, 2001
Wang Y, Beydoun MA: The obesity epidemic in the United Statesgender, age,
socioeconomic, racial/ethnic, and geographic characteristics: a systematic
review and meta-regression analysis. Epidemiol Rev 29:628, 2007
Weiden PJ: Switching antipsychotics as a treatment strategy for antipsychoticinduced weight gain and dyslipidemia. J Clin Psychiatry 68 (suppl 4):34
39, 2007
Weiden PJ, Buckley PF: Reducing the burden of side effects during long-term
antipsychotic therapy: the role of switching medications. J Clin Psychiatry 68 (suppl 6):1423, 2007
Weiden PJ, Daniel DG, Simpson G, et al: Improvement in indices of health status in outpatients with schizophrenia switched to ziprasidone. J Clin Psychopharmacol 23:595600, 2003a
Weiden PJ, Simpson GM, Potkin SG, et al: Effectiveness of switching to ziprasidone for stable but symptomatic outpatients with schizophrenia. J Clin
Psychiatry 64:580588, 2003b
Weiden PJ, Newcomer JW, Loebel AD, et al: Long-term changes in weight and
plasma lipids during maintenance treatment with ziprasidone. Neuropsychopharmacology 33:985994, 2008
Wirshing DA, Wirshing WC, Kysar L, et al: Novel antipsychotics: comparison
of weight gain liabilities. J Clin Psychiatry 60:358363, 1999
World Health Organization: Obesity: Preventing and Managing the Global Epidemic: Report of a WHO Consultation. Geneva, Switzerland, World
Health Organization 2004
Wu RR, Zhao JP, Guo XF, et al: Metformin addition attenuates olanzapineinduced weight gain in drug-naive first-episode schizophrenia patients: a
double-blind, placebo-controlled study. Am J Psychiatry 165:352358, 2008


Medical Illness and Schizophrenia

Zipursky RB, Gu H, Green AI, et al: Course and predictors of weight gain in
people with first-episode psychosis treated with olanzapine or haloperidol.
Br J Psychiatry 187:537543, 2005

Glucose Intolerance
and Diabetes in
Patients With
David C. Henderson, M.D.
Kathleen Miley, B.S.

In recent years, tremendous progress has been made in the treatment of schizophrenia and other psychotic disorders. In particular, the
introduction of atypical antipsychotic agents has helped numerous patients gain control of both positive and negative symptoms of schizophrenia. Studies also suggest that the newer antipsychotic agents
increase compliance, prevent relapse (Csernansky et al. 2002), and may
offer improvement in the treatment of cognitive dysfunction (Meltzer
2001). These treatment successes, combined with a lower risk of extrapyramidal symptoms (e.g., parkinsonism) and improved control of
positive and negative symptoms, have led many psychiatrists to favor
the newer, atypical antipsychotic medications over the older, typical
agents. Furthermore, the atypical antipsychotic agent clozapine has
been found to be a highly effective agent in improving symptoms for
those patients who have had treatment resistance to other antipsychotic
medications (Lewis et al. 2006).


Medical Illness and Schizophrenia

The use of atypical antipsychotic medications, however, has not

proven to be trouble free. Numerous reports in the psychiatric and
medical literature have suggested an association between atypical antipsychotic agents and impaired glucose metabolism, diabetes mellitus,
and diabetic ketoacidosis (Elias and Hofflich 2008; Henderson 2002;
Henderson et al. 2007; Perez-Iglesias et al. 2007; Saddichha et al. 2008).
After a brief review of diabetes pathophysiology, especially type 2 diabetes mellitus, and its short- and long-term medical consequences, we
examine glucose intolerance and diabetes in individuals with schizophrenia before and following the introduction of typical and atypical
antipsychotic medications. Finally, we make recommendations for
monitoring and screening for glucose intolerance and diabetes in patients with schizophrenia.

Diagnostic Criteria and

Pathophysiology of Diabetes Mellitus
Both diabetes and glucose intolerance are characterized by problems in
glucose-insulin regulation. According to the American Diabetes Association (ADA), if an individual has the typical symptoms of diabetes,
such as polyuria, polydipsia, and unexplained weight loss, plus a casual plasma glucose level 200 mg/dL (11.1 mmol/L), a fasting plasma
glucose level 126 mg/dL (7.0 mmol/L), or a 2-hour plasma glucose
level 200 mg/dL (11.1 mmol/L) post 75-g oral glucose load, then he or
she has diabetes mellitus (American Diabetes Association 2008). Similarly, if an individual has a 2-hour plasma glucose level 140 mg/dL
and <200 mg/dL post 75-g oral glucose load, he or she is said to have impaired glucose tolerance. Impaired fasting glucose is defined as a level
from 100 to 125 mg/dL, whereas fasting glucose values <100 mg/dL are
considered to be in the normal range (American Diabetes Association
Two pathophysiological processes can lead to the development of
glucose intolerance and diabetes mellitus, the first being a problem
with insulin secretion, as in type 1 and the later stages of type 2 diabetes
mellitus, and the second being a problem with insulin action, otherwise
known as insulin resistance, as seen in type 2 diabetes mellitus (Lebovitz 2001). In the first case, the beta cells of the pancreas have been destroyed by an autoimmune process, and hyperglycemia occurs because
not enough insulin is secreted to facilitate glucose uptake in skeletal
muscle tissue, inhibit glucose production in hepatic tissue, and suppress lipolysis in adipose tissue. In type 2 diabetes mellitus, although

Glucose Intolerance and Diabetes


enough insulin may be secreted, insulin resistance prevents the insulin

from working at the sites of skeletal muscle, hepatic, and adipose tissues. Indeed, a reactive hyperinsulinemia may occur, which may help
control plasma glucose levels initially, but eventually the beta cells in
the pancreas begin to deteriorate at a genetically determined rate, and
the compensatory hyperinsulinemia decreases (Lebovitz 2001). As this
occurs, the postprandial plasma glucose levels increase progressively,
and the individual progresses from normal glucose tolerance to impaired glucose tolerance to type 2 diabetes mellitus. Thus, patients with
type 2 diabetes mellitus have both insulin resistance and an insulin
secretory deficit due to decreased beta cell function.
Both type 1 and type 2 diabetes mellitus are diseases with multifactorial inheritance. Researchers are confident that no single gene causes
type 1 or type 2 diabetes mellitus, and that development of this heterogeneous group of disorders is likely the result of multiple genes and environmental factors. For example, the interaction of genetic factors such
as beta cell abnormalities and a predisposition for central obesity with
excess caloric intake, high fat ingestion, and decreased physical activity
can lead to type 2 diabetes mellitus in some individuals (Lebovitz 2001).
Drugs are another environmental factor that can cause diabetes mellitus by either destroying the beta cells in the pancreas or by causing insulin resistance by one of the following mechanisms: 1) increasing
appetite and caloric intake, leading to obesity; 2) altering fat distribution (central obesity is a risk factor for diabetes mellitus); 3) decreasing
physical activity because of increased sedation; 4) decreasing oxidative
metabolism in tissues; 5) interfering with the insulin action cascade; 6)
increasing counterregulatory hormones; or 7) increasing free fatty acid
release from adipose tissue (Lebovitz 2001). Later in this chapter, we
discuss the possible mechanisms by which atypical antipsychotic medications may cause glucose intolerance or diabetes mellitus, exacerbate
existing diabetes, or induce diabetic ketoacidosis in patients with

Medical Complications of Glucose

Intolerance and Diabetes Mellitus
Hyperglycemia and diabetes mellitus are associated with acute and
chronic complications associated with significant morbidity and mortality. Diabetic ketoacidosis, an acute complication of diabetes mellitus,
is seen more often in patients with type 1 than in patients with type 2
diabetes mellitus and is a serious and potentially fatal complication.


Medical Illness and Schizophrenia

Ketoacidosis is defined by low serum pH (7.35), low serum bicarbonate levels (15), and an anion gap in the presence of ketonemia (Westphal 1996). The diabetic patient is also susceptible to a variety of chronic
complications that affect the cardiovascular system, nervous system,
eyes, kidneys, and wound-healing capabilities. Most of these complications are a result of microvascular and macrovascular disease that is
more extensive and appears much earlier in the diabetic patient than in
the general population (Newcomer 2001).
Macrovascular disease in the form of atherosclerosis increases the
risk of cardiovascular and cerebrovascular events such as myocardial
infarction and stroke, thus accounting for much of the disability and
death among diabetic patients (Haupt and Newcomer 2001; Henderson
2001). Results from studies conducted with large samples of patients
without diabetes show that even modest increases in fasting plasma
glucose levels that do not meet the diagnostic criteria for diabetes mellitus put patients at increased risk for cardiovascular disease and complications, including coronary artery disease, myocardial infarction,
and other vascular problems, as well as increased risk for cardiovascular death (Gerstein et al. 1999). In peripheral sites, atherosclerosis can
cause claudication and diabetic foot, a condition in which patients
develop nonhealing ulcers that are prone to infection on their lower extremities and feet as a result of vascular insufficiency and sensory deficits from impairments in the peripheral nervous system. Diabetic
neuropathy is a complication that contributes significantly to morbidity
in diabetic patients because it not only contributes to diabetic foot but
also can affect any part of the nervous system, resulting in sensory deficits, paresthesias, motor abnormalities, or autonomic dysfunction
(Henderson 2001). A large percentage of diabetic patients also experience ophthalmic complications such as diabetic retinopathy and diseases of the anterior chamber that affect vision (e.g., cataracts), leading
to blindness and significant disability.
Finally, many diabetic patients suffer a great deal from microvascular- and macrovascular-induced nephropathy, which can cause hypertension, proteinuria, and a decrease in the glomerular filtration rate,
leading to renal failure. Indeed, diabetic nephropathy is the most prevalent cause of end-stage renal failure (Schernthaner 2008) and is a leading cause of morbidity and mortality in diabetic patients.

Metabolic Syndrome
Obesity (especially visceral), insulin resistance, and dyslipidemia,
along with hypertension (Henderson et al. 2004), are major, modifiable

Glucose Intolerance and Diabetes


cardiovascular risk factors associated with macrovascular complications. These metabolic abnormalities and hypertension are key components of metabolic syndrome, which is highly predictive of overt type 2
diabetes mellitus and cardiovascular disease (Wannamethee et al.
2005). The prevalence of metabolic syndrome and cardiovascular disease in the schizophrenia population taking atypical antipsychotics is
much higher than in the general population (Bobes et al. 2007; Casey et
al. 2004; Newcomer 2004, 2007a, 2007b; Newcomer and Haupt 2006;
Osby et al. 2000). McEvoy et al. (2005) found that the age-adjusted prevalence rates of metabolic syndrome in patients in the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) schizophrenia trial
were 40.9% based on criteria from the third report of the National Cholesterol Education Program Expert Panel on Detection, Evaluation, and
Treatment of High Blood Cholesterol in Adults and 42.7% based on criteria from the American Heart Association.

Lifestyle Factors
Allison et al. (2003) reported that patients treated with antipsychotic
medications who have experienced weight gain have a reduced quality
of life, poorer self-reported general health, and decreased vitality. By
increasing a patients risk of obesity, antipsychotic agents may be placing patients at risk for associated morbidity and mortality (Pi-Sunyer
2002). Patients who gain greater than 7% of their total body weight are
at risk for developing hypertension and type 2 diabetes mellitus. The
risk for diabetes mellitus is increased approximately twofold in mildly
obese, fivefold in moderately obese, and 10-fold in severely obese persons (Pi-Sunyer 1993).
Although patients with schizophrenia tend to consume fewer calories, their food choices differ from those of the general population and
they are less likely to make healthy dietary choices (Brown et al. 1999;
Henderson et al. 2006a; McCreadie et al. 1998; Pereira et al. 1997; Strassnig et al. 2003). Studies suggest that high dietary saturated fat and low
omega-3 polyunsaturated fatty acid play a key role in the development
of type 2 diabetes mellitus (Peet 2004a) and contribute to a poor clinical
outcome for people with schizophrenia (Christensen and Christensen
1988; Peet 2003, 2004b). No studies have reported on whether atypical
antipsychotic-induced insulin resistance has any relationship to poor dietary profile in this population. However, one study showed that clozapine-treated patients with schizophrenia consumed almost twice as
much sugar as those taking other antipsychotic agents (Stokes and Peet
2004). Coccurello et al. (2006) and Fell et al. (2007) observed that olanza-


Medical Illness and Schizophrenia

pine treatment in mice altered dietary macronutrient selection such that

they preferred a diet high in fat and sugar. Identification of any link between poor dietary profile and glucose metabolism abnormalities induced by atypical antipsychotic drugs in the schizophrenia population
would be helpful to emphasize preventive measures through dietary
modification prior to taking aggressive therapeutic approaches.
Central obesity is a significant risk factor for the development of type
2 diabetes mellitus in patients with schizophrenia as well as in the general population. The prevalence of obesity and overweight in people
with type 2 diabetes mellitus is as high as 80%90% (Carolino et al.
2008). Apparently, however, the duration of obesity is a greater determinant of risk than that conferred by simply being obese. That is, if a patient gains a considerable amount of weight and maintains this weight,
his or her risk of developing hyperglycemia or type 2 diabetes mellitus
appears to be increased (Henderson 2001).

Glucose Intolerance and Diabetes Mellitus

in Patients With Schizophrenia
Findings From the Pre-Antipsychotic Era
Early reports dating back to the 1920s, before the use of antipsychotic
agents, suggest that individuals with schizophrenia and other psychotic
disorders exhibited an elevated risk for developing glucose intolerance
or diabetes mellitus (Braceland et al. 1945; Haupt and Newcomer 2001;
Marinow 1971; Saddichha et al. 2008; Waitzkin 1966). Specifically, the
reports indicate a pattern of insulin resistance in patients with schizophrenia independent of adverse medication effects. These studies, however, suffered from several methodological problems: flaws in the
diagnostic criteria for schizophrenia, and lack of control for age, weight,
fat distribution, ethnicity, diet, or exercise, all of which are variables
now known to play a role in an individuals risk for developing glucoregulatory disturbances (Haupt and Newcomer 2001). Thus, because no
well-controlled studies exist, the debate continues as to whether individuals with schizophrenia, when unmedicated, are at increased risk
for developing diabetes compared with the general population.

Conventional Antipsychotic Agents, Diabetes, and

Glucose Intolerance
Conventional antipsychotic agents, which have primarily antidopaminergic activity, may alter glucose-insulin homeostasis (Hgg et al. 1998).

Glucose Intolerance and Diabetes


In particular, the low-potency phenothiazines may induce diabetes mellitus or aggravate existing diabetes mellitus (Hgg et al. 1998; Haupt and
Newcomer 2001). Because of this finding, chlorpromazine has been used to
prevent hypoglycemia in patients with malignant insulinoma. Furthermore, chlorpromazine has been shown to induce hyperglycemia in healthy
volunteers as well as in patients with latent diabetes (Hgg et al. 1998).
Other conventional antipsychotic agents, such as the high-potency
agent haloperidol, are associated with a decrease in the prevalence rate
of diabetes in the schizophrenia population. For example, Mukherjee et
al. (1996) found an overall diabetes prevalence rate of 15% in 95 patients
with schizophrenia. In patients younger than age 50 years, there were
no cases of diabetes mellitus. For patients ages 5059 years, however,
the prevalence rate was 12.9%, and for patients ages 6069 years, the
prevalence rate was 18.9%. Finally, for those ages 7074 years, the prevalence rate was 16.7%. After controlling for age, gender, and cumulative
duration of antipsychotic treatment, Mukherjee et al. (1996) found that
medication-free patients were more likely to develop diabetes mellitus
than were those receiving treatment with conventional agents. Of note,
the prevalence rates quoted in this study exceeded those expected for
type 2 diabetes mellitus in the general population, lending further evidence to the argument that schizophrenia may indeed be an independent risk factor for the development of diabetes mellitus (Henderson
2002; Henderson et al. 2005a).
Several other reports have suggested higher rates of type 2 diabetes
mellitus in patients with schizophrenia than in the general population,
even before widespread use of atypical antipsychotics (Dixon et al.
2000). In a sample of 26 medication-naive first-episode schizophrenia
patients, Ryan et al. (2003) found evidence of insulin resistance and impaired glucose tolerance, with 15% of subjects exhibiting abnormal glucose metabolism on an oral glucose tolerance test. Several more recent
studies of neuroleptic-naive first-episode schizophrenia patients have
also confirmed the presence of glucose-insulin homeostasis dysfunction in this patient population (Cohn et al. 2006; Spelman et al. 2007; van
Nimwegen et al. 2008; Venkatasubramanian et al. 2007).
In summary, examination of reports discussing the prevalence rates
of diabetes mellitus or impaired glucose tolerance in patients with
schizophrenia shows that among the conventional antipsychotic
agents, the effects of glucose regulation may vary in magnitude across
individual agents. Specifically, the phenothiazines may increase a patients susceptibility for developing diabetes mellitus, but no significant
association has been found between diabetes mellitus and the use of
other conventional antipsychotic medications.


Medical Illness and Schizophrenia

Atypical Antipsychotic Medications, Diabetes, and

Glucose Intolerance
The greatest benefits of the introduction of the atypical antipsychotic
medications as a groupclozapine, olanzapine, quetiapine, risperidone, and ziprasidonehave been improvements in the treatment of
negative symptoms and concomitant improvements in cognitive function (Lewis et al. 2006; Meltzer 2001). The improvements in cognitive
function, which include the aspects of executive function, verbal fluency, attention, and memory and learning, can lead to improved functioning in both home and work environments (Meltzer 2001). However,
these benefits have not come without a cost, because many atypical antipsychotic agents are associated with significant weight gain, which
has an adverse effect on health and medication compliance.
In the Clinical Antipsychotic Trials for Intervention Effectiveness
(CATIE) schizophrenia study, funded by the National Institute of Mental Health, 74% of patients discontinued their study medication (either
perphenazine, olanzapine, quetiapine, risperidone, or ziprasidone) before 18 months in phase 1 (Lieberman et al. 2005). The time to discontinuation of treatment for any cause was significantly longer in the group
taking olanzapine. Although symptom outcomes were similar in phase 1
for all groups, as indicated by total scores on the Positive and Negative
Syndrome Scale for Schizophrenia, the groups demonstrated significant differences in weight gain and significant glucose metabolism
changes (with olanzapine resulting in the greatest negative effect). In
the phase 2 tolerability arm, the time to treatment discontinuation was
longer for both olanzapine (6.3 months) and risperidone (7 months)
than for quetiapine (4 months). In subjects who had discontinued their
previous phase 1 drug because of side effects and were switched to another drug, olanzapine was more effective than quetiapine and ziprasidone, and risperidone was more effective than quetiapine (Stroup et
al. 2006). However, olanzapine, compared with the other agents, was
again associated with greater and more significant weight gain and increases in lipids and glucose (Nasrallah 2006). The CATIE trial also
found that at baseline, the subjects with schizophrenia had elevated
rates of metabolic syndrome and its individual components compared
with matched controls from the general population (McEvoy et al.
2005). Additionally, elevated risk for cardiovascular disease was observed at baseline compared with this same pool of matched controls
(Goff et al. 2005). (For more discussion of the CATIE trial findings, see
Chapter 3, Medical Outcomes From the CATIE Schizophrenia

Glucose Intolerance and Diabetes


Although CATIE has been the largest randomized, double-blind,

non-industry sponsored trial to report metabolic data, numerous earlier
reports in the medical and psychiatric literature have linked the use of
atypical antipsychotic agents to the development of glucose intolerance, new-onset diabetes mellitus, diabetic ketoacidosis, and exacerbation of existing type 1 or type 2 diabetes mellitus, even in patients who
are not obese (Henderson 2002; Henderson et al. 2007). Koller and colleagues from the U.S. Food and Drug Administration research program
conducted a MedWatch surveillance program analysis and reported
384 cases of clozapine-related diabetes (Koller et al. 2001), of which 242
were new-onset diabetes cases and 54 were cases of exacerbation in patients with preexisting diabetes. Their report also identified 80 cases of
clozapine-induced probable diabetic ketoacidosis and 25 deaths. Of
note, the majority of diabetic ketoacidosis episodes occurred within the
first 6 months of clozapine treatment, and one patient developed diabetes mellitus immediately after he accidentally took a 500-mg dose of
clozapine. In another study, Koller and Doraiswamy (2002) found a total of 289 cases of olanzapine-related diabetes, of which 225 were newonset diabetes, with 100 cases of diabetic ketoacidosis and 25 deaths.
Given that the MedWatch surveillance program is able to identify only
a small percentage of cases, the incidence of clozapine- and olanzapineinduced diabetes mellitus and diabetic ketoacidosis is likely much
greater than reported.
Hgg et al. (1998), using an oral glucose tolerance test, found that
12% of patients treated with clozapine developed type 2 diabetes and
10% developed impaired glucose tolerance compared to 6% and 3%, respectively, of patients treated with conventional depot antipsychotic
medications. In an analysis of 45 published cases of new-onset diabetes
mellitus or diabetic ketoacidosis following treatment with atypical
antipsychotic agents, Jin et al. (2002) reported that 20 patients had
received clozapine, 19 olanzapine, 3 quetiapine, and 3 risperidone. In
a study of the association between the atypical antipsychotic agents and
diabetes mellitus in patients with schizophrenia or related psychoses,
Melkersson et al. (1999) found elevated fasting insulin levels and
reduced growth hormonedependent insulin-like growth factor I in
28 patients taking conventional antipsychotic agents compared with
13 patients taking clozapine, thus suggesting insulin resistance with
secondary increased hyperinsulinemia related to clozapine therapy.
Our group conducted a 5-year naturalistic study examining, in 96 patients with schizophrenia treated with clozapine, the incidence of treatment-emergent diabetes in relation to other factors such as weight gain,
lipid level abnormalities, age, clozapine dose, and concomitant treat-


Medical Illness and Schizophrenia

ment with valproic acid (Henderson et al. 2000). Patients with diabetes
prior to clozapine initiation experienced a twofold increase in insulin requirements or required a switch to insulin from an oral hypoglycemic
agent. At baseline, the patients mean weight was 79.6 kg and mean body
mass index was 26.9 kg/m2. Their mean age at the time of clozapine initiation was 36.4 years. The sample was 26.8% female and 91% Caucasian.
During the 5-year follow-up, 30 patients (36%) were diagnosed with diabetes. Patients experienced significant weight gain, which continued
until approximately month 46 from the beginning of clozapine treatment, but weight gain was not a significant risk factor for the development of diabetes in this study. After the conclusion of the 5-year study,
this cohort was followed for an additional 5 years, resulting in a 10-year
naturalistic study that examined the metabolic and cardiovascular side
effects of clozapine treatment. At the end of the 10-year study, Henderson et al. (2005c) reported a mean weight gain of approximately 13.6 kg
and an estimated 10-year cardiovascular disease mortality rate of 9%.
Over the same period, 33 (34%) of the 96 patients developed diabetes, to
generate a 43% estimated 10-year incidence rate of new-onset diabetes.
In another study, we examined data for patients with schizophrenia
treated at our hospital during a 7-year period and found that 18.4%
of patients with schizophrenia were diagnosed with diabetes mellitus
compared to 6.6% in the general hospital population (P< 0.001) (Henderson et al. 2007). Twenty-three patients with schizophrenia were
identified with diabetic ketoacidosis; of these, 11 had new-onset diabetes presenting as diabetic ketoacidosis, 8 had diabetic ketoacidosis
with known diabetes mellitus, 2 had new-onset diabetes mellitus
hyperosmolar hyperglycemic syndrome, and 2 had hyperosmolar
hyperglycemic syndrome with known diabetes mellitus. The incidence
of diabetes presenting as diabetic ketoacidosis in patients with schizophrenia was more than 10-fold higher than that reported in the general
population: 14.93 per 10,000 patient years in patients with schizophrenia versus 1.4 per 10,000 patient years in the general population
(P< 0.000001) and versus 1.98 per 10,000 patient years in the general
hospital population (P<0.000001). The incidence of diabetic ketoacidosis in patients taking each of the atypical antipsychotic drugs over the
7-year period was as follows: clozapine, 2.2%; olanzapine, 0.8%; and
risperidone, 0.2% (no incidence with ziprasidone or quetiapine). Of the
11 patients with diabetes presenting as diabetic ketoacidosis, the mean
hemoglobin A1c level at admission was 13.3% 1.9% (10.4%16.9%).
The elevated hemoglobin A1c levels observed suggest that patients had
undiagnosed diabetes mellitus for at least several weeks before their diabetic ketoacidosis episode.

Glucose Intolerance and Diabetes


A few reports have linked the atypical antipsychotic agents risperidone and quetiapine to new-onset diabetes mellitus and glucose intolerance, but not nearly as many as those described for clozapine and
olanzapine (Henderson 2001; Henderson et al. 2005a; Perez-Iglesias et
al. 2007). Griffiths and Springuel (2001), using the Canadian Adverse
Drug Reaction Monitoring Program, reported 37 cases of suspected glucose metabolism disorders associated with atypical antipsychotic
agents. With clozapine there were 8 cases of diabetes mellitus, 5 cases
of diabetic ketoacidosis, and 4 cases of hyperglycemia. For olanzapine
2 cases of diabetes mellitus were reported, along with 3 cases of diabetic
ketoacidosis, 2 cases of diabetic coma, and 3 cases of hyperglycemia.
For quetiapine, 1 case of diabetes mellitus and 2 cases of diabetic ketoacidosis were reported. Finally, for risperidone 1 case of diabetes mellitus,
1 case of labile blood glucose levels, 3 cases of hyperglycemia, and
2 cases of hypoglycemia in patients with a previous history of diabetes
mellitus were reported. Of note, there were 3 deaths from the 10 cases
of diabetic ketoacidosis.
Wirshing et al. (2001) reported two cases of new-onset diabetes in risperidone-treated patients. Both subjects were African American, were
overweight or obese prior to risperidone initiation, and gained significant weight while receiving treatment with risperidone. These cases
highlight the importance of understanding risk factors for diabetes in
patients receiving treatment with atypical antipsychotic agents.
Finally, Fryburg et al. (2001) reported a double-blind controlled trial
comparing olanzapine and ziprasidone therapy on the basis of metabolic indices. In a comparison of baseline to 6-week follow-up, olanzapine subjects showed a significant increase in body weight, fasting
serum insulin, cholesterol, triglycerides and a parameter for insulin resistance, suggesting increased insulin resistance in the olanzapine
group but not the ziprasidone group. Similarly, Perez-Iglesias et al.
(2007) conducted a 12-week study in which 128 patients taking either
haloperidol, olanzapine, or risperidone were assessed. Although all
treatment groups showed a worsening lipid profile, increases in triglyceride levels were observed only in the olanzapine group.
All of the reports indicate that the strength of the association between the atypical antipsychotic medications and disturbances in glucose regulation can vary across the different medications, with many
more cases of medication-induced hyperglycemia, diabetes mellitus,
and diabetic ketoacidosis occurring with clozapine and olanzapine
treatment than with treatment with the atypical agents risperidone,
quetiapine, and ziprasidone (Haupt and Newcomer 2001). Also, it is
important to recognize that the reports of diabetic ketoacidosis associ-


Medical Illness and Schizophrenia

ated with clozapine and olanzapine treatments and the increased incidence of new-onset type 2 diabetes may represent two distinct
populations with varying risks for developing the diseases (Henderson
2002; Henderson et al. 2007).

Causative Mechanisms With

Atypical Antipsychotic Agents
The atypical antipsychotic medications could lead to hyperglycemia
and diabetes mellitus in a number of ways. As discussed earlier in this
chapter, decreased sensitivity (increased resistance) to insulin and decreased insulin secretion as a result of decreased beta cell function are
involved in the development of type 2 diabetes. A few controlled studies suggest that the atypical antipsychotic medications affect insulin resistance rather than causing a primary defect in insulin secretion
(Henderson 2007; Henderson et al. 2005a). The insulin resistance seen
during atypical antipsychotic treatment may be a result of increased
central adiposity or may arise from the direct effect of the medications
action on the glucose transporter function (Haupt and Newcomer
2001). Dwyer et al. (1999) studied the effects of atypical antipsychotic
agents on glucose transporter function. They suggested that a structurefunction relationship exists in which similar drugs, such as clozapine
and olanzapine, achieve relatively higher intracellular concentrations
and bind to, and thus interfere with, the function of the glucose transporter proteins.
Another mechanism by which the atypical antipsychotic agents may
lead to hyperglycemia and diabetes mellitus is by antagonism of the serotonin 5-HT1A receptors. Antagonism of these receptors may decrease
pancreatic beta cell responsiveness to blood sugar levels, thus resulting
in disturbances in glucose metabolism secondary to decreased insulin
secretion (Gilles et al. 2005). Melkersson et al. (2000), however, found
that patients treated with olanzapine had higher fasting insulin levels
on comparison of baseline to 5-month follow-up, suggesting that impaired beta cell function is an unlikely cause for olanzapine-associated
diabetes mellitus. Their group found that although 11 of 14 (79%) olanzapine-treated patients were normoglycemic and only three showed increased blood glucose values, the majority of patients (10 of 14, or 71%)
had insulin levels above the normal limit. The report also noted
increased rates of hyperleptinemia, hypertriglyceridemia, and hypercholesterolemia. In short, Melkersson et al. found that olanzapine treatment was associated with weight gain and elevated levels of insulin,
leptin, and blood levels as well as insulin resistance, with three patients

Glucose Intolerance and Diabetes


diagnosed with diabetes mellitus. The elevated insulin values argue

against the theory that antagonism of the serotonin receptors by some
atypical antipsychotic agents, a property that could theoretically lead to
decreased beta cell insulin production, causes hyperglycemia and diabetes mellitus. It is also worth noting that olanzapine has little affinity
for 5-HT1A receptors.
Notably, in several of the case studies and research studies mentioned in this chapter, weight gain was not associated with the development of diabetes mellitus or diabetic ketoacidosis during atypical
antipsychotic therapy. Indeed, many of the individuals who developed
diabetes mellitus or diabetic ketoacidosis were of normal weight. Thus,
although all obese individuals are at increased risk for the development
of type 2 diabetes, the added obesity caused by the atypical antipsychotic medications does not appear to be the sole reason for the development of diabetes mellitus or diabetic ketoacidosis in patients with
schizophrenia. Nevertheless, if patients with schizophrenia gain considerable amounts of weight and maintain that weight, their risk of developing type 2 diabetes is significantly increased.

The potential benefit of improving health behavior was demonstrated in
a longitudinal study of 85,000 nurses. Results indicated that 83% of coronary heart disease events were avoided by 3% of individuals who maintained a desirable weight, ate a healthy diet, exercised regularly, drank
alcoholic beverages moderately, and did not smoke (Stampfer et al. 2000).
In nurses who achieved three of the above components, 51% of coronary
events were avoided. Although achieving a healthy lifestyle may be more
difficult for patients with schizophrenia, the potential benefits warrant
intensive efforts toward this goal on the part of patients and caregivers.
Switching to a more weight-neutral atypical antipsychotic agent offers promise in halting or reversing weight gain and improving glucose
metabolism, but many patients and their clinicians are reluctant to risk
a worsening or return of psychotic symptoms and risk relapse (Casey et
al. 2003; Luebbe et al. 2006; Spurling et al. 2007; Weiden et al. 2003).
Menza et al. (2004) reported on a 52-week multimodal weight control
program using nutrition, exercise, and behavioral interventions in
31 patients with schizophrenia or schizoaffective disorder and found,
using a last observation carried forward (LOCF) analysis, a significant
reduction in weight, body mass index (BMI), hemoglobin A1c, and systolic and diastolic blood pressure, with 20 of 31 completing the program.


Medical Illness and Schizophrenia

A number of pharmacological interventions have been evaluated to

improve glucose metabolism. One study randomized 26 first-episode
schizophrenia patients to either olanzapine 10 mg or olanzapine 10 mg
plus reboxetine 4 mg (Poyurovsky et al. 2003). Patients receiving the
combination had less weight gain (2.5 2.7 kg) than patients receiving
olanzapine alone (5.5 3.1 kg). In another study, fluoxetine was ineffective in diminishing olanzapine-induced weight gain in first-episode
schizophrenia patients (Poyurovsky et al. 2002). In a 16-week doubleblind trial, Radulovic et al. (2002) evaluated the addition of sibutramine
with behavioral weight counseling to an ongoing antipsychotic regimen
in stable, overweight, or obese outpatients with schizophrenia. Twentyone patients were assigned to either sibutramine (up to 15 mg/day) or
placebo in a 2:1 ratio. Nineteen completed at least 4 weeks of doubleblind treatment (sibutramine: 14; placebo: 5), with 11 completing the
full 16 weeks. There were no significant differences between groups on
weight loss (placebo group 4.2 kg; sibutramine group 3.7 kg) or BMI
(1.46 kg/m 2 in the placebo group; 1.32 kg/m 2 in the sibutramine
group) in an LOCF analysis. The small sample size and high dropout
rate may have limited the opportunity to detect differences between
groups. In a placebo-controlled trial of sibutramine with a behavioral
nutrition program in olanzapine-treated subjects (Henderson et al.
2005b), the mean weight loss was 3.8 1.1 kg in the sibutramine group
and 0.8 0.7 kg in the placebo group (P< 0.05). However, no improvements in lipids were detected in this study. Another study examined
the effect of amantadine for weight loss in olanzapine-treated patients
(Deberdt et al. 2005). An LOCF analysis indicated a modest but significant difference in change in weight comparing amantadine to placebo
group at weeks 8, 12, and 16. At week 16 the result was 0.194.58 kg
for amantadine versus 1.28 4.26 kg for placebo (P= 0.045).
In a study of 24 patients with treatment-resistant schizophrenia
who began taking aripiprazole with decreased doses of clozapine,
Karunakaren et al. (2006) reported an average weight loss of 5.1 kg over
34 weeks. Mitsonis et al. (2007) reported on 27 patients for whom aripiprazole was added to stable doses of clozapine. Although no significant weight loss was noted, patients total scores on the Positive and
Negative Syndrome Scale for Schizophrenia improved significantly. In
a 6-week open-label trial to examine the effects of adjunctive aripiprazole in 10 clozapine-treated subjects, patients decreased in weight (from
99.5 18.1 kg to 96.9 17.4 kg) (P=0.003) from baseline to study endpoint (Henderson et al. 2006b). Significant decreases were also seen in
fasting total serum cholesterol (from 211 27 mg/dL to 18427 mg/dL)
(P=0.002) and triglycerides (from 274229 to 176106 mg/dL) (P=0.04).

Glucose Intolerance and Diabetes


Morrison et al. (2002) openly treated 19 adolescents taking olanzapine, risperidone, quetiapine, or valproate, with metformin 500 mg
three times a day. The mean weight loss at 12 weeks was 2.93 3.13 kg,
with 15 of 19 patients losing some weight. In a 12-week study comparing the addition of fluvoxamine to low-dose clozapine (250 mg/day)
with high-dose clozapine monotherapy (600 mg/day), Lu et al. (2004)
found that subjects on monotherapy had significantly greater increases
in weight, BMI, serum glucose, and triglyceride levels. Wu et al. (2008)
randomized 40 first-episode schizophrenia patients to treatment with
either olanzapine 15 mg/day plus metformin 750 mg/day or olanzapine plus placebo for 12 weeks. They found that weight, BMI, waist circumference, insulin, and insulin resistance increased less with the
combination. Baptista et al. (2008) reported on a 12-week study with
metformin (8501,700 mg) plus sibutramine (1020 mg, n = 13) or placebo (n =15) in olanzapine-treated patients with chronic schizophrenia.
Weight loss was similar in both groups, although the combination did
prevent a triglyceride increase.

Monitoring and Screening

The development of glucose intolerance and diabetes mellitus may be a
very serious comorbid complication of treatment with antipsychotic
agents, thus contributing to morbidity and mortality in patients with
schizophrenia. The development of hyperglycemia in patients with
schizophrenia often goes unrecognized, as it does in approximately
50% of the general population. Because patients taking antipsychotic
medications may be at increased risk for developing diabetes, close
monitoring and screening for obesity and hyperglycemia are imperative for individualizing treatment decisions and reducing the risks of
morbidity and mortality. Before starting treatment with an atypical
antipsychotic, the clinician should perform a risk factor assessment for
diabetes mellitus and other metabolic disorders. This risk assessment
should include baseline serum lipid and glucose values (preferably
fasting), weight, BMI, age, ethnicity, family history of diabetes in a firstdegree relative, level of physical activity, and diet (American Diabetes
Association et al. 2004).
The American Diabetes Association (2008) recommends that routine
screening for diabetes begin at age 45 in whites with no other risk
factors for diabetes, and further recommends repeat testing for this
group every 3 years thereafter if the results are within normal range.


Medical Illness and Schizophrenia

The association recommends beginning surveillance at an earlier age

and more frequent monitoring for individuals with any of the following
risk factors for diabetes and cardiovascular disease: obesity, a firstdegree relative with diabetes, membership in a high-risk ethnic population (African Americans; Hispanic Americans; Asian Americans, including Indian Asians; Pacific Islanders; and Native Americans),
previous diagnosis of gestational diabetes or delivery of a baby larger
than 4.4 kg, hypertension (>130/85 mmHg), a high-density lipoprotein
level 40 mg/dL in men and 50 mg/dL in women, a triglyceride level
150 mg/dL, or previous diagnosis of impaired fasting glucose (fasting
plasma glucose level 100 mg/dL but <126 mg/dL) or impaired glucose tolerance (oral glucose tolerance test revealing 2-hour postload
glucose level of 140 mg/dL but <200 mg/dL).
Because schizophrenia is associated with increased rates of diabetes
mellitus or impaired glucose regulation, and because the newer, atypical antipsychotic agents appear to increase this risk even further, it
seems reasonable that patients with schizophrenia, and especially patients taking atypical antipsychotic medications, should be screened for
hyperglycemia before age 45 and monitored more frequently than every 3 years. This is especially appropriate in light of case reports of newonset diabetes mellitus associated with atypical antipsychotic therapy
in patients under age 20. Additionally, patients with higher risk factors
should be examined more closely. African American, Asian, and Latino
patients should be monitored closely, particularly if they have other
risk factors for diabetes independent of ethnicity. If factors such as
weight gain, elevated lipids, or hypertension develop, early interventions are necessary to prevent the onset of diabetes. Although very little
is known about the underlying reasons that some patients treated with
atypical antipsychotic agents experience diabetic ketoacidosis, undertaking a risk assessment with each patient may help in choosing safer
agents for each individual. We recommend monitoring weight changes
and blood pressure at each office visit, as well as fasting glucose and lipids at baseline and every 3 months. Also, patients must be instructed to
notify the treatment team if they experience any physical symptoms or
medical disorders, because a number of diabetic ketoacidosis episodes
have occurred in conjunction with other acute medical disorders such
as infections and pancreatitis. A recent study found that hemoglobin
A1c levels were significantly elevated in patients with schizophrenia
with new-onset diabetes mellitus presenting as diabetic ketoacidosis
(Henderson et al. 2007). Therefore, more frequent monitoring of hemoglobin A1c levels may aid in forewarning physicians of a patients risk
for developing diabetes mellitus or diabetic ketoacidosis. Finally, pa-

Glucose Intolerance and Diabetes


tients treated with atypical antipsychotic agents, particularly clozapine

and olanzapine, should be screened more frequently.
In contrast to use of some of the atypical antipsychotic agents, treatment with ziprasidone and aripiprazole are associated with minimal
impact on metabolic parameters. Ziprasidone may offer significant advantages because it appears to be weight neutral (Allison et al. 1999;
Gardner et al. 2005; Henderson 2007); however, additional clinical experience is necessary with this agent before definitive conclusions can
be reached. Aripiprazole also appears to be associated with minimal
weight gain and little or no negative impact on serum lipids or glucose
metabolism (Melkersson and Dahl 2004).
Therefore, a reasonable approach would be to take baseline measures (fasting blood glucose, lipids, weight, waist measurements, blood
pressure) for all patients and to repeat fasting glucose and lipids every
6 months if possible. Weight and blood pressure can be monitored more
frequently in the office. Monitoring of glycohemoglobin may be useful
in this population if fasting glucose is difficult to obtain. However, glycohemoglobin may not be sensitive enough to detect lower levels of hyperglycemia, glycohemoglobin changes will lag 23 months behind
changes in glucose homeostasis, and criteria for diagnosing diabetes
with glycohemoglobin do not exist. Education and referral to weight reduction and exercise programs may play a significant preventive role.
In addition, weight gain should initiate meaningful dietary intervention. Finally, the clinician should be alert to the possibility of diabetic
If a patient develops diabetes while receiving treatment with an
atypical antipsychotic agent, consideration should be given to switching to another antipsychotic agent. Although diabetic ketoacidosis appears to be uncommon, it is of great concern because it increases the risk
of death. Patients who experience an episode of diabetic ketoacidosis
should be switched to a different antipsychotic agent that has a lower
impact on glucose metabolism. Reports and clinical experience suggest
that in a case of atypical antipsychotic agentassociated diabetes or diabetic ketoacidosis, discontinuation of the antipsychotic agent may result in a complete resolution of the hyperglycemia and diabetes.
Clozapine-treated patients have few options. Clozapine remains the
most effective agent for the treatment-resistant schizophrenia population, and patients often have failed to benefit from several other antipsychotic agents. Interventions to reduce the factors that contribute to
impaired glucose tolerance, such as weight loss programs, encouraging
exercise, and lowering the clozapine dose, may lead to improvement.
Other approaches, including augmentation with a more neutral anti-


Medical Illness and Schizophrenia

psychotic agent such as aripiprazole, may result in improvements in

patients weight and metabolic parameters (Henderson et al. 2006b).
For patients treated with other atypical agents, switching to another
agent should be considered in addition to the preceding interventions.
Finally, diabetic patients who are placed on these agents must be monitored more closely to prevent a worsening of glycemic control.
The widely distributed guidelines published by the American Diabetes Association et al. (2004) recommend monitoring of all patients who
are taking atypical antipsychotic agents by performing a baseline assessment, including fasting glucose and lipids, waist and weight measurements, and blood pressure. They recommend repeating the
measures 3 months later and then annually. Regardless of whether
these specific guidelines are followed, individual clinicians and programs should develop and adhere to a monitoring system to improve
the safety of patients taking these medications.

Atypical antipsychotic medications have helped to improve the lives of
patients with schizophrenia by alleviating positive and negative symptoms and bringing some improvement in cognitive function. This improvement is helping patients to function in the home and in the
workplace, thus improving the patients quality of life. At the same
time, however, the atypical antipsychotics appear to be associated with
the development of glucose intolerance, new-onset diabetes mellitus,
diabetic ketoacidosis, and exacerbation of existing diabetes mellitus.
These disturbances in glucose metabolism have their own medical consequences, including cardiovascular disease, cerebrovascular disease,
diabetic retinopathy, neuropathy, and diabetic nephropathy, all of
which can lead to considerable morbidity and mortality. Thus, to minimize morbidity and mortality associated with the use of atypical antipsychotic medications, close screening and monitoring for diabetes
mellitus should become a priority for all clinicians treating patients
with schizophrenia who are receiving atypical antipsychotic therapy.

Key Clinical Points

Many clinicians choose to prescribe atypical antipsychotic medications,

such as clozapine, olanzapine, quetiapine, risperidone, ziprasidone, and
aripiprazole, instead of the older agents, because the newer agents
have fewer extrapyramidal symptoms and may increase adherence and

Glucose Intolerance and Diabetes


prevent relapse. However, these medications may also increase the risk
of metabolic disorders, such as impaired glucose metabolism, diabetes
mellitus, and diabetic ketoacidosis.

The choice of antipsychotic medications should be based on perceived

efficacy and knowledge of short- and long-term side effects. A generally
reasonable approach is to choose the safest medications available first
and, if the patient has no response, work the way up the risk ladder. For
metabolic problems, prevention is preferable over intervention.

Diabetes mellitus and glucose intolerance are characterized by problems

in glucose-insulin regulation. The symptoms that warrant a diabetes
diagnosis include polyuria, polydipsia, unexplained weight loss, a casual
plasma glucose level 200 mg/dL, a fasting plasma glucose 126 mg/dL,
or a 2-hour plasma glucose level 200 mg/dL post 75-g oral glucose
load. Glucose intolerance is characterized by a 2-hour plasma glucose
level 140 mg/dL and <200 mg/dL post 75-g oral glucose load.

Glucose metabolism abnormalities, including onset of diabetes mellitus,

diabetic ketoacidosis, cerebrovascular disease, cardiovascular disease,
myocardial infarction and stroke, diabetic retinopathy, and neuropathy, are associated with considerable medical morbidity and mortality.

Atypical antipsychotic agents have been associated with significant weight

gain, particularly the development of central obesity, which can lead to
several comorbid cardiovascular conditions as well as type 2 diabetes.

The incidence of disturbances in glucose regulation varies across antipsychotics; many more cases of medication-induced hyperglycemia, diabetes mellitus, or diabetic ketoacidosis occur with clozapine and olanzapine
than with risperidone, quetiapine, ziprasidone, and aripiprazole.

Mechanisms by which atypical antipsychotic agents could lead to glucose abnormalities and diabetes mellitus include causing insulin resistance by either increasing central or visceral adiposity or affecting the
glucose transporter function antagonism of serotonin 5-HT1A, which
may decrease pancreatic beta cell responsiveness to blood sugar levels
and treatment-induced weight gain.

Close and frequent monitoring and screening for obesity and hyperglycemia is imperative for individualizing treatment decisions and reducing the risks for morbidity and mortality for patients with
schizophrenia taking antipsychotic mediations.

Lifestyle interventions and recommendations should be given at the

start of the medications and reinforced at every visit.


Medical Illness and Schizophrenia

Allison DB, Mentore JL, Heo M, et al: Antipsychotic-induced weight gain: a
comprehensive research synthesis. Am J Psychiatry 156:16861696, 1999
Allison DB, Mackell JA, McDonnell DD: The impact of weight gain on quality
of life among persons with schizophrenia. Psychiatr Serv 54:565567, 2003
American Diabetes Association: Diagnosis and classification of diabetes mellitus. Diabetes Care 31 (suppl 1):S55S60, 2008
American Diabetes Association, American Psychiatric Association, American
Association of Clinical Endocrinologists, et al: Consensus Development
Conference on Antipsychotic Drugs and Obesity and Diabetes. J Clin Psychiatry 65:267272, 2004
Baptista T, Uzcategui E, Rangel N, et al: Metformin plus sibutramine for olanzapine-associated weight gain and metabolic dysfunction in schizophrenia: a 12-week double-blind, placebo-controlled pilot study. Psychiatry Res
159:250253, 2008
Bobes J, Arango C, Aranda P, et al: Cardiovascular and metabolic risk in outpatients with schizophrenia treated with antipsychotics: results of the CLAMORS study. Schizophr Res 90:162173, 2007
Braceland FJ, Meduna LJ, Vaichulis JA: Delayed action of insulin in schizophrenia. Am J Psychiatry 102:108110, 1945
Brown S, Birtwistle J, Roe L, et al: The unhealthy lifestyle of people with schizophrenia. Psychol Med 29:697701, 1999
Carolino ID, Molena-Fernandes CA, Tasca RS, et al: Risk factors in patients with
type 2 diabetes mellitus. Rev Lat Am Enfermagem 16:238244, 2008
Casey DE, Carson WH, Saha AR, et al: Switching patients to aripiprazole from
other antipsychotic agents: a multicenter randomized study. Psychopharmacology (Berl) 166:391399, 2003
Casey DE, Haupt DW, Newcomer JW, et al: Antipsychotic-induced weight gain
and metabolic abnormalities: implications for increased mortality in patients with schizophrenia. J Clin Psychiatry 65 (suppl 7):418, 2004
Christensen O, Christensen E: Fat consumption and schizophrenia. Acta Psychiatr Scand 78:587591, 1988
Coccurello R, Caprioli A, Ghirardi O, et al: Chronic administration of olanzapine induces metabolic and food intake alterations: a mouse model of the
atypical antipsychotic-associated adverse effects. Psychopharmacology
(Berl) 186:561571, 2006
Cohn TA, Remington G, Zipursky RB, et al: Insulin resistance and adiponectin
levels in drug-free patients with schizophrenia: a preliminary report. Can J
Psychiatry 51:382386, 2006
Csernansky JG, Mahmoud R, Brenner R, et al: A comparison of risperidone and
haloperidol for the prevention of relapse in patients with schizophrenia.
N Engl J Med 346:1622, 2002
Deberdt W, Winokur A, Cavazzoni PA, et al: Amantadine for weight gain associated with olanzapine treatment. Eur Neuropsychopharmacol 15:1321, 2005

Glucose Intolerance and Diabetes


Dixon L, Weiden P, Delahanty J, et al: Prevalence and correlates of diabetes in

national schizophrenia samples. Schizophr Bull 26:903912, 2000
Dwyer DS, Pinkofsky HB, Liu Y, et al: Antipsychotic drugs affect glucose uptake and the expression of glucose transporters in PC12 cells. Prog Neuropsychopharmacol Biol Psychiatry 23:6980, 1999
Elias AN, Hofflich H: Abnormalities in glucose metabolism in patients with
schizophrenia treated with atypical antipsychotic medications. Am J Med
121:98104, 2008
Fell MJ, Anjum N, Dickinson K, et al: The distinct effects of subchronic antipsychotic drug treatment on macronutrient selection, body weight, adiposity,
and metabolism in female rats. Psychopharmacology (Berl) 194:221231, 2007
Fryburg DA, OSullivan RL, Siu C, et al: Insulin resistance in olanzapine and
ziprasidone treated subjects: interim results of a double blind controlled sixweek trial (NR261), in 2001 New Research Program and Abstracts, American
Psychiatric Association 154th Annual Meeting, New Orleans, LA, May 510,
2001. Washington, DC, American Psychiatric Association, 2001
Gardner DM, Baldessarini RJ, Waraich P: Modern antipsychotic drugs: a critical
overview. CMAJ 172:17031711, 2005
Gerstein HC, Pais P, Pogue J, et al: Relationship of glucose and insulin levels to
the risk of myocardial infarction: a case-control study. J Am Coll Cardiol
33:612619, 1999
Gilles M, Wilke A, Kopf D, et al: Antagonism of the serotonin (5-HT)-2 receptor
and insulin sensitivity: implications for atypical antipsychotics. Psychosom Med 67:748751, 2005
Goff DC, Sullivan LM, McEvoy JP, et al: A comparison of ten-year cardiac risk
estimates in schizophrenia patients from the CATIE study and matched
controls. Schizophr Res 80:4553, 2005
Griffiths J, Springuel P: Atypical antipsychotics: impaired glucose metabolism.
CMAJ 165:943945, 947949, 2001
Hgg S, Joelsson L, Mjrndal T, et al: Prevalence of diabetes and impaired glucose tolerance in patients treated with clozapine compared with patients
treated with conventional depot neuroleptic medications. J Clin Psychiatry
59:294299, 1998
Haupt DW, Newcomer JW: Hyperglycemia and antipsychotic medications.
J Clin Psychiatry 62 (suppl 27):1526, 2001
Henderson DC: Clinical experience with insulin resistance, diabetic ketoacidosis, and type 2 diabetes mellitus in patients treated with atypical antipsychotic agents. J Clin Psychiatry 62 (suppl 27):1014, 2001
Henderson DC: Atypical antipsychotic-induced diabetes mellitus: how strong
is the evidence? CNS Drugs 16:7789, 2002
Henderson DC: Weight gain with atypical antipsychotics: evidence and insights. J Clin Psychiatry 68 (suppl 12):1826, 2007
Henderson DC, Cagliero E, Gray C, et al: Clozapine, diabetes mellitus, weight
gain, and lipid abnormalities: a five-year naturalistic study. Am J Psychiatry 157:975981, 2000


Medical Illness and Schizophrenia

Henderson DC, Daley TB, Kunkel L, et al: Clozapine and hypertension: a chart
review of 82 patients. J Clin Psychiatry 65:686689, 2004
Henderson DC, Cagliero E, Copeland PM, et al: Glucose metabolism in patients
with schizophrenia treated with atypical antipsychotic agents: a frequently
sampled intravenous glucose tolerance test and minimal model analysis.
Arch Gen Psychiatry 62:1928, 2005a
Henderson DC, Copeland PM, Daley TB, et al: A double-blind, placebocontrolled trial of sibutramine for olanzapine-associated weight gain. Am
J Psychiatry 162:954962, 2005b
Henderson DC, Nguyen DD, Copeland PM, et al: Clozapine, diabetes mellitus,
hyperlipidemia, and cardiovascular risks and mortality: results of a 10year naturalistic study. J Clin Psychiatry 66:11161121, 2005c
Henderson DC, Borba CP, Daley TB, et al: Dietary intake profile of patients with
schizophrenia. Ann Clin Psychiatry 18:99105, 2006a
Henderson DC, Kunkel L, Nguyen DD, et al: An exploratory open-label trial of
aripiprazole as an adjuvant to clozapine therapy in chronic schizophrenia.
Acta Psychiatr Scand 113:142147, 2006b
Henderson DC, Cagliero E, Copeland PM, et al: Elevated hemoglobin A1c as a
possible indicator of diabetes mellitus and diabetic ketoacidosis in schizophrenia patients receiving atypical antipsychotics. J Clin Psychiatry
68:533541, 2007
Jin H, Meyer JM, Jeste DV: Phenomenology of and risk factors for new-onset diabetes mellitus and diabetic ketoacidosis associated with atypical antipsychotics: an analysis of 45 published cases. Ann Clin Psychiatry 14:5964,
Karunakaran K, Tungaraza TE, Harborne GC: Is clozapine-aripiprazole combination a useful regime in the management of treatment-resistant schizophrenia? J Psychopharmacol 21:453456, 2006
Koller E, Doraiswamy PM: Olanzapine-associated diabetes mellitus. Pharmacotherapy 22:841852, 2002
Koller E, Schneider B, Bennett K, et al: Clozapine-associated diabetes. Am J Med
111:716723, 2001
Lebovitz HE: Diagnosis, classification, and pathogenesis of diabetes mellitus.
J Clin Psychiatry 62 (suppl 27):59, 2001
Lewis SW, Barnes TR, Davies L, et al: Randomized controlled trial of effect of
prescription of clozapine versus other second-generation antipsychotic
drugs in resistant schizophrenia. Schizophr Bull 32:715723, 2006
Lieberman JA, Stroup TS, McEvoy JP, et al: Effectiveness of antipsychotic drugs
in patients with chronic schizophrenia. N Engl J Med 353:12091223, 2005
Lu ML, Lane HY, Lin SK et al: Adjunctive fluvoxamine inhibits clozapine-related
weight gain and metabolic disturbances. J Clin Psychiatry 65:766771, 2004
Luebbe G, Kropp S, Harms G, et al: Extensive weight loss in a patient with
schizophrenia after switching from olanzapine to aripiprazole: a case report. Pharmacopsychiatry 39:76, 2006
Marinow A: Diabetes in chronic schizophrenia. Dis Nerv Syst 32:777778, 1971

Glucose Intolerance and Diabetes


McCreadie R, Macdonald E, Blacklock C, et al: Dietary intake of schizophrenic

patients in Nithsdale, Scotland: case-control study. BMJ 317:784785, 1998
McEvoy JP, Meyer JM, Goff DC, et al: Prevalence of the metabolic syndrome in
patients with schizophrenia: baseline results from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) schizophrenia trial and
comparison with national estimates from NHANES III. Schizophr Res
80:1932, 2005
Melkersson K, Dahl ML: Adverse metabolic effects associated with atypical antipsychotics: literature review and clinical implications. Drugs 64:701723,
Melkersson KI, Hulting AL, Brismar KE: Different influences of classical antipsychotics and clozapine on glucose-insulin homeostasis in patients with
schizophrenia or related psychoses. J Clin Psychiatry 60:783791, 1999
Melkersson KI, Hulting AL, Brismar KE: Elevated levels of insulin, leptin, and
blood lipids in olanzapine-treated patients with schizophrenia or related
psychoses. J Clin Psychiatry 61:742749, 2000
Meltzer HY: Putting metabolic side effects into perspective: risks versus benefits of atypical antipsychotics. J Clin Psychiatry 62 (suppl 27):3539, 2001
Menza M, Vreeland B, Minsky S, et al: Managing atypical antipsychotic-associated weight gain: 12-month data on a multimodal weight control program.
J Clin Psychiatry 65:471477, 2004
Mitsonis CI, Dimopoulos NP, Kararizou EG, et al: Aripiprazole augmentation
in the management of residual symptoms in clozapine-treated outpatients
with chronic schizophrenia: an open-label pilot study. Prog Neuropsychopharmacol Biol Psychiatry 31:373377, 2007
Morrison JA, Cottingham EM, Barton BA: Metformin for weight loss in pediatric patients taking psychotropic drugs. Am J Psychiatry 159:655657,
Mukherjee S, Decina P, Bocola, V, et al: Diabetes mellitus in schizophrenic patients. Compr Psychiatry 37:6873, 1996
Nasrallah HA: Metabolic findings from the CATIE trial and their relation to tolerability. CNS Spectr 11 (suppl 7):3239, 2006
Newcomer JW: Metabolic disturbances associated with antipsychotic use. J Clin
Psychiatry 62 (suppl 27):34, 2001
Newcomer JW: Metabolic risk during antipsychotic treatment. Clin Ther
26:19361946, 2004
Newcomer JW: Metabolic considerations in the use of antipsychotic medications: a review of recent evidence. J Clin Psychiatry 68 (suppl 1):2027,
Newcomer JW: Metabolic syndrome and mental illness. Am J Manag Care 13
(7 suppl):S170S177, 2007b
Newcomer JW, Haupt DW: The metabolic effects of antipsychotic medications.
Can J Psychiatry 51:480491, 2006
Osby U, Correia N, Brandt L, et al: Mortality and causes of death in schizophrenia in Stockholm county, Sweden. Schizophr Res 45:2128, 2000


Medical Illness and Schizophrenia

Peet M: Nutrition and schizophrenia: an epidemiological and clinical perspective. Nutr Health 17:211219, 2003
Peet M: Diet, diabetes and schizophrenia: review and hypothesis. Br J Psychiatry 184(suppl):S102S105, 2004a
Peet M: International variations in the outcome of schizophrenia and the prevalence of depression in relation to national dietary practices: an ecological
analysis. Br J Psychiatry 184:404408, 2004b
Pereira MA, FitzerGerald SJ, Gregg EW, et al: A collection of Physical Activity
Questionnaires for health-related research. Med Sci Sports Exerc 29 (6 suppl):
S1S205, 1997
Perez-Iglesias R, Crespo-Facorro B, Amado JA, et al: A 12-week randomized
clinical trial to evaluate metabolic changes in drug-naive, first-episode psychosis patients treated with haloperidol, olanzapine, or risperidone. J Clin
Psychiatry 68:17331740, 2007
Pi-Sunyer FX: Medical hazards of obesity. Ann Intern Med 119:655660, 1993
Pi-Sunyer FX: The medical risks of obesity. Obes Surg 12 (suppl 1):6S11S,
Poyurovsky M, Pashinian A, Gil-Ad I, et al: Olanzapine-induced weight gain
in patients with first-episode schizophrenia: a double-blind, placebocontrolled study of fluoxetine addition. Am J Psychiatry 159:10581060,
Poyurovsky M, Isaacs I, Fuchs C, et al: Attenuation of olanzapine-induced
weight gain with reboxetine in patients with schizophrenia: a doubleblind, placebo-controlled study. Am J Psychiatry 160:297302, 2003
Radulovic L, Weiden P, Allison DB: A double-blind, placebo controlled trial of
adjunctive sibutramine for obesity in patients with schizophrenia, in Proceedings of the 2002 Meeting of the New Clinical Drug Evaluation Unit.
Rockville, MD, New Clinical Drug Evaluation Unit, 2002, pp 115
Ryan MC, Collins P, Thakore JH: Impaired fasting glucose tolerance in firstepisode, drug-naive patients with schizophrenia. Am J Psychiatry 160:284
289, 2003
Saddichha S, Manjunatha N, Ameen S, et al: Diabetes and schizophreniaeffect
of disease or drug? Results from a randomized, double-blind, controlled
prospective study in first-episode schizophrenia. Acta Psychiatr Scand
117:342347, 2008
Schernthaner G: Kidney disease in diabetology: lessons from 2007. Nephrol
Dial Transplant 23:11121115, 2008
Spelman LM, Walsh PI, Sharifi N, et al: Impaired glucose tolerance in first
episode drug-naive patients with schizophrenia. Diabet Med 24:481485,
Spurling RD, Lamberti JS, Olsen D, et al: Changes in metabolic parameters with
switching to aripiprazole from another second-generation antipsychotic:
a retrospective chart review. J Clin Psychiatry 68:406409, 2007
Stampfer MJ, Hu FB, Manson JE, et al: Primary prevention of coronary heart
disease in women through diet and lifestyle. N Engl J Med 343:1622, 2000

Glucose Intolerance and Diabetes


Stokes C, Peet M: Dietary sugar and polyunsaturated fatty acid consumption as

predictors of severity of schizophrenia symptoms. Nutr Neurosci 7:247
249, 2004
Strassnig M, Brar JS, Ganguli R: Nutritional assessment of patients with schizophrenia: a preliminary study. Schizophr Bull 29:393397, 2003
Stroup TS, Lieberman JA, McEvoy JP, et al: Effectiveness of olanzapine, quetiapine, risperidone, and ziprasidone in patients with chronic schizophrenia following discontinuation of a previous atypical antipsychotic. Am J
Psychiatry 163:611622, 2006
van Nimwegen LJM, Storosum JG, Blumer RME, et al: Hepatic insulin resistance in antipsychotic naive patients with schizophrenia, a detailed study
of glucose metabolism with stable isotopes. J Clin Endocrinol Metab
93:572577, 2008
Venkatasubramanian G, Chittiprol S, Neelakantachar N, et al: Insulin and insulin-like growth factor-1 abnormalities in antipsychotic-naive schizophrenia. Am J Psychiatry 164:15571560, 2007
Waitzkin L: A survey of unknown diabetics in a mental hospital, I: men under
age fifty. Diabetes 15:97104, 1966
Wannamethee SG, Shaper AG, Lennon L, et al: Metabolic syndrome vs
Framingham Risk Score for prediction of coronary heart disease, stroke,
and type 2 diabetes mellitus. Arch Intern Med 165:26442650, 2005
Weiden PJ, Daniel DG, Simpson G, et al: Improvement in indices of health status in outpatients with schizophrenia switched to ziprasidone. J Clin Psychopharmacol 23:595600, 2003
Westphal SA: The occurrence of diabetic ketoacidosis in non-insulin-dependent
diabetes and newly diagnosed diabetic adults. Am J Med 101:1924, 1996
Wirshing DA, Pierre JM, Eyeler J, et al: Risperidone-associated new-onset diabetes. Biol Psychiatry 50:148149, 2001
Wu RR, Zhao JP, Jin H, et al: Metformin addition attenuates olanzapineinduced weight gain in drug-naive first-episode schizophrenia patients:
a double-blind, placebo-controlled study. Am J Psychiatry 165:352358,

This page intentionally left blank

Effects of
Antipsychotics on
Serum Lipids
Jonathan M. Meyer, M.D.

Patients with severe mental illnesses, such as schizophrenia or bipolar disorder, are a medically vulnerable population at high risk for
cardiovascular mortality (Newcomer and Hennekens 2007), with standardized mortality ratios from cardiovascular disease that are two
times greater than in the general population (Osby et al. 2000, 2001).
Although much of psychiatric care for individuals with schizophrenia
is focused on suicide prevention, cardiovascular disease remains the
single largest cause of death among males and females with schizophrenia.
Given this sobering data, it is imperative that those who care for patients with severe mental illness have a working knowledge of the risks
associated with cardiovascular disease and the patterns of risk factors
seen in this patient population. Recognition and treatment of diabetes
mellitus is covered in Chapter 5, Glucose Intolerance and Diabetes in
Patients With Schizophrenia, but the importance of diabetes relates
not only to the adverse effects of hyperglycemia but also to its impact
on cardiovascular risk. Diabetes mellitus is considered equivalent to
having established coronary heart disease (CHD) for future risk of major cardiovascular events (e.g., myocardial infarction, sudden death)


Medical Illness and Schizophrenia

(Expert Panel on Detection, Evaluation and Treatment of High Blood

Cholesterol in Adults 2001). This view of diabetes-related CHD risk is
based on indications that patients with established diabetes have the
same future myocardial infarction incidence as nondiabetics who have
already experienced a myocardial infarction (Haffner et al. 1998).
For nondiabetics, the focus for preventing cardiovascular disease remains on modification of the traditional risk factors of hypertension,
hyperlipidemia, and smoking. Baseline data from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE), sponsored by the
National Institute of Mental Health, provide the most timely and complete picture of the risk patterns in patients with chronic schizophrenia
residing in the United States (Goff et al. 2005). Among the 1,460 schizophrenia patients assessed at study entry, 10-year CHD risk, calculated
using Framingham scores, was significantly elevated compared with
age-, gender-, and race- and ethnicity-matched controls from a general
population database in both males (9.4% vs. 7.0%) and females (6.3% vs.
4.2%) (P= 0.0001). In particular, schizophrenia patients compared with
controls had significantly higher rates of smoking (68% vs. 35%), diabetes (13% vs. 3%), and hypertension (27% vs. 17%) and lower highdensity lipoprotein (HDL) cholesterol levels (43.7 vs. 49.3 mg/dL)
(P< 0.001). Moreover, CATIE subjects also had greater prevalence of
central adiposity and elevated serum triglycerides, both of which are
components of the metabolic syndrome and are associated with insulin
resistance and future diabetes risk (McEvoy et al. 2005). The importance
of monitoring serum triglyceride values during antipsychotic treatment
will become readily apparent, because this is the lipid parameter most
greatly affected by offending medications. Another concerning finding
from CATIE (covered in Chapter 3, Medical Outcomes From the
CATIE Schizophrenia Study) was the significant undertreatment of
hypertension and dyslipidemia, reinforcing the idea that clinicians
must redouble efforts to both monitor and treat the basic cardiovascular
risks in patients with schizophrenia (Nasrallah et al. 2006).
The induction of hyperlipidemia during antipsychotic therapy thus
represents a serious condition, not only because of its inherent impact
on cardiovascular risk but also because of its occurrence in a group that
already possesses considerable risk (Saari et al. 2004). What has become
evident in recent years is that the atypical antipsychotics have a decreased liability for neurological side effects, but certain agents in this
class have a marked propensity for adverse metabolic outcomes, especially hyperlipidemia (Meyer and Koro 2004). One is tempted to hypothesize that the widespread use of atypical antipsychotics with
metabolic liabilities may be responsible for the widening mortality gap

Effects of Antipsychotics on Serum Lipids


between schizophrenia patients and the general population (Saha et al.

2007), but the fact remains that more must be done to prevent morbidity
and early mortality from CHD. This chapter is intended to guide the clinician in selecting medications and appropriate monitoring strategies
for hyperlipidemia based on the best available data. This discussion is
bolstered by the publication of the double-blind, controlled data from
phases 1 and 2 of the CATIE schizophrenia trial (Daumit et al. 2008; Lieberman et al. 2005; Meyer et al. 2008a, 2008b; Stroup et al. 2006). The
large population under study in CATIE is one of the best sources of prospective information regarding certain compounds, especially perphenazine and quetiapine, for which prospective data were previously
sorely lacking.
Due to increased interest in improving medical outcomes for patients with severe mental illness (Marder et al. 2004), minimization of
iatrogenically induced lipid problems and appropriate monitoring of
those at risk are increasingly becoming the standards of care for this patient population. Recognizing which antipsychotics impose the greatest
risk for hyperlipidemia and understanding the common dyslipidemia
patterns seen during use of these antipsychotics are necessary for providing high-quality care to antipsychotic-treated patients.

Hyperlipidemia and
Coronary Heart Disease
A complete review of lipid physiology is beyond the scope of this chapter; however, some information is worth noting herein. The major
classes of lipoproteins are divided on the basis of their weight under
centrifugation (Kwiterovich 2000). The lightest are chylomicrons derived from dietary triglycerides and very-low-density lipoproteins
(VLDL) that are endogenously produced triglyceride-rich particles. The
next heavier particles are low-density lipoproteins (LDL) and highdensity lipoproteins (HDL). Total cholesterol, as obtained on a lipid
panel, reflects the combination of serum concentrations of VLDL (calculated as serum triglyceride levels divided by 5), LDL, and HDL. Normal
serum total cholesterol concentration is defined as a fasting level under
200 mg/dL, normal HDL is 40 mg/dL, and the ideal serum LDL concentration is determined by cardiovascular risk factors as noted below.
The rate-limiting step in cholesterol synthesis is a reaction catalyzed by
the enzyme 3-hydroxy-3-methylglutaryl coenzyme A reductase, the
target for a class of lipid-lowering agents called statins. The Multiple
Risk Factor Intervention Trial clearly established the linear association


Medical Illness and Schizophrenia

between CHD and serum cholesterol, but further research clarified that
risk is more properly associated with abnormalities involving the LDL
and HDL cholesterol fractions and separately by serum triglycerides
(Stamler et al. 1986).
A series of landmark studies published in the mid-1990s demonstrated the significant association between reduction in LDL levels and
reduced cardiovascular events and mortality. The Scandinavian Simvastatin Survival Study was a randomized, double-blind, placebocontrolled trial of cholesterol lowering in 4,444 patients with coronary
heart disease who were already following a lipid-lowering diet and
were followed for a median of 5.4 years (Randomised Trial of Cholesterol Lowering 1994). This secondary prevention study found that statin therapy reduced total cholesterol by 28% and LDL by 38%, and led
to a 42% reduction in coronary deaths and a 30% decrease in all-cause
mortality compared to the placebo cohort. A randomized, double-blind,
placebo-controlled secondary prevention trial of pravastatin in 9,014
patients (ages 3175) with ischemic heart disease followed for 6.1 years
on average showed a 24% reduction in nonfatal myocardial infarction
or death due to CHD (Prevention of Cardiovascular Events and
Death 1998). A similar 24% reduction in nonfatal myocardial infarction or death due to CHD was seen in a randomized, double-blind,
placebo-controlled secondary prevention trial of pravastatin after myocardial infarction in 4,159 patients who had normal cholesterol levels
(Sacks et al. 1996).
Moreover, two primary prevention studies of patients without CHD
but with either hypercholesterolemia or low HDL levels and average
total cholesterol identified profound benefits from statin therapy.
The West of Scotland Coronary Prevention Study was a randomized,
double-blind, placebo-controlled trial of pravastatin in 6,595 men, ages
4564 years, with mean total cholesterol of 272 mg/dL and no history
of myocardial infarction, followed for a mean of 4.9 years (Shepherd et
al. 1995). A 32% reduction in the development of CHD and a 22% reduction in total mortality were found in this trial (Shepherd et al. 1995). The
Air Force/Texas Coronary Atherosclerosis Prevention Study was a primary prevention trial of lovastatin versus placebo in men and women
with no clinical evidence of cardiovascular disease and average total
cholesterol and LDL levels, but low HDL (Downs et al. 1998). Over the
5.2 years of this study, a 37% reduction in risk of major coronary events
was seen in the lovastatin group compared to the placebo group.
To establish the 2001 National Cholesterol Education Program
guidelines, the Expert Panel on Detection, Evaluation and Treatment
of High Blood Cholesterol in Adults (2001) used the results of these

Effects of Antipsychotics on Serum Lipids


and other studies to establish LDL goals based on the number of underlying major cardiovascular risk factors: smoking, hypertension (having
blood pressure 140/90 mmHg or taking antihypertensive medication),
low HDL cholesterol (<40 mg/dL), family history of premature CHD
(CHD in male first-degree relative <55 years or female first-degree
relative <65 years), and age (men 45 years or women 55 years). The
fasting LDL goal is <160 mg/dL for those with zero or one risk factor,
<130 mg/dL for those with two or more risk factors, and <100 mg/dL
for those with CHD or CHD-equivalent disorders (e.g., diabetes mellitus). Recent updates to the 2001 guidelines suggest even lower LDL
goals (<70 mg/dL) are appropriate for those who are at very high risk
by virtue of having diabetes mellitus, multiple metabolic syndrome risk
factors, and persistent risk factors such as smoking in conjunction with
established CHD or acute coronary syndrome (Stone et al. 2005).
Although LDL continues to be the main focus of lipid-lowering therapy due to evidence from multiple large studies, other lipid parameters
must be addressed and are secondary targets for therapy. Significant
data from the Framingham Heart Study show a 2%4% increase in risk
of CHD-related death for each 1-mg/dL decrease in HDL, with over
30% of deaths occurring in those with normal total cholesterol values
(Wilson et al. 1988). Because low HDL is one of the metabolic syndrome
criteria, the 2005 National Cholesterol Education Program update suggests targeting low HDL in those who have already met LDL goals but
continue to manifest the characteristic dyslipidemia associated with
metabolic syndrome (i.e., low HDL, elevated fasting triglycerides)
(Stone et al. 2005). Elevated fasting triglycerides occur as a direct result
of insulin resistance, because insulin-dependent lipases in fat cells are
normally inhibited by insulin. As insulin resistance worsens, inappropriately high levels of lipolysis lead to the release of excess amounts of
free fatty acids that are hepatically transformed into triglycerides
(Smith 2007). Elevated fasting triglyceride levels thus become a sensitive marker of insulin resistance, and therefore are included as a
metabolic syndrome criterion. Moreover, a fasting triglyceride:HDL
cholesterol ratio 3.0 performs better than fasting glucose in predicting
insulin resistance among prediabetic individuals (McLaughlin et al.
The sensitivity of the triglyceride:HDL ratio to insulin resistance derives from the fact that increased triglyceride levels interfere with an important regulatory function governing the production of apolipoprotein
B100, a core lipoprotein in very-low-density, intermediate-density, and
low-density lipoprotein particles (Smith 2007). The overproduction of
apolipoprotein B100 results in more of these triglyceride-rich particles,


Medical Illness and Schizophrenia

with resultant hypertriglyceridemia. In addition, the greater presence of

these light triglyceride-rich lipoproteins causes the transfer of triglycerides to HDL at the expense of HDL cholesterol content. After passage
through the liver, where triglyceride is cleaved by enzymatic processes,
the remaining HDL particle is smaller than normal and more readily
cleared in the kidney, resulting in the characteristic low serum HDL levels seen with insulin-resistant states (Smith 2007). The triglyceride:HDL
ratio thus reflects the combined effects of low HDL and elevated triglycerides seen in insulin-resistant patients.
Although serum triglyceride concentrations are considered a secondary focus of lipid-lowering treatment, evidence from several studies link elevated triglycerides (i.e., 200 mg/dL) to increased CHD risk
(Cullen 2000; Jeppesen et al. 1998; Rubins 2000), an increased risk that
is independent of HDL levels (Jeppesen et al. 1998). Evidence also has
been reported that supports monitoring of nonfasting triglycerides, because atherosclerosis may be a postprandial phenomenon in which
atherogenic remnant lipoproteins (chylomicrons and VLDL) play a critical role (Eberly et al. 2003). These triglyceride-rich particles are smaller
than other lipid components and more readily penetrate arterial intimal
cells. Individuals are in a nonfasting state most of the day with respect
to serum triglycerides, with triglyceride levels peaking 4 hours after an
oral fat load and returning to baseline values only after 8 hours (Nordestgaard et al. 2007).
Large population studies indicate a significant linear correlation between nonfasting triglyceride values and directly measured remnant lipoproteins (Nordestgaard et al. 2007). Data from a prospective
Copenhagen, Denmark, study (N =13,981, mean follow-up 26 years) solidified the link between nonfasting triglycerides and cardiovascular
events by identifying the significant relationship between nonfasting
triglyceride levels in men and women and risk of major cardiovascularrelated events, including CHD, myocardial infarction, and mortality
(Nordestgaard et al. 2007). Compared with individuals with nonfasting
triglyceride levels <88.5 mg/dL, women and men with levels of 177.0
264.6 mg/dL had adjusted hazard ratios for myocardial infarction of 2.5
and 1.6, respectively. The superiority of nonfasting triglycerides over
fasting triglycerides is also seen in prospective data from the Womens
Health Study (N = 26,509, median follow-up 11.4 years) (Bansal et
al. 2007). Although no relationship was found between increasing tertiles of fasting triglyceride values and risk of cardiovascular events in
fully adjusted models, nonfasting triglyceride tertiles were significantly
associated with cardiovascular risk, with triglyceride levels measured
24 hours postprandially showing the strongest association. The impor-

Effects of Antipsychotics on Serum Lipids


tance of fasting and nonfasting hypertriglyceridemia is of further concern for those who care for schizophrenia patients, given the association
between elevated triglycerides and therapy with low-potency phenothiazines and the dibenzodiazepine-derived compounds clozapine,
olanzapine, and quetiapine (a dibenzothiazepine).

Hyperlipidemia and
Typical Antipsychotics
Although typical antipsychotics are used in less than 10% of patients
with schizophrenia in the United States, they are commonly available
throughout the world and represent an important class of psychotropic
medications. Moreover, a review of the lipid effects of typical antipsychotics illustrates an important concept seen in the atypical antipsychotic data: medications with similar modes of therapeutic action can
have disparate metabolic profiles. Within a decade after the widespread
use of chlorpromazine and other low-potency phenothiazines, several
studies emerged examining the metabolic profiles of this class of antipsychotics (Clark and Johnson 1960; Clark et al. 1967; Mefferd et al.
1958). In general, these compounds were found to elevate serum triglycerides and total cholesterol but with greater effects on triglyceride
concentrations. Subsequent studies of the phenothiazine chlorpromazine and related compounds (Clark et al. 1970, 1972) confirmed these
findings that high serum triglycerides seemed to be the primary significant lipid abnormality, but elevated total cholesterol could also be
found. What also emerged from this early literature was the fact that
certain D2 antagonists did not exert the same changes on serum lipids
as seen with the lower-potency phenothiazines.
The lipid neutrality of high-potency typical antipsychotics was seen
in the early uncontrolled studies of butyrophenone derivatives published in the mid-1960s (Braun and Paulonis 1967; Clark et al. 1968;
Simpson and Cooper 1966; Simpson et al. 1967) and a placebo-controlled
trial in 1971 (Serafetinides et al. 1971). Comparative trials of low-potency
phenothiazines and butyrophenones (primarily haloperidol) published
in the 1970s confirmed the lipid neutrality of high-potency agents,
whereas low-potency phenothiazine treatment was associated with
hyperlipidemia, primarily in the form of hypertriglyceridemia (Serafetinides et al. 1972; Vaisanen et al. 1979).
Surprisingly, only a limited amount of subsequent work has been
published in the past two decades covering lipid changes during typical antipsychotic therapy. Table 61, which is a summary of studies,


Medical Illness and Schizophrenia

case reports, and case series related to lipids and antipsychotics that
have been published since 1980, demonstrates the paucity of controlled
data on typical antipsychotics after 1980. Two papers were published
by a Japanese group in 1984 and 1985 that analyzed serum total cholesterol and triglycerides in male chronic schizophrenia inpatients taking
phenothiazines or butyrophenones compared with age- and sexmatched controls (Sasaki et al. 1984, 1985). The cohort exposed to phenothiazines had mean serum triglycerides of 163 mg/dL, compared
with 104 mg/dL for the butyrophenone group and 127 mg/dL for the
control group, with no significant differences in total cholesterol across
the three groups. The phenothiazine cohort also had higher serum LDL
cholesterol and decreased HDL cholesterol concentrations compared
with the other study arms. Cross-sectional studies from Pakistan
(Shafique et al. 1988) and a psychiatric hospital in Spain (Martnez et al.
1994) substantiated earlier findings, although specific analysis on the
basis of type of neuroleptic prescribed was not performed in the Spanish study.
Lastly, the CATIE schizophrenia trial employed a medium-potency
phenothiazine, perphenazine, as one of its treatment arms in phase 1,
thus providing controlled comparative data versus atypical antipsychotics in a randomized, double-blind study (Lieberman et al. 2005). As
shown in Table 62, use of perphenazine was associated with modest
increases in serum triglycerides and total cholesterol, with a greater effect on triglycerides. The extent of lipid changes related to perphenazine use in neuroleptic-naive patients might actually be greater than
seen in CATIE phase 1, because 22% of the CATIE sample at baseline
was taking olanzapine. Given olanzapines known effects on serum lipids, olanzapine-exposed patients switched to perphenazine in phase 1
would likely have minimal further increases in serum triglycerides.
Further analysis of CATIE data at the end of phase 1 (Table 63), obtained after approximately 9 months of exposure, revealed that perphenazine exposure was associated with favorable effects on serum
HDL in whites (+2.7 mg/dL) in a manner that was significantly different from olanzapine (1.7 mg/dL); however, perphenazine induced a
mean decrease of 1.3 mg/dL in nonwhites, and this was significantly
different from that seen for ziprasidone (+4.3 mg/dL) (Meyer et al.
2008a). Whether this represents a specific moderating effect of race and
ethnicity on perphenazines metabolic effects or merely the results of
an unusual cohort remains to be seen. The effect of perphenazine
on fasting triglycerides depended on the subjects baseline triglyceride
valuesthose with low fasting triglycerides experienced a mean increase of 28.7 mg/dL, comparable to most of the other agents studied,

Effects of Antipsychotics on Serum Lipids


whereas those with high baseline triglycerides (>148 mg/dL) experienced a mean decrease of 27.5 mg/dL. With the increasing use of atypical antipsychotics throughout the world, these may be the last
randomized, controlled data published on the lipid effects of a typical

Hyperlipidemia and
Atypical Antipsychotics
Atypical antipsychotics have been designed primarily to mimic the
identifiable features of clozapines pharmacology: weaker D 2 antagonism than typical antipsychotics, combined with potent serotonin
5-HT2A antagonism (Meyer and Simpson 1997). A recurring theme, related to the metabolic effects of this antipsychotic class, echoes the earlier findings for typical agents, namely that drugs with similar
mechanisms of action may have disparate metabolic adverse effects.
This analogy is more than superficial, because the structurally related
dibenzodiazepine-derived atypical antipsychoticsolanzapine, clozapine, and quetiapineall appear to have significantly greater effects on
serum triglycerides than total cholesterol, much like the low-potency
phenothiazines, whereas other atypical antipsychoticsziprasidone,
risperidone, and aripiprazolehave minimal lipid effects, as did the
high-potency typical antipsychotics.

Through 2002 virtually no randomized, prospective, controlled data
had been published on the lipid effects of the atypical antipsychotics,
but the early literature is replete with case reports and retrospective
studies, primarily focused on clozapine, the only atypical antipsychotic
available until 1994. The first reports of hyperlipidemia with atypical
antipsychotics were small studies of fluperlapine, a dibenzodiazepinederived compound never marketed. Two trial reports published in the
mid-1980s documented elevated triglycerides, in one case as high as 900
mg/dL (Fleischhacker et al. 1986; Mller-Oerlinghausen 1984). Despite
clozapines use from the 1960s onward, the first case report of hyperlipidemia (presenting as hypertriglyceridemia) occurred in 1994
(Vampini et al. 1994) and was followed in 1995 by a report of four clozapine-treated patients with hypertriglyceridemia whose triglycerides
normalized upon switching to risperidone (Ghaeli and Dufresne 1995).
The following year (1996), the authors of that case series published


Medical Illness and Schizophrenia

a chart review comparing serum lipids in patients exposed to clozapine

or typical antipsychotics (primarily butyrophenones) for at least 1 year
with no prior history of hyperlipidemia or use of lipid-lowering agents
(Ghaeli and Dufresne 1996). This retrospective study found serum triglycerides 114 mg/dL higher in the clozapine cohort (P < 0.001) but
nearly identical total cholesterol levels (clozapine 217.0 52.9 mg/dL
vs. typical antipsychotics 215.0 43.2 mg/dL). A 1998 study comparing
30 patients taking clozapine to 30 patients taking typical antipsychotics
for at least 1 year also found higher triglycerides in the clozapine group
(202.9 131.1 mg/dL) compared with the typical group (134.4 51.9
mg/dL), but again without significant differences in total cholesterol
(clozapine 197.146.4 mg/dL vs. typical antipsychotics 194.951.5 mg/dL)
(Spivak et al. 1998). A small prospective study (Dursun et al. 1999) and
two subsequent chart reviews of long-term clozapine-treated patients,
one with 70 patients (Spivak et al. 1999) and the other with 222 patients
(Gaulin et al. 1999), demonstrated that the mean increase in serum triglycerides ranged from 41% to 45%.
More recent cross-sectional and short- and long-term prospective
clozapine studies have noted the same pattern of marked elevations of
serum triglycerides with modest increases in total cholesterol (Atmaca
et al. 2003b; Baymiller et al. 2003; Henderson et al. 2000; Leonard et al.
2002; Wirshing et al. 2002) compared with baseline and with typical
antipsychotics (Lund et al. 2001). A 1-year prospective study of 50 clozapine-treated patients (Baymiller et al. 2003) found mean increases of
41.7% in serum triglycerides, exactly in the range predicted by prior retrospective data, and only a 7.5% increase in total cholesterol. The authors also noted that no significant changes occurred in HDL and LDL
and that triglyceride elevations peaked between days 41 and 120 and
then declined, but still remained elevated at the 1-year interval. As part
of a 5-year naturalistic study of 82 clozapine-treated patients with
schizophrenia, Henderson et al. (2000) found ongoing increases in serum triglycerides (linear coefficient 2.75 mg/dL per month, P=0.04). A
subsequent 10-year follow-up study of clozapine-treated patients
found that triglycerides had plateaued significantly, with a linear coefficient over that time frame of 0.5 mg/dL/month (P=0.04) (Henderson
et al. 2005). In neither study were total cholesterol changes significant.
Other cross-sectional, retrospective and other nonprospective clozapine studies are noted in Table 61, but the efficacy arm of CATIE
phase 2 (hereafter referred to as phase 2b) generated prospective randomized data on a sample of 49 schizophrenia patients compared with
50 subjects assigned to other agents (McEvoy et al. 2006). As seen in
Table 62, the exposure-adjusted mean increases in serum lipids for

Effects of Antipsychotics on Serum Lipids


clozapine-treated subjects were greatest for triglycerides and substantially less for total cholesterol. The samples for each of the other phase
2b drug arms are extremely small (<15), so the findings for the other
atypical antipsychotics in phase 2b are of limited reliability.

The first published studies of olanzapine-associated hypertriglyceridemia appeared in 1999 and revealed patterns of dyslipidemia similar
to those reported for clozapine. A study of nine patients followed for an
average of 16 months while taking olanzapine (mean age 41 years,
mean olanzapine dose 19 mg/day) showed a mean increase in serum
triglycerides of 41%, with no significant changes in total cholesterol levels and a mean weight gain of 10 kg (Sheitman et al. 1999). A subsequent
report on 25 inpatients followed prospectively for 12 weeks found a
mean fasting triglyceride increase of 37%. Both weight and triglyceride
increases were significant (P <0.05), and a significant association was
found between weight gain and triglyceride change (P<0.02); however,
after controlling for baseline weight, analysis of covariance showed no
independent increase in triglycerides (Osser et al. 1999).
Subsequent work involving olanzapine illustrates not only its much
greater effect on serum triglycerides than on total cholesterol, but also
the risk of severe hypertriglyceridemia. Melkersson et al. (2000) followed
a group of 14 patients with schizophrenia taking olanzapine for an average of 5 months and noted that 62% had elevated fasting triglycerides
(mean 273.45 mg/dL) and 85% exhibited hypercholesterolemia (mean
257.14 mg/dL). Although these lipid changes were comparable to those
reported for clozapine, the issue of severe hypertriglyceridemia was
noted in Meyers (2001) case series of 12 olanzapine and two quetiapine
patients with fasting triglyceride levels exceeding 500 mg/dL, including
one patient taking olanzapine whose fasting serum triglyceride levels
were measured at 7,688 mg/dL, and a subsequent case of olanzapineassociated severe hypertriglyceridemia (triglycerides 5,093 mg/dL) reported by Stoner et al. (2002) in a patient who also developed new-onset
type 2 diabetes mellitus. As discussed later in the section Monitoring
Recommendations for Hyperlipidemia During Antipsychotic Therapy,
one concern with serum triglyceride levels of 1,000 mg/dL or greater is
the development of acute pancreatitis, and the majority of cases that are
associated with atypical antipsychotic treatment are related to olanzapine or clozapine exposure (Koller et al. 2003).
The cases of severe hypertriglyceridemia are extreme findings, but
the majority of studies published since 2001 have confirmed the greater


Medical Illness and Schizophrenia

effects of olanzapine on serum lipids, primarily on triglycerides, compared to more lipid-neutral medications such as high-potency typical
antipsychotics or risperidone, ziprasidone, and aripiprazole. The earlier comparative literature comprised retrospective, cross-sectional, or
small prospective studies, but confirmed the greater effects of olanzapine on serum lipids than other atypical antipsychotic medications
(with the exception of clozapine) (Atmaca et al. 2003a, 2003b; Bouchard
et al. 2001; Garyfallos et al. 2003; Kinon et al. 2001, 2004; Meyer 2002;
Wirshing et al. 2002). Also, two retrospective large database studies
were performed using data from a Great Britain General Practice Research Practice Database (Koro et al. 2002) and the state of California
MediCal claims system (Lambert et al. 2005). Koro et al.s (2002) casecontrol study of 18,309 schizophrenia patients, which included 1,268 incident cases of hyperlipidemia, compared typical antipsychotics, risperidone, and olanzapine to no antipsychotic usage, and found that
risperidone use was not associated with increased odds of hyperlipidemia compared to typical antipsychotic use or no antipsychotic exposure, whereas olanzapine use was associated with a nearly fivefold
increase in the odds of developing hyperlipidemia compared with no
antipsychotic exposure and more than a threefold increase compared
with typical antipsychotic use. Lambert et al.s (2005) case-control study
of MediCal claims after schizophrenia diagnosis, in patients with antipsychotic monotherapy within 12 weeks prior to hyperlipidemia claim,
found greater risk of hyperlipidemia for olanzapine compared to typical antipsychotics or risperidone. Increasing the exposure window to 24
or 52 weeks did not affect the results.
A cross-sectional study by Almras et al. (2004) is noteworthy in that
other lipid parameters were studied in addition to triglycerides and total cholesterol. In Quebec, Canada, the researchers examined only male
schizophrenia patients treated with olanzapine (n =42) or risperidone
(n = 45) for >6 months and compared the results of a comprehensive
lipid panel with those from a reference group of nondiabetic males
(mean ages: 28.4 risperidone; 31.7 olanzapine; 32.8 controls). The olanzapine-treated cohort had significantly higher serum cholesterol, LDL,
triglycerides, cholesterol:HDL ratio, and apolipoprotein B, and significantly lower HDL, smaller LDL peak particle size, and lower apolipoprotein A1. Interestingly, compared to the control group, the
olanzapine subjects had no significant differences on total cholesterol,
triglycerides, or LDL, but they did have lower HDL levels and a higher
cholesterol:HDL ratio. Additionally, the risperidone cohort had lower
total cholesterol and LDL levels but lower HDL levels than the reference group.

Effects of Antipsychotics on Serum Lipids


Given the modest size of the early prospective literature and the uncontrolled nature of many retrospective studies, questions were raised
about the extent to which olanzapine induces greater dyslipidemia than
risperidone. Phases 1 and 2a of the CATIE schizophrenia trial provided
an answer (Lieberman et al. 2005; Meyer et al. 2008a; Stroup et al. 2006),
as shown in Tables 62 and 63. This large, randomized study demonstrated convincingly that olanzapine is one of the greatest offending
agents with respect to its effects on serum lipids, again mostly on serum
triglycerides, whereas both risperidone and ziprasidone appear neutral. Phase 2a is likely more reflective of the true extent of olanzapines
effects, as the CATIE trial design mandated that those who entered that
arm of the trial not be exposed to an agent to which they had been randomized in phase 1. Thus, all of the subjects in the olanzapine arm of
phase 2a were new starts to that medication. This fact explains why the
triglyceride increase in phase 2a olanzapine subjects was 94.1 mg/dL,
compared with only a 40.5-mg/dL increase seen in the phase 1 olanzapine cohort, 22% of whom were taking olanzapine at study baseline.
The analysis of nonfasting triglycerides also revealed the greater deleterious impact of olanzapine compared with other agents, particularly
when nonswitchers were removed from the analysis (Meyer et al.
The differential impact of olanzapine compared with other agents
can also be seen in switch studies (Casey et al. 2003; Meyer et al. 2005;
Spurling et al. 2007; Su et al. 2005; Weiden et al. 2003, 2007). Institutionalized populations may experience less adverse metabolic effects, as
seen in a switch study of developmentally disabled adults (McKee et al.
2005) and two studies of olanzapine in elderly inpatient populations,
one including patients with schizophrenia or schizoaffective disorder
(Barak and Aizenberg 2003) and the other including patients with dementia of the Alzheimer type with behavioral disturbances or psychosis (De Deyn et al. 2004). For the latter two studies in particular, what is
not known is whether it is subject age or the fact of being confined with
a fixed dietary regimen that mitigates the development of dyslipidemia.

Quetiapine and zotepine are structurally related to the other dibenzodiazepine-derived antipsychotics clozapine and olanzapine, but limited information has been published regarding their metabolic effects.
The extent of available information was so sparse that the American Diabetes Association et al.s (2004) consensus paper on metabolic effects


Medical Illness and Schizophrenia

of atypical antipsychotics was unclear whether quetiapine could be differentiated from risperidone in its effects on serum lipids. Compared
with clozapine and olanzapine, quetiapine generally has a lower risk of
causing significant weight gain (Lieberman et al. 2005; Wetterling 2001).
Nonetheless, the available data suggest that quetiapine shares the propensity with other benzodiazepine-derived atypical antipsychotics to
elevate serum triglyceride levels, as illustrated by case reports by
Domon and Cargile (2002) and Meyer (2001), two 6-week prospective
comparative studies by Atmaca et al. (2003a, 2003b), and an 8-week
study by Shaw et al. (2001). As with other dibenzodiazepine-derived
atypical antipsychotics, there were lesser effects on total cholesterol
than on triglycerides (Shaw et al. 2001). The extent of the lipid effects
was best seen in the CATIE trials (Lieberman et al. 2005; Meyer et al.
2008a, 2008b; Stroup et al. 2006), especially phase 2a, in which quetiapines nearly 40-mg/dL elevation in serum triglycerides was second
only to that of olanzapine and more than that for risperidone or ziprasidone, despite weight gain only marginally greater than for risperidone.
Analysis of 3-month changes in nonfasting lipids during CATIE phase
1 also found that quetiapine was associated with a mean increase (+59.8
mg/dL) that was nearly identical to that of olanzapine (+61.5 mg/dL)
when nonswitchers were removed from the analysis (Meyer et al.
Zotepine is associated with weight gain similar to that experienced
with olanzapine and clozapine (Wetterling 2001), but the sum total of
the published literature on lipid changes during zotepine therapy is one
case report in which serum triglycerides peaked at 1,247 mg/dL and
normalized upon switch to a high-potency typical agent (Wetterling

Nonbenzodiazepine Agents: Risperidone, Ziprasidone,

and Aripiprazole
The limited effects of the nonbenzodiazepine agentsrisperidone,
ziprasidone, and aripiprazoleon serum lipids has been demonstrated
in large prospective trials, because the pharmaceutical industry routinely includes multiple lipid measures as part of study protocols. The
more benign effects of risperidone were largely known from comparative trials versus olanzapine, as confirmed by the results from CATIE
phases 1 and 2a (Lieberman et al. 2005; Meyer et al. 2008a, 2008b; Stroup
et al. 2006). Although doubts about the differential effects of quetiapine
and risperidone on serum lipids were expressed several years ago in the
consensus paper by the American Diabetes Association et al. (2004),

Effects of Antipsychotics on Serum Lipids


when quetiapine is prescribed in full therapeutic doses for schizophrenia, it appears to elevate serum triglycerides in a manner not typically
seen with risperidone.
Ziprasidone appears to have even less adverse effect on serum lipids
than risperidone, as evidenced by the statistically significant improvement in serum triglycerides and total cholesterol over 6 weeks when patients were switched to ziprasidone from risperidone (Kingsbury et al.
2001). The lipid neutrality of ziprasidone has been subsequently confirmed in a retrospective chart review (Brown and Estoup 2005) and in
multiple prospective trials of short duration (6 weeks) (Simpson et al.
2004; Weiden et al. 2003) and longer duration (26 weeks) (Breier et al.
2005; Cohen et al. 2004; Simpson et al. 2005; Weiden et al. 2007), including CATIE phases 1 and 2a (Lieberman et al. 2005; Meyer et al. 2008a,
2008b; Stroup et al. 2006) in which the net impact of ziprasidone treatment was to lower serum lipid levels.
Early clinical trials data suggested that aripiprazole also had nominal effects on serum lipids (Goodnick and Jerry 2002), a finding confirmed in large prospective trials. In a 26-week trial versus olanzapine
(McQuade et al. 2004), the mean change in fasting triglycerides was
+79.4 mg/dL with olanzapine but only +6.5 mg/dL with aripiprazole
(P< 0.05), whereas HDL decreased by 3.39 mg/dL with olanzapine and
increased 3.61 mg/dL with aripiprazole (P< 0.05). Changes in serum
LDL for the drug cohorts were not significantly different, but the incidence of new dyslipidemias was significantly greater for olanzapine on
the basis of the proportion of new subjects with endpoint serum
LDL >130 mg/dL (38% olanzapine vs. 19% aripiprazole, P< 0.05). Recent switch studies confirm that patients switched to aripiprazole from
metabolically more offending medications experience improvements in
serum lipids, again primarily triglycerides, but with measurable
changes for other parameters (Spurling et al. 2007).


Lipid changes during antipsychotic therapy: 19842008

Study design

MllerMulticenter fluperlapine trial in

43 schizophrenia and depression



In 16 of 28 patients, TG increased significantly and serum TC increased

Chart review of chronic phenothiaPatients chronically treated with phenothiazines (chlorpromazine,

zine users (mean exposure 8 years);
levomepromazine, perphenazine) had significantly higher HDL
10-week prospective data in 8 new (P <0.001) and higher TG (P <0.05) than normal controls. HDL
decreased 24% within 1 week of new phenothiazine exposure, with no
phenothiazine-treated patients
significant changes in TC and TG levels after 10 weeks.

Sasaki et al.

Chart review of males with chronic

schizophrenia; excluded patients
with DM or taking lipid-lowering
17 phenothiazine
14 haloperidol
14 healthy controls

After mean 8 years of antipsychotic exposure, no effect of

butyrophenones on lipids, but significantly elevated mean fasting TG
levels for the phenothiazine group (163 mg/dL) compared to the
butyrophenone group (104 mg/dL) and controls (127 mg/dL). No
significant differences in TC, LDL, or HDL values across the three

et al. 1986

Double-blind, prospective 6-week

6 haloperidol
6 fluperlapine

No significant differences in mean HDL or TC between groups or

compared to baseline by day 28. One fluperlapine subject developed
serum TG level of 900 mg/dL on day 7 and required treatment on
day 28.

Medical Illness and Schizophrenia

Sasaki et al.


Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Shafique et al.

Chart review of males with chronic

schizophrenia treated for 612
35 phenothiazine
30 butyrophenone
22 combined drug classes

Significant elevations of TC, VLDL, and LDL levels in patients taking

phenothiazines and LDL in patients taking butyrophenone. VLDL and
LDL levels significantly higher and HDL levels lower in patients taking
combined therapy.

Martnez et al.

Chart review of 311 chronically

Neuroleptic administration was associated with changes in HDL and TG
hospitalized schizophrenia patients
in males but not in females.
225 neuroleptic exposed (mostly
haloperidol, thioridazine, or
fluphenazine) for prior 2 years
86 no psychotropic medications

Vampini et al.

Case report:
1 taking clozapine 400 mg/day

Increased TG levels after 15 months.

Ghaeli and

Case series:
4 clozapine-treated patients
(doses 600900 mg/day)
switched to risperidone

Increased serum TG levels in patients taking clozapine decreased upon

switching to risperidone, and increased upon clozapine rechallenge.

Effects of Antipsychotics on Serum Lipids



Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Ghaeli and

Chart review of 67 schizophrenia

39 clozapine
21 high-potency typical
2 medium-potency typical
5 low-potency typical

Mean TG significantly higher in clozapine vs. typical groups

(264.6 mg/dL vs. 149.8 mg/dL, P<0.001). Difference was not explained
by concomitant illness or medication, age, or gender. No significant
difference in TC.

Spivak et al.

Chart review of schizophrenia

30 clozapine
(mean 295.0165 mg/day)
30 typical antipsychotics
(mean 348.9298.8 mg/day in
CPZ equivalents)

Mean TG significantly higher in clozapine vs. typical group after

1 year of treatment (202.9 mg/dL vs. 134.4 mg/dL, P<0.01). No
significant difference in serum TC.

Dursun et al.

Prospective 12-week study

8 clozapine
(mean 35273 mg/day)

Small increase in TG levels (11%) but no significant changes in other lipid

levels after 12 weeks.

Medical Illness and Schizophrenia





Lipid changes during antipsychotic therapy: 19842008 (continued)


Gaulin et al.

Chart review of schizophrenia

222 treated with clozapine or

After a mean treatment period of 590 days for clozapine and 455 days for
haloperidol, 45% increase in serum TG in clozapine-treated patients
(P<0.01) and insignificant decrease in serum TG in haloperidol-treated

Osser et al.

Prospective study in schizophrenia

25 olanzapine
(mean 13.84.4 mg/day)

37% (60 mg/dL) increase in fasting TG from baseline (P<0.05). Fasting

TC did not increase.

Sheitman et al.

Prospective study in schizophrenia

9 olanzapine (mean 19 mg/day)

After 16 months, mean increase in serum TG was 70 mg/dL

(5 patients had >50% increase in TG). No significant change in TC, HDL,
or LDL.

Spivak et al.

Chart review of schizophrenia

70 clozapine
(mean 332.9168.1 mg/day)
30 typical antipsychotics
(mean 347.3247.3 mg/day in
CPZ equivalents)

Mean TG increased in the clozapine group and decreased in the typical

antipsychotic group after 6 months of treatment (P<0.005).

et al. 2000

5-year naturalistic study

82 clozapine

Significant changes in serum TG (P =0.04)

(linear coefficient=2.75 mg/dL per month, SE=1.28).


Study design

Effects of Antipsychotics on Serum Lipids


Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


et al. 2000

Prospective cohort study

14 olanzapine
(median 12.5 mg/day)

62% prevalence of hypertriglyceridemia (mean 273.45 mg/dL) and 85%

prevalence of hypercholesterolemia (mean 257.14 mg/dL) after a
median exposure of 5 months (range 2.416.8 months).

Wu et al. 2000

Case report:
1 Chinese male, age 25, taking

Dose-dependent increases in fasting serum TG and glucose.

Baptista et al.

Cross-sectional cohort study of

women matched for age, BMI, and
day of menses
26 typical antipsychotics
(>6 consecutive months of
26 controls

Antipsychotic-treated subjects had a trend for lower HDL and increased

insulin resistance.

Bouchard et al.

Retrospective study
22 olanzapine
(mean 12.84.4 mg/day)
22 risperidone
(mean 2.81.8 mg/day)

After mean exposure of 17.9 months (olanzapine) and 17.4 months

(risperidone), olanzapine patients compared to risperidone patients
had significantly higher TG (185 mg/dL vs. 115 mg/dL, P<0.01),
significantly higher VLDL (0.9 mol/L vs. 0.5 mol/L, P<0.03), and a
trend for a higher cholesterol/HDL ratio (5.3 vs. 4.3, P=0.06) and lower
HDL (P=0.08). No significant differences in TC, fasting glucose, or

Medical Illness and Schizophrenia





Lipid changes during antipsychotic therapy: 19842008 (continued)


Domon and
Webber 2001

Case report:
After 3 months, weight 91 kg, fasting glucose 90 mg/dL, fasting TG
1 African American male, age 15,
155 mg/dL, fasting TC 131 mg/dL; 5 months later, weight peaked
taking olanzapine 20 mg/day
at 108 kg, then declined with onset of type 2 DM. Maximum TG
298 mg/dL with olanzapine. Both DM and hyperlipidemia resolved
over 8 weeks after olanzapine was discontinued.

et al. 2001

Prospective 6-week switch study

37 ziprasidone
(mean 124.3 mg/day)
Prior meds:
15 from olanzapine
12 from risperidone
10 from typical antipsychotics

Significant decreases in serum TC (17.57 mg/dL, P <0.001) and serum

TG (62.38 mg/dL, P=0.018). Change in TG correlated with weight
change (r=0.409, P=0.018, r2 =0.167). Change in TC did not correlate
with weight change.

Kinon et al.

Retrospective cohort
573 olanzapine (520 mg/day)
103 haloperidol (520 mg/day)

Median nonfasting endpoint serum TC was significantly higher for

olanzapine-treated than for haloperidol-treated patients
(205.7 mg/dL vs. 189.9 mg/dL, P=0.002).

Lund et al.

Retrospective cohort study

2,461 typical antipsychotics
552 clozapine

Clozapine significantly increased relative risk of hyperlipidemia in

patients ages 2034 years (RR 2.4; 95% CI 1.15.2) but not in older

Meyer 2001

Case series
12 olanzapine (520 mg/day)
2 quetiapine (200250 mg/day)

Hypertriglyceridemia reported up to 7,668 mg/dL with olanzapine and

1,932 mg/dL with quetiapine. Time to peak TG ranged from 1 to
23.5 months.


Study design

Effects of Antipsychotics on Serum Lipids


Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Nguyen and
Murphy 2001

Case report:
1 boy, age 10, taking olanzapine
5 mg/day for ADHD

Over 3 months, experienced 9-kg weight gain with resulting TC of

193 mg/dL and TG of 183 mg/dL. 5 weeks after olanzapine was
discontinued, TC 151 mg/dL, TG 61 mg/dL, and over 9-kg weight loss.

Shaw et al.

8-week open trial in adolescents with TC remained unchanged.

15 quetiapine (300800 mg/day)

Domon and
Cargile 2002

Case report:
1 African American female,
age 17, taking quetiapine
600 mg/day and metformin
1,000 mg po bid for type 2 DM

Goodnick and
Jerry 2002

Meta-analysis of trial data of 1,919

Meta-analysis of short-term trial data showed that increases in TC
patients treated with one of the
following aripiprazole administration were lower than for haloperidol,
following: aripiprazole, olanzapine, risperidone, or placebo. A 26-week trial comparing aripiprazole to
risperidone, haloperidol, or placebo olanzapine found significant differences after 4 weeks: olanzapine
increased TC, whereas aripiprazole decreased TC. Data from a 1-year
study found that aripiprazole produced less of an increase in TC than

On admission, TC 235 mg/dL, TG 456 mg/dL. Quetiapine discontinued

with complete resolution of DM and hyperlipidemia. At 6 weeks
postquetiapine (and 4 weeks without metformin), TC 226 mg/dL and
TG 163 mg/dL.

Medical Illness and Schizophrenia





Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Koro et al.

Case-control database study of 18,309 Olanzapine use was associated with nearly a fivefold increase in the odds
schizophrenia patients, with 1,268
of developing hyperlipidemia compared with no antipsychotic
incident cases of hyperlipidemia;
exposure and more than a threefold increase compared with those
typical or atypical antipsychotic use receiving typical antipsychotics.

et al. 2002

Chart review, 13 males, 8 females

21 clozapine
(mean 485 mg/day)

Martin and

Retrospective chart review of child

After mean exposure of 4.91.0 months, no significant changes in serum
and adolescent inpatients, mean age TG or TC levels were seen in the group as a whole.
12.8 years
22 risperidone
(mean 2.72.2 mg/day)

Meyer 2002

Retrospective chart review of 1-year

47 risperidone
47 olanzapine

11 patients (52%) had hypertriglyceridemia, 3 (14%) had

hypercholesterolemia. Mean TC 19346 mg/dL (range 131309 mg/dL),
mean TG 196142 mg/dL (range 39665 mg/dL).

After 52 weeks, TC increased 24 mg/dL with olanzapine vs. 7 mg/dL

with risperidone (P=0.029), and fasting TG increased by 88 mg/dL
with olanzapine vs. 30 mg/dL with risperidone (P=0.042). In the
nongeriatric cohort, TC increased 31 mg/dL with olanzapine vs.
7 mg/dL with risperidone (P =0.004), and fasting TG increased
105 mg/dL with olanzapine vs. 32 mg/dL with risperidone

Effects of Antipsychotics on Serum Lipids



Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Stoner et al.

Case report:
Baseline lipid panel: TC 227 mg/dL, TG 134 mg/dL, HDL 38 mg/dL.
1 African American male, age 42,
After 8 weeks of olanzapine serum: TG 5,093 mg/dL, TC 375 mg/dL,
with treated hyperlipidemia
with evidence of new-onset type 2 DM (fasting glucose 395 mg/dL,
taking olanzapine 15 mg/day
hemoglobin A1c 11.9%).

et al. 2002

Case report:
1 male, age 48, taking
olanzapine 1015 mg/day
for 1 month

6-kg weight gain; 3.6% decrease in basal and 11.4% decrease in 3-hour
energy expenditure; decrease in HDL; increase in TG and LDL. No
change in insulin sensitivity using euglycemic hyperinsulinemic clamp.


Case report:
1 patient taking zotepine

Maximum TG of 1,247 mg/dL, which normalized upon switch to highpotency typical antipsychotic.

Wirshing et al.

Chart review
39 clozapine
42 olanzapine
49 risperidone
13 quetiapine
41 haloperidol
41 fluphenazine

Clozapine was associated with greatest increase in TC, whereas

risperidone and fluphenazine were associated with decreases.
Clozapine and olanzapine were associated with greatest increase in TG

Medical Illness and Schizophrenia





Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Atmaca et al.

6-week prospective schizophrenia

14 quetiapine
(mean 535.7 mg/day)
14 olanzapine
(mean 15.7 mg/day)
14 risperidone
(mean 6.7 mg/day)
14 clozapine
(mean 207.1 mg/day)
11 controls
(no psychotropic medication)

Compared to controls, significant increases in fasting TG seen for

olanzapine (+31.23 mg/dL, P <0.001), clozapine (+36.28 mg/dL,
P<0.001), and quetiapine (+11.64 mg/dL, P<0.05), but not risperidone
(+3.87 mg/dL, P<0.76).

Atmaca et al.

6-week prospective schizophrenia

15 haloperidol
15 olanzapine
15 quetiapine

Serum TG increases were much greater for olanzapine compared to

quetiapine or haloperidol, and for quetiapine compared to haloperidol.

Effects of Antipsychotics on Serum Lipids



Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Barak and

Prospective 6-month study of

elderly inpatients (mean age 71.7)
with schizophrenia or schizoaffective disorder (16 female, 5 male)
21 olanzapine
(mean 12.9 mg/day)

After a mean duration of 289 days of olanzapine treatment, no significant

change from baseline serum lipid levels was found for TG or TC.

Baymiller et al.

Prospective 1-year open-label study

50 clozapine
(mean 454109 mg/day)

From baseline, 54.7 mg/dL (41.7%) increase in serum TG (P=0.001) and

14.4 mg/dL (7.5%) increase in TC (P<0.001). No significant changes in
HDL or LDL. Serum TG peaked between days 41 and 120 and then
declined, but still remained elevated. Use of propranolol exacerbated
increases in TC and TG.

Garyfallos et al.

8-week prospective randomized

study of acute schizophrenia
spectrum inpatients
25 olanzapine
(mean 18.0 mg/day)
25 risperidone
(mean 7.7 mg/day)

Compared to baseline, significant increases in fasting TG

(+43.5 mg/dL, P<0.001) for olanzapine but not risperidone (+7.5 mg/dL,
P>0.05). Nonsignificant increases in TC: +10.2 mg/dL for olanzapine,
+0.7 mg/dL for risperidone. Between-group difference in change was
significant for both TG and cholesterol parameters (P<0.001 for each).

Medical Illness and Schizophrenia





Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Weiden et al.

6-week ziprasidone switch study;

all labs were nonfasting;
mean ziprasidone dose 91 mg
Prior medications:
104 from olanzapine
58 from risperidone
108 from typical antipsychotics

Median changes from baseline for TG: 50 mg/dL in those switched from
olanzapine (P<0.0001) and 29 mg/dL in those switched from
risperidone (P <0.01). Median changes in cholesterol: 17 mg/dL
(P<0.0001) in those switched from olanzapine and 12 mg/dL
(P<0.005) for those switched from risperidone. Cholesterol levels
declined in 76% of patients switched from olanzapine and 72%
switched from risperidone. Reductions in lipid levels in patients
switched from typical antipsychotics were nonsignificant.

Effects of Antipsychotics on Serum Lipids

Melkersson and Cross-sectional study of lipids and

TG elevated in 44% of clozapine group and 56% of olanzapine group.
Dahl 2003
insulin parameters in long-term
Cholesterol elevated in 39% of clozapine group and 63% of olanzapine
treatment (mean 8.2 years for
group. High LDL in 17% of clozapine- and 38% of olanzapine-treated
clozapine, 1.2 years for olanzapine)
patients. Normal HDL in all but one of each drug group. No significant
18 clozapine
between-group differences in hyperlipidemia rates or median lipid
(median 400 mg/day)
16 olanzapine
(median 10 mg/day)


Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Almras et al.

Cross-sectional study of male

Risperidone subjects had significantly lower serum cholesterol, LDL, TG,
schizophrenia patients in Canada; all cholesterol:HDL ratio, and apolipoprotein B. Risperidone cohort had
subjects taking drug >6 months;
significantly higher HDL, larger LDL peak particle size, and greater
results compared with reference
apolipoprotein A1. Compared to reference group, no differences for
group of nondiabetic males (mean
olanzapine group for cholesterol, TG, or LDL, but lower HDL and
ages: 28.4 risperidone, 31.7
higher cholesterol:HDL ratio. Risperidone cohort had lower cholesterol
olanzapine, 32.8 controls)
and LDL but also lower HDL than the reference group.
42 olanzapine
(mean 12.4 mg/day)
45 risperidone
(mean 2.9 mg/day)

Cohen et al.

Chart review of 10 autistic adults (mean Data on lipids available for only 5 subjects; 4 of 5 had a decrease in TC,
age 43) switched to ziprasidone and
but mean decrease (2.6 mg/dL) was not significant; 3 of 5 had a
followed >6 months
decrease in serum TG, but mean decrease (21.8 mg/dL) was not
10 ziprasidone
(mean 128 mg/day)

Medical Illness and Schizophrenia





Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


De Deyn et al.

Randomized, fixed-dose, 10-week

Serum values not provided, but report states that cholesterol and TG
trial of olanzapine (1.0, 2.5, 5.0, 7.5
levels were not significantly different among or between treatment
mg/day) vs. placebo for psychosis/ groups.
behavioral disturbance in patients
with Alzheimers disease; mean age
76.6 years
520 olanzapine
129 placebo

Kinon et al.

Randomized, prospective 4-month

study comparing those who
switched to olanzapine with those
who continued prior medication
27 olanzapine
27 risperidone/typical

After 4 months, no significant within-group change in mean TC

(4.7 mg/dL, P=0.69) or TG (6.6 mg/dL, P=0.81) in patients switched
to olanzapine, or in those who continued taking typical antipsychotics/
risperidone (TC: 0.6 mg/dL, P =0.69; TG: +13.6 mg/dL, P=0.28).
Elevations in lipids noted prior to month 4 in olanzapine cohort, but
declined over time.

Effects of Antipsychotics on Serum Lipids



Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


McQuade et al.

26-week double-blind, randomized

trial in acutely relapsed
schizophrenia patients
156 aripiprazole
(mean 25.1 mg/day)
161 olanzapine
(mean 16.5 mg/day)

At week 26, mean change in fasting TG +79.4 mg/dL for olanzapine,

+6.5 mg/dL for aripiprazole (P <0.05); HDL 3.39 mg/dL for
olanzapine, +3.61 mg/dL for aripiprazole (P<0.05).
Changes in TC and LDL favored aripiprazole but were not significant:
TC +16.3 mg/dL olanzapine, 1.13 mg/dL aripiprazole; LDL +2.27 mg/
dL olanzapine, 3.86 mg/dL aripiprazole. Incidence of new
dyslipidemias was significantly greater for olanzapine
on the basis of elevated TC (>200 mg/dL: 47% olanzapine vs.
17% aripiprazole), LDL (>130 mg/dL: 38% olanzapine vs.
19% aripiprazole), and TG (>150 mg/dL: 50% olanzapine vs.
18% aripiprazole) (P<0.05 for each).

Medical Illness and Schizophrenia





Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Simpson et al.

6-week double-blind, randomized

trial in patients with acute
136 ziprasidone
(mean 129.9 mg/day)
133 olanzapine
(mean 11.3 mg/day)

Endpoint change in fasting TC was significant for olanzapine (median

+19.5 mg/dL, P<0.0001) vs. ziprasidone (median 1.0 mg/dL, P =0.48
from baseline) (P<0.0001 between groups). TG increased median
+26 mg/dL for olanzapine (P =0.0003), and decreased a median of
2 mg/dL for ziprasidone group (P =0.77); between-group difference
was significant (P<0.003). LDL increased by a median of 13 mg/dL for
olanzapine (P<0.0001) and decreased 1 mg/dL for ziprasidone group
(P=0.78); between-group difference was significant (P <0.0004).
Apolipoprotein B increased 9.0 mg/dL for olanzapine (P <0.0001)
and decreased 3.0 mg/dL for ziprasidone (P=0.17); between-group
difference was significant (P<0.0001). No impact or between-group
differences in HDL, apolipoprotein A1, or lipoprotein (a) levels.

Waage et al.

Case report:
1 male, age 42, with chronic
paranoid psychosis and

Baseline TC 189 mg/dL. After 7 months, TC 282 mg/dL. At time of

admission with pancreatitis, 19 months after starting olanzapine, TC
552 mg/dL and TG 2,044 mg/dL.

Effects of Antipsychotics on Serum Lipids



Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Ball et al. 2005

Case report:
1 male, age 42, with
schizoaffective disorder and
hypercholesterolemia and
hypertriglyceridemia that
resolved after switch to

Peak lipid levels: TC 477 mg/dL and TG 4,758 mg/dL. Patient

hospitalized subsequently due to noncompliance, with lipid levels
at time of admission: cholesterol 213 mg/dL, TG 298 mg/dL, and
LDL 146 mg/dL while taking simvastatin. Trial of aripiprazole,
up to 45 mg/day for 5 weeks, with lipids (off of simvastatin):
cholesterol 163 mg/dL, TG 145 mg/dL, LDL 100 mg/dL.

Breier et al.

28-week randomized, double-blind

All between-group lipid changes significantly favored ziprasidone.
study in patients with schizophrenia Endpoint change in fasting TC for olanzapine (+3.09 mg/dL, P =0.07)
277 olanzapine
vs. ziprasidone (12.74 mg/dL, P=0.08) (P<0.0001 between groups).
(mean 15.27 mg/day)
TG increased +34.52 mg/dL for olanzapine (P=0.09) and decreased
271 ziprasidone
21.24 mg/dL for ziprasidone (P=0.11); between-group difference was
(mean 115.96 mg/day)
significant (P<0.001). LDL increased by a median of 0.77 mg/dL for
olanzapine (P=0.06) and decreased 10.42 mg/dL for ziprasidone
(P=0.07); between-group difference was significant (P <0.001). HDL
decreased 2.32 mg/dL for olanzapine (P=0.02) and increased
0.77 mg/dL for ziprasidone (P=0.02); between-group difference was
significant (P=0.001).

Medical Illness and Schizophrenia





Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Brown and
Estoup 2005

Chart review study from Veterans

Administration medical center
88 ziprasidone
103 olanzapine

Olanzapine associated with significant increases in TC (+16 mg/dL,

P=0.01) and TG (+61 mg/dL, P=0.05), but not HDL or LDL.
Ziprasidone associated with significant decrease in TC (15 mg/dL,
P=0.01) and LDL (18 mg/dL, P=0.001), and increase in HDL
(+3 mg/dL, P =0.05). Among those who switched between agents,
significant differences were found for TC (P=0.05) and LDL (P =0.01).

Graham et al.

Prospective study of resting energy

expenditure and metabolic
outcomes in 9 adults started on
olanzapine and followed 12 weeks
9 olanzapine
(range 2.520 mg/day)

Median changes were 7 mg/dL for HDL (P =0.19), +1 mg/dL for LDL
(P=0.47), +59 mg/dL for TG (P=0.04). The TG change was a 62.8%
increase from baseline.

et al. 2005

Review of data on
96 subjects treated with
clozapine and followed for
10 years

A significant linear increase in TG levels was found for the duration of

the follow-up (0.5 mg/dL/month, P =0.04) but not for TC (P=0.13).
Elevations in TG and TC were associated with new-onset DM but not
cardiovascular mortality.

Effects of Antipsychotics on Serum Lipids



Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Lambert et al.

Case-control study of MediCal claims For the 12-week exposure window, olanzapine (OR 1.20, 95%
after schizophrenia diagnosis and
CI 1.081.33) was significantly associated with hyperlipidemia risk
exposure to only one antipsychotic
compared with typical agents, but not compared with exposure to
within 12 weeks prior to
clozapine, risperidone, or quetiapine. The odds ratio for olanzapine was
hyperlipidemia claim; cases were
greater than for risperidone (P=0.002). Increasing the exposure window
age- and gender-matched to patients to 24 or 52 weeks did not affect the results, although clozapines
with schizophrenia who did not
association did become significant at 24 weeks (OR 1.22,
develop hyperlipidemia
95% CI 1.031.45).

et al. 2005

Phase 1 of CATIE schizophrenia trial (See Tables 62 and 63)

McKee et al.

Chart review of 41 adults with

At endpoint, no significant changes in serum TC, LDL, or TG.
developmental delays switched
from typical antipsychotics to
olanzapine or risperidone (and some
then to olanzapine) and followed
up to 2 years. At endpoint:
33 olanzapine
8 risperidone

Medical Illness and Schizophrenia





Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


Meyer et al.

20-week switch study in overweight At endpoint, nonsignificant decreases in serum TG (13.1 mg/dL) and
or obese olanzapine-treated patients HDL (1.6 mg/dL). Overall, the prevalence of metabolic syndrome
with schizophrenia following
decreased from 53.5% to 36.6% over 20 weeks (P <0.005).
metabolic syndrome parameters
71 risperidone

Simpson et al.

6-month continuation of prior 6-week There were significant within-group median increases from baseline in
randomized, double-blind study of
TC (+13.0 mg/dL, P=0.03) and LDL (+17.0 mg/dL, P =0.04) with
olanzapine vs. ziprasidone
olanzapine, and nonsignificant changes in TC (1.0 mg/dL, P =0.98),
55 ziprasidone
and LDL (9.0 mg/dL, P=0.29) with ziprasidone.
(mean 135.2 mg/day)
71 olanzapine
(mean 12.6 mg/day)

Su et al. 2005

Crossover switch study in 15

After switch to olanzapine, nonsignificant changes in LDL (1.0 mg/dL),
schizophrenia patients followed for
HDL (+0.4 mg/dL), and TC (+20.5 mg/dL), but significant increase
3 months after switch
in TG (+84.3 mg/dL, P<0.05). In those switched to risperidone,
7 risperidone olanzapine
nonsignificant changes in LDL (+5.9 mg/dL), HDL (+6.8 mg/dL),
8 olanzapine risperidone
and TC (8.6 mg/dL), but significant decrease in TG (86.0 mg/dL,

Effects of Antipsychotics on Serum Lipids



Lipid changes during antipsychotic therapy: 19842008 (continued)

Study design


McEvoy et al.

Phase 2 nonresponse arm of CATIE

schizophrenia trial

(See Table 62, Phase 2b results)

Stroup et al.

Phase 2 intolerability arm of CATIE

schizophrenia trial

(See Table 62, Phase 2a results)

Meyer et al.

End of CATIE phase 1 detailed


(See Table 63)

Meyer et al.

CATIE phase 1 nonfasting

triglycerides analysis

Mean 3-month changes: quetiapine (+54.7 mg/dL), olanzapine

(+23.4 mg/dL), ziprasidone (+0.0 mg/dL), risperidone (18.4 mg/dL),
perphenazine (1.3 mg/dL).

Note. ADHD= attention-deficit/hyperactivity disorder; BMI= body mass index; CI=confidence interval; CPZ=chlorpromazine;
DM=diabetes mellitus; HDL= high-density lipoprotein; LDL= low-density lipoprotein; OR=odds ratio; RR=relative risk; SE=standard
error of measurement; TC =total cholesterol; TG = triglycerides; VLDL= very-low-density lipoprotein.

Medical Illness and Schizophrenia




TABLE 62. Exposure-adjusted lipid changes for CATIE schizophrenia trial phases 1, 2a, and 2b (mean SE)


Phase 1 (N=1,460)

9.4 2.4

1.5 2.7

Phase 2a (n= 444)

17.5 5.2

5.9 4.7

Phase 1 (N=1,460)
Phase 2a (n= 444)




6.6 2.4

1.3 2.4

8.2 3.2

6.5 5.3

3.1 5.2

10.7 5.1

1.0 7.1

11.0 8.1

7.4 8.7

40.5 8.9

9.2 10.1

21.2 9.6

2.4 9.1

16.5 12.2

94.1 21.8

39.3 22.1

5.2 21.6

3.5 20.9

43.8 21.2

5.3 32.0

7.1 36.2

30.0 39.0

Cholesterol (mg/dL)

Phase 2b (n=99)
Triglycerides (mg/dL)

Phase 2b (n=99)

Blood chemistry change from phase 2 baseline to average of two largest values.

Effects of Antipsychotics on Serum Lipids



Medical Illness and Schizophrenia


TABLE 63. Mean changes in high-density lipoprotein (HDL) and fasting

triglycerides at end of CATIE phase 1
HDL, mg/dLa


Fasting triglycerides, mg/dLb

Below median Above median







1.7 (0.6)

0.9 (0.9)

49.0 (10.8)

5.2 (17.4)


2.7 (0.7)

1.3 (1.0)

28.7 (11.6)

27.5 (22.3)


0.2 (0.6)

0.1 (1.1)

29.8 (10.8)

13.0 (18.4)


0.1 (0.6)

0.9 (0.9)

19.7 (11.2)

67.1 (21.2)


0.6 (0.9)

4.3 (1.4)

26.0 (15.6)

96.4 (28.5)






Note. Table entries are analysis of covariance least-squares adjusted means

(standard error of measurement). NS=not significant (P 0.05).
Data presented in separate columns due to significant race treatment effect.
Data presented in separate columns due to significant baselinetreatment effect. Median triglyceride level=148 mg/dL.
Between-group comparisons significant for perphenazine vs. olanzapine
(P<0.001) and for perphenazine vs. quetiapine (P=0.002).
Between-group comparisons significant for ziprasidone vs. olanzapine
(P=0.002) and for ziprasidone vs. perphenazine (P= 0.001).
Between-group comparison significant for olanzapine vs. ziprasidone

Effects of Antipsychotics on Serum Lipids


Patient Variables and Possible Mechanisms

for Antipsychotic-Related Hyperlipidemia
Based on a review of data on certain metabolic outcomes, such as the development of type 2 diabetes mellitus or diabetic ketoacidosis, ethnicity
and obesity stand out as important predictors of risk and are additive
with that imposed by the antipsychotic medication itself (Jin et al. 2002,
2004). As of 2008, no important trends regarding patient risk for hyperlipidemia can be assigned on the basis of ethnicity, gender, patient
weight, or even medication dosing within the range commonly used to
treat schizophrenia. There are case reports of significant hypertriglyceridemia with low-dose olanzapine (e.g., 5 mg) (Meyer 2001) but no data
for those medications employed at extremely low doses (relative to their
antipsychotic dosing), such as quetiapine, which is frequently used in
the United States as a sedative at doses of 25100 mg. The only demographic variable that may be predictive of decreased risk for dyslipidemia is age, with the caveat that both olanzapine studies in older
subjects previously cited were performed among subjects in controlled
settings, presumably with controlled diets as well (Barak and Aizenberg
2003; De Deyn et al. 2004). Conversely, reports exist of dyslipidemia in
adolescents exposed to antipsychotic medications, primarily olanzapine, so younger age is not protective (Domon and Cargile 2002; Domon
and Webber 2001; Martin and LEcuyer 2002; Nguyen and Murphy 2001;
Penzak and Chuck 2002; Shaw et al. 2001; Stigler et al. 2004).
That an antipsychotic agent can induce dyslipidemia is not entirely
surprising given that hyperlipidemia can occur with a variety of medications, including certain diuretics, progestins, -adrenergic antagonists, immunosuppressive agents, protease inhibitors, and some
anticonvulsants (Echevarria et al. 1999; Mantel-Teeuwisse et al. 2001;
Penzak and Chuck 2002). In their comprehensive review of the subject,
Mantel-Teeuwisse et al. (2001) noted that global changes in serum lipids are described with some medications, whereas certain agents appear to have specific effects on particular lipid fractions. For example,
isotretinoin, acitretin, certain protease inhibitors, low-potency phenothiazines, and dibenzodiazepine-derived antipsychotics primarily elevate serum triglyceride levels. Moreover, in a manner that parallels the
differential metabolic effects of the atypical antipsychotics, the protease
inhibitors also vary dramatically in their metabolic effects (Penzak and
Chuck 2002).
Any agent may induce hyperlipidemia through several possible
means, although none of these has been proven definitively for the


Medical Illness and Schizophrenia

atypical antipsychotics. Nonetheless, several biologically plausible

hypotheses have been advanced, focusing on weight gain, dietary
changes, and the development of insulin resistance, to explain the high
incidence of hyperlipidemia with certain antipsychotic medications.
The variation in weight gain liability is quite marked among the
atypical antipsychotics (Allison et al. 1999), with clozapine and olanzapine associated with the greatest gains. Because obesity and weight
gain have a demonstrable negative impact on serum lipid profiles,
those atypical antipsychotics most likely to cause significant weight
gain are also correlated with the greatest impact on serum lipids (Stone
et al. 2005). Nonetheless, recent switch data suggest that the more metabolically offending medications may have direct, weight-independent
effects on serum lipids. In a long-term switch study, Weiden et al. (2007)
charted the time course of weight and lipid changes over 58 weeks after
switching subjects from typical antipsychotics, risperidone, and olanzapine to ziprasidone. Those switching from high-potency typical antipsychotics experienced no significant lipid changes, but those
previously taking olanzapine and risperidone experienced an immediate reduction in lipids (triglycerides more than total cholesterol) over
the first 6 weeks after the switch, as well as slow but steady weight loss
over the course of the study. The rapid improvement in serum lipids
during a time frame when weight loss had been minimal points to a direct, weight-independent effect of certain antipsychotic medications on
serum lipids.
Accumulating data suggest that the metabolic pathway most likely
to mediate the increase in serum triglycerides relates to the development of insulin resistance. In those who have become less sensitive to
the action of insulin, the inability of insulin to adequately suppress lipolysis in adipose cells results in an outflow of free fatty acids and dyslipidemia, characterized primarily by elevated serum triglycerides
(Reaven 2005). The literature certainly supports the concept that some
patients who develop new-onset type 2 diabetes mellitus experience
hypertriglyceridemia (Meyer 2001), often with reversal of these problems upon discontinuation of the offending agent (Casey et al. 2003;
Meyer et al. 2005; Weiden et al. 2007). Although the majority of patients
with elevated triglycerides related to antipsychotic treatment do not
have overt type 2 diabetes mellitus, many do show signs of insulin resistance, as seen in studies of glucose-insulin parameters in patients
taking clozapine or olanzapine (Melkersson and Dahl 2003).
More recent data suggest that quetiapine might induce triglyceride
changes by another mechanism. Prospective data presented at the 2007
American Psychiatric Association annual meeting (Newcomer et al.

Effects of Antipsychotics on Serum Lipids


2007) indicated that quetiapine-exposed patients experienced changes

in serum triglyceride levels without an apparent change in glucose tolerance, as measured by standard glucose tolerance testing. Compelling
biological data support Weiden et al.s (2007) switch study findings and
point to direct effects of the more metabolically offending medications
on insulin sensitivity independent of changes in adiposity. Single doses
of medications such as olanzapine and clozapine have been shown to
induce loss of insulin sensitivity in a dose-dependent fashion in laboratory animals in a manner not seen with risperidone or ziprasidone
(Houseknecht et al. 2007). Evidence for this effect can be directly measured within 2 hours of drug exposure, using the hyperinsulinemiceuglycemic clamp technique, as both the decreased ability to metabolize
glucose and the failure to adequately suppress endogenous glucose
production from the liver. The means by which clozapine and olanzapine induce these effects is not known, but the propensity to cause insulin resistance with single doses strongly suggests that these two agents
have weight-independent effects on glucose-insulin homeostasis, effects that will be exacerbated by future medication-related weight gain.

Monitoring Recommendations for

Hyperlipidemia During Antipsychotic
Because multiple cardiovascular risk factors exist in patients with
schizophrenia (Goff et al. 2005), caution must be exercised in the choice
of antipsychotic therapy to minimize the added morbidity and mortality of hyperlipidemia. Unfortunately, hyperlipidemia is undertreated
in schizophrenia patients (Nasrallah et al. 2006), thereby exposing patients to ongoing substantial additional cardiovascular risk. For example, for a normotensive smoker exposed to a dibenzodiazepine-derived
atypical antipsychotic, the 10-year risk for a major cardiovascular event
(e.g., sudden death, acute myocardial infarction) might be increased
significantly from baseline after only 12 weeks of therapy (Daumit et al.
2008). Hyperlipidemia is associated with long-term cardiovascular consequences, yet monitoring for hyperlipidemia is not solely a long-term
issue, because severe hypertriglyceridemia also represents a risk for
acute pancreatitis (Koller et al. 2003).
The following guidelines are based on previously published recommendations (American Diabetes Association et al. 2004; De Hert et
al. 2006; Marder et al. 2004; Melkersson et al. 2004; Meyer et al. 2006;
Sernyak 2007) and my own clinical experience. Some of the consider-


Medical Illness and Schizophrenia

ations inherent in these recommendations include the following: patients with schizophrenia have multiple risk factors for cardiovascular
disease; certain antipsychotics are associated with greater adverse effects on serum lipids; many health care providers outside of the psychiatric arena are unaware of the potential metabolic complications of
atypical antipsychotic therapy; and schizophrenia patients often receive
limited or no medical care outside of that provided by the mental health
practitioner, so the burden of medical monitoring necessarily falls on
those who prescribe antipsychotic medications. Although the following
recommendations are specific to monitoring of serum lipids, these are
understood to be part of the monitoring recommended elsewhere as part
of routine medical care for those taking atypical antipsychotics.

Baseline Assessment
For each patient with schizophrenia, the clinician should document
in the medical record the individuals smoking status, as well as the
patients and first-degree family members history of cardiovascular
disease, hyperlipidemia, and glucose intolerance.
The clinician should obtain weight, waist circumference, blood pressure, and (ideally) fasting lipid panel for each patient. Total cholesterol and HDL values are valid on nonfasting specimens, but
triglyceride and LDL levels are not. These measures are recommended for all patients with schizophrenia, regardless of medication regimen, given the limited health care access for these patients.

For patients taking agents associated with lower risk for hyperlipidemia (high-potency typical antipsychotics, ziprasidone, risperidone, aripiprazole), an annual fasting lipid panel is sufficient unless
dyslipidemia is suspected from the baseline evaluation.
For patients taking agents associated with higher risk for hyperlipidemia (low-potency typical antipsychotics, quetiapine, olanzapine,
clozapine), a quarterly fasting lipid panel is necessary for the first
year to detect cases of severe hypertriglyceridemia. Testing frequency may be decreased to semiannually if fasting lipids remain
normal but should continue on a quarterly basis in those identified
with abnormal values.
All patients with persistent dyslipidemia should be referred for
lipid-lowering therapy or considered for switch to a less offending
agent if possible.

Effects of Antipsychotics on Serum Lipids


Key Clinical Points

Patients with severe mental illness have twice the risk for cardiovascular
mortality versus their nonmentally ill counterparts. Therefore, the
induction of hyperlipidemia secondary to antipsychotic treatment represents a serious condition not only because of its impact on cardiovascular risk but also because it is occurring in a group that possesses
considerable risk.

High-potency typical antipsychotic agents (butyrophenones) are lipid

neutral, whereas low-potency agents (phenothiazines) are associated
with hyperlipidemia, primarily in the form of hypertriglyceridemia.

The structurally related dibenzodiazepine-derived atypical antipsychotics (clozapine, olanzapine, and quetiapine) are associated with greater
elevations in serum triglycerides than in total cholesterol, whereas the
non-dibenzodiazepine agents (risperidone, ziprasidone, and aripiprazole) have minimal effects on lipids.

Known mechanisms by which antipsychotics cause hyperlipidemia include weight gain, dietary changes, and the direct development of insulin resistance.

A thorough baseline assessment in all patients should include a cardiovascular assessment. Subsequent monitoring for hyperlipidemia is recommended annually for all patients prescribed antipsychotics and as
frequently as quarterly for those prescribed higher risk agents.

Allison DB, Mentore JL, Heo M, et al: Antipsychotic-induced weight gain: a
comprehensive research synthesis. Am J Psychiatry 156:16861696, 1999
Almras N, Desprs JP, Villeneuve J, et al: Development of an atherogenic metabolic risk factor profile associated with the use of atypical antipsychotics.
J Clin Psychiatry 65:557564, 2004
American Diabetes Association, American Psychiatric Association, American
Association of Clinical Endocrinologists, et al: Consensus Development
Conference on Antipsychotic Drugs and Obesity and Diabetes. Diabetes
Care 27:596601, 2004
Atmaca M, Kuloglu M, Tezcan E, et al: Serum leptin and triglyceride levels in
patients on treatment with atypical antipsychotics. J Clin Psychiatry
64:598604, 2003a


Medical Illness and Schizophrenia

Atmaca M, Kuloglu M, Tezcan E, et al: Weight gain, serum leptin and triglyceride levels in patients with schizophrenia on antipsychotic treatment with
quetiapine, olanzapine and haloperidol. Schizophr Res 60:99100, 2003b
Ball MP, Hooper ET, Skipwith DF, et al: Clozapine-induced hyperlipidemia resolved after switch to aripiprazole therapy. Ann Pharmacother 39:1570
1572, 2005
Bansal S, Buring JE, Rifai N, et al: Fasting compared with nonfasting triglycerides and risk of cardiovascular events in women. JAMA 298:309316,
Baptista T, Lacruz A, Angeles F, et al: Endocrine and metabolic abnormalities
involved in obesity associated with typical antipsychotic drug administration. Pharmacopsychiatry 34:223231, 2001
Barak Y, Aizenberg D: Effects of olanzapine on lipid abnormalities in elderly
psychotic patients. Drugs Aging 20:893896, 2003
Baymiller SP, Ball P, McMahon RP, et al: Serum glucose and lipid changes during the course of clozapine treatment: the effect of concurrent beta-adrenergic antagonist treatment. Schizophr Res 59:4957, 2003
Bouchard RH, Demers M-F, Simoneau I, et al: Atypical antipsychotics and cardiovascular risk in schizophrenic patients. J Clin Psychopharmacol 21:110
111, 2001
Braun GA, Paulonis ME: Sterol metabolism: biochemical differences among the
butyrophenones. Int J Neuropsychiatry 3(suppl):2627, 1967
Breier A, Berg PH, Thakore JH, et al: Olanzapine versus ziprasidone: results of
a 28-week double-blind study in patients with schizophrenia. Am J Psychiatry 162:18791887, 2005
Brown RR, Estoup MW: Comparison of the metabolic effects observed in patients treated with ziprasidone versus olanzapine. Int Clin Psychopharmacol 20:105112, 2005
Casey DE, Carson WH, Saha AR, et al: Switching patients to aripiprazole from
other antipsychotic agents: a multicenter randomized study. Psychopharmacology 166:391399, 2003
Clark ML, Johnson PC: Amenorrhea and elevated serum cholesterol produced
by a trifluoro-methylated phenothiazine. J Clin Endocrinol Metabol
20:641646, 1960
Clark ML, Ray TS, Paredes A, et al: Chlorpromazine in women with chronic
schizophrenia: the effect on cholesterol levels and cholesterol-behavior relationships. Psychosom Med 29:634642, 1967
Clark ML, Braun GA, Hewson JR, et al: Trifluperidol and cholesterol in man.
Clin Pharmacol Ther 9:333340, 1968
Clark M, Dubowski K, Colmore J: The effect of chlorpromazine on serum cholesterol in chronic schizophrenic patients. Clin Pharmacol Ther 11:883889,
Clark ML, Huber WK, Sullivan J, et al: Evaluation of loxapine succinate in
chronic schizophrenia. Dis Nerv Syst 33:783791, 1972

Effects of Antipsychotics on Serum Lipids


Cohen SA, Fitzgerald BJ, Khan SR, et al: The effect of a switch to ziprasidone in
an adult population with autistic disorder: chart review of naturalistic,
open-label treatment. J Clin Psychiatry 65:110113, 2004
Cullen P: Evidence that triglycerides are an independent coronary heart disease
risk factor. Am J Cardiol 86:943949, 2000
Daumit GL, Goff DC, Meyer JM, et al: Antipsychotic effects on estimated 10year coronary heart disease risk in the CATIE schizophrenia study.
Schizophr Res 105:175187, 2008
De Deyn PP, Carrasco MM, Deberdt W, et al: Olanzapine versus placebo in the
treatment of psychosis with or without associated behavioral disturbances in
patients with Alzheimers disease. Int J Geriatr Psychiatry 19:115126, 2004
De Hert M, van Eyck D, De Nayer A: Metabolic abnormalities associated with
second generation antipsychotics: fact or fiction? Development of guidelines for screening and monitoring. Int Clin Psychopharmacol 21 (suppl
2):S11S15, 2006
Domon SE, Cargile CS: Quetiapine-associated hyperglycemia and hypertriglyceridemia. J Am Acad Child Adolesc Psychiatry 41:495496, 2002
Domon SE, Webber JC: Hyperglycemia and hypertriglyceridemia secondary to
olanzapine. J Child Adolesc Psychopharmacol 11:285288, 2001
Downs JR, Clearfield M, Weis S, et al: Primary prevention of acute coronary
events with lovastatin in men and women with average cholesterol levels:
results of AFCAPS/TexCAPS. Air Force/Texas Coronary Atherosclerosis
Prevention Study. JAMA 279:16151622, 1998
Dursun SM, Szemis A, Andrews H, et al: The effects of clozapine on levels of
total cholesterol and related lipids in serum of patients with schizophrenia:
a prospective study. J Psychiatry Neurosci 24:453455, 1999
Eberly LE, Stamler J, Neaton JD, et al: Relation of triglyceride levels, fasting and
nonfasting, to fatal and nonfatal coronary heart disease. Arch Intern Med
163:10771083, 2003
Echevarria KL, Hardin TC, Smith JA: Hyperlipidemia associated with protease
inhibitor therapy. Ann Pharmacother 33:859863, 1999
Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults: Executive Summary of the Third Report of the National
Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA 285:24862497, 2001
Fleischhacker WW, Stuppack C, Moser C, et al: Fluperlapine vs haloperidol: a
comparison of their neuroendocrinological profiles and the influence on
serum lipids. Pharmacopsychiatry 19:111114, 1986
Garyfallos G, Dimelis D, Kouniakis P, et al: Olanzapine versus risperidone:
weight gain and elevation of serum triglyceride levels. Eur Psychiatry
18:320321, 2003
Gaulin BD, Markowitz JS, Caley CF, et al: Clozapine-associated elevation in serum triglycerides. Am J Psychiatry 156:12701272, 1999


Medical Illness and Schizophrenia

Ghaeli P, Dufresne RL: Elevated serum triglycerides on clozapine resolve with

risperidone. Pharmacotherapy 15:382385, 1995
Ghaeli P, Dufresne RL: Serum triglyceride levels in patients treated with clozapine. Am J Health Syst Pharm 53:20792081, 1996
Goff DC, Sullivan L, McEvoy JP, et al: A comparison of ten-year cardiac risk estimates in schizophrenia patients from the CATIE study and matched controls. Schizophr Res 80:4553, 2005
Goodnick PJ, Jerry JM: Aripiprazole: profile on efficacy and safety. Expert Opin
Pharmacother 3:17731781, 2002
Graham KA, Perkins DO, Edwards LJ, et al: Effect of olanzapine on body composition and energy expenditure in adults with first-episode psychosis.
Am J Psychiatry 162:118123, 2005
Haffner SM, Lehto S, Rnnemaa T, et al: Mortality from coronary heart disease
in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med 339:229234, 1998
Henderson DC, Cagliero E, Gray C, et al: Clozapine, diabetes mellitus, weight
gain, and lipid abnormalities: a five-year naturalistic study. Am J Psychiatry 157:975981, 2000
Henderson DC, Nguyen DD, Copeland PM, et al: Clozapine, diabetes mellitus,
hyperlipidemia and cardiovascular risks and mortality: results of a 10-year
naturalistic study. J Clin Psychiatry 66:11161121, 2005
Houseknecht KL, Robertson AS, Zavadoski W, et al: Acute effects of atypical
antipsychotics on whole body insulin resistance in rats: implications for adverse metabolic effects. Neuropsychopharmacology 32:289297, 2007
Jeppesen J, Hein HO, Suadicani P, et al: Triglyceride concentration and ischemic
heart disease: an eight-year follow-up in the Copenhagen Male Study. Circulation 97:10291036, 1998 [Erratum published in: Circulation 97(11):1995,
Jin H, Meyer JM, Jeste DV: Phenomenology of and risk factors for new-onset diabetes mellitus and diabetic ketoacidosis associated with atypical antipsychotics: an analysis of 45 published cases. Ann Clin Psychiatry 14:5964,
Jin H, Meyer JM, Jeste DV: Atypical antipsychotics and glucose dysregulation:
a systematic review. Schizophr Res 71:195212, 2004
Kingsbury SJ, Fayek M, Trufasiu D, et al: The apparent effects of ziprasidone on
plasma lipids and glucose. J Clin Psychiatry 62:347349, 2001
Kinon BJ, Basson BR, Gilmore JA, et al: Long-term olanzapine treatment: weight
change and weight-related health factors in schizophrenia. J Clin Psychiatry 62:92100, 2001
Kinon BJ, Liu-Seifert H, Ahl J, et al: Longitudinal effect of olanzapine on fasting
serum lipids: a randomized, prospective, 4-month study. Ann N Y Acad Sci
1032:295296, 2004
Koller E, Cross JT, Doraiswamy PM, et al: Pancreatitis associated with atypical antipsychotics: from the Food and Drug Administrations MedWatch surveillance system and published reports. Pharmacotherapy 23:11231130, 2003

Effects of Antipsychotics on Serum Lipids


Koro CE, Fedder DO, LItalien GJ, et al: An assessment of the independent effects of olanzapine and risperidone exposure on the risk of hyperlipidemia
in schizophrenic patients. Arch Gen Psychiatry 59:10211026, 2002
Kwiterovich PO Jr: The metabolic pathways of high-density lipoprotein, lowdensity lipoprotein, and triglycerides: a current review. Am J Cardiol
86:5L10L, 2000
Lambert BL, Chang KY, Tafesse E, et al: Association between antipsychotic
treatment and hyperlipidemia among California Medicaid patients with
schizophrenia. J Clin Psychopharmacol 25:1218, 2005
Leonard P, Halley A, Browne S: Prevalence of obesity, lipid and glucose abnormalities in outpatients prescribed clozapine. Ir Med J 95:119120, 2002
Lieberman JA, Stroup TS, McEvoy JP, et al: Effectiveness of antipsychotic drugs in
patients with chronic schizophrenia. N Engl J Med 353:12091223, 2005
Lund BC, Perry PJ, Brooks JM, et al: Clozapine use in patients with schizophrenia and the risk of diabetes, hyperlipidemia, and hypertension: a claimsbased approach. Arch Gen Psychiatry 58:11721176, 2001
Mantel-Teeuwisse AK, Kloosterman JM, Maitland-van der Zee AH, et al: Druginduced lipid changes: a review of the unintended effects of some commonly used drugs on serum lipid levels. Drug Saf 24:443456, 2001
Marder SR, Essock SM, Miller AL, et al: Physical health monitoring of patients
with schizophrenia. Am J Psychiatry 161:13341349, 2004
Martin A, LEcuyer S: Triglyceride, cholesterol and weight changes among risperidone-treated youths: a retrospective study. Eur Child Adolesc Psychiatry 11:129133, 2002
Martnez JA, Velasco JJ, Urbistondo MD: Effects of pharmacological therapy on
anthropometric and biochemical status of male and female institutionalized psychiatric patients. J Am Coll Nutr 13:192197, 1994
McEvoy JP, Meyer JM, Nasrallah HA, et al: Prevalence of the metabolic syndrome in patients with schizophrenia: baseline results from the Clinical
Antipsychotic Trials of Intervention Effectiveness (CATIE) schizophrenia
trial and comparison with national estimates from NHANES III. Schizophr
Res 80:1932, 2005
McEvoy JP, Lieberman JA, Stroup TS, et al: Effectiveness of clozapine versus
olanzapine, quetiapine, and risperidone in patients with chronic schizophrenia who did not respond to prior atypical antipsychotic treatment. Am
J Psychiatry 163:600610, 2006
McKee JR, Bodfish JW, Mahorney SL, et al: Metabolic effects associated with
atypical antipsychotic treatment in the developmentally disabled. J Clin
Psychiatry 66:11611168, 2005
McLaughlin T, Abbasi F, Cheal K, et al: Use of metabolic markers to identify
overweight individuals who are insulin resistant. Ann Intern Med 139:802
809, 2003
McQuade RD, Stock E, Marcus R, et al: A comparison of weight change during
treatment with olanzapine or aripiprazole: results from a randomized,
double-blind study. J Clin Psychiatry 65 (suppl 18):4756, 2004


Medical Illness and Schizophrenia

Mefferd RB, Labrosse EH, Gawienowski AM, et al: Influence of chlorpromazine

on certain biochemical variables of chronic male schizophrenics. J Nerv
Ment Dis 127:167179, 1958
Melkersson KI, Dahl ML: Relationship between levels of insulin or triglycerides
and serum concentrations of the atypical antipsychotics clozapine and
olanzapine in patients on treatment with therapeutic doses. Psychopharmacology 170:157166, 2003
Melkersson KI, Hulting AL, Brismar KE: Elevated levels of insulin, leptin, and
blood lipids in olanzapine-treated patients with schizophrenia or related
psychoses. J Clin Psychiatry 61:742749, 2000
Melkersson KI, Dahl ML, Hulting AL: Guidelines for prevention and treatment
of adverse effects of antipsychotic drugs on glucose-insulin homeostasis
and lipid metabolism. Psychopharmacology 175:16, 2004
Meyer JM: Novel antipsychotics and severe hyperlipidemia. J Clin Psychopharmacol 21:369374, 2001
Meyer JM: A retrospective comparison of weight, lipid, and glucose changes between risperidone- and olanzapine-treated inpatients: metabolic outcomes
after 1 year. J Clin Psychiatry 63:425433, 2002
Meyer JM, Koro CE: The effects of antipsychotic therapy on serum lipids: a comprehensive review. Schizophr Res 70:117, 2004
Meyer JM, Simpson GM: Psychopharmacology from chlorpromazine to olanzapine: a brief history of antipsychotics. Psychiatr Serv 48:11371140, 1997
Meyer JM, Pandina G, Bossie CA, et al: Effects of switching from olanzapine to
risperidone on the prevalence of the metabolic syndrome in overweight or
obese patients with schizophrenia or schizoaffective disorder: analysis of a
multicenter, rater-blinded, open-label study. Clin Ther 27:19301941, 2005
Meyer JM, Loh C, Leckband SG, et al: Prevalence of the metabolic syndrome in
veterans with schizophrenia. J Psychiatr Pract 12:510, 2006
Meyer JM, Davis VG, Goff DC, et al: Change in metabolic syndrome parameters
with antipsychotic treatment in the CATIE schizophrenia trial: prospective
data from phase 1. Schizophr Res 101:273286, 2008a
Meyer JM, Davis VG, Goff DC, et al: Impact of antipsychotic treatment on nonfasting triglycerides in the CATIE schizophrenia trial phase 1. Schizophr
Res 103:104109, 2008b
Mller-Oerlinghausen B: A short survey on untoward effects of fluperlapine.
Arzneimittelforschung 34:131134, 1984
Nasrallah HA, Meyer JM, Goff DC, et al: Low rates of treatment for hypertension, dyslipidemia and diabetes in schizophrenia: data from the CATIE
schizophrenia trial sample at baseline. Schizophr Res 86:1522, 2006
Newcomer JW, Hennekens CH: Severe mental illness and risk of cardiovascular
disease. JAMA 298:17941796, 2007
Newcomer JW, Ratner RE, Eriksson JW, et al: Metabolic changes following 24
weeks treatment with quetiapine, olanzapine, or risperidone in patients
with schizophrenia. Poster presented at the annual meeting of the American Psychiatric Association, San Diego, CA, May 2007

Effects of Antipsychotics on Serum Lipids


Nguyen M, Murphy T: Olanzapine and hypertriglyceridemia. J Am Acad Child

Adolesc Psychiatry 40:133, 2001
Nordestgaard BG, Benn M, Schnohr P, et al: Nonfasting triglycerides and risk
of myocardial infarction, ischemic heart disease, and death in men and
women. JAMA 298:299308, 2007
Osby U, Correia N, Brandt L, et al: Mortality and causes of death in schizophrenia in Stockholm county, Sweden. Schizophr Res 45:2128, 2000
Osby U, Brandt L, Correia N, et al: Excess mortality in bipolar and unipolar disorder in Sweden. Arch Gen Psychiatry 58:844850, 2001
Osser DN, Najarian DM, Dufresne RL: Olanzapine increases weight and serum
triglyceride levels. J Clin Psychiatry 60:767770, 1999
Penzak SR, Chuck SK: Management of protease inhibitorassociated hyperlipidemia. Am J Cardiovasc Drugs 2:91106, 2002
Prevention of cardiovascular events and death with pravastatin in patients with
coronary heart disease and a broad range of initial cholesterol levels. The
Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID)
Study Group. N Engl J Med 339:13491357, 1998
Randomised trial of cholesterol lowering in 4444 patients with coronary heart
disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 344:1383
1389, 1994
Reaven GM: Compensatory hyperinsulinemia and the development of an atherogenic lipoprotein profile: the price paid to maintain glucose homeostasis in insulin-resistant individuals. Endocrinol Metab Clin North Am 34:4962, 2005
Rubins HB: Triglycerides and coronary heart disease: implications of recent
clinical trials. J Cardiovasc Risk 7:339345, 2000
Saari K, Koponen H, Laitinen J, et al: Hyperlipidemia in persons using antipsychotic medication: a general population-based birth cohort study. J Clin
Psychiatry 65:547550, 2004
Sacks FM, Pfeffer MA, Moye LA, et al: The effect of pravastatin on coronary
events after myocardial infarction in patients with average cholesterol levels. Cholesterol and Recurrent Events Trial investigators. N Engl J Med
335:10011009, 1996
Saha S, Chant D, McGrath J: A systematic review of mortality in schizophrenia:
is the differential mortality gap worsening over time? Arch Gen Psychiatry
64:11231131, 2007
Sasaki J, Kumagae G, Sata T, et al: Decreased concentration of high density lipoprotein cholesterol in schizophrenic patients treated with phenothiazines. Atherosclerosis 51:163169, 1984
Sasaki J, Funakoshi M, Arakawa K: Lipids and apolipoproteins in patients
treated with major tranquilizers. Clin Pharmacol Ther 37:684687, 1985
Serafetinides EA, Colmore JP, Rahhal DK, et al: Trifluperidol in chronic male
psychiatric patients. Behav Neuropsychiatry 3:1012, 1971
Serafetinides EA, Collins S, Clark ML: Haloperidol, clopenthixol, and chlorpromazine in chronic schizophrenia: chemically unrelated antipsychotics as
therapeutic alternatives. J Nerv Ment Dis 154:3142, 1972


Medical Illness and Schizophrenia

Sernyak MJ: Implementation of monitoring and management guidelines for

second-generation antipsychotics. J Clin Psychiatry 68 (suppl 4):1418,
Shafique M, Khan IA, Akhtar MH, et al: Serum lipids and lipoproteins in
schizophrenic patients receiving major tranquilizers. J Pak Med Assoc
38:259261, 1988
Shaw JA, Lewis JE, Pascal S, et al: A study of quetiapine: efficacy and tolerability
in psychotic adolescents. J Child Adolesc Psychopharmacol 11:415424,
Sheitman BB, Bird PM, Binz W, et al: Olanzapine-induced elevation of plasma
triglyceride levels (letter). Am J Psychiatry 156:14711472, 1999
Shepherd J, Cobbe SM, Ford I, et al: Prevention of coronary heart disease with
pravastatin in men with hypercholesterolemia. West of Scotland Coronary
Prevention Study Group. N Engl J Med 333:13011307, 1995
Simpson GM, Cooper TB: The effect of three butyrophenones on serum cholesterol levels. Curr Ther Res Clin Exp 8:249255, 1966
Simpson GM, Cooper TB, Braun GA: Further studies on the effect of butyrophenones on cholesterol synthesis in humans. Curr Ther Res Clin Exp 9:413
418, 1967
Simpson GM, Glick ID, Weiden PJ, et al: Randomized, controlled, double-blind
multicenter comparison of the efficacy and tolerability of ziprasidone and
olanzapine in acutely ill inpatients with schizophrenia or schizoaffective
disorder. Am J Psychiatry 161:18371847, 2004
Simpson GM, Weiden P, Pigott T, et al: Six-month, blinded, multicenter continuation study of ziprasidone versus olanzapine in schizophrenia. Am J Psychiatry 162:15351538, 2005
Smith DA: Treatment of the dyslipidemia of insulin resistance. Med Clin North
Am 91:11851210, 2007
Spivak B, Roitman S, Vered Y, et al: Diminished suicidal and aggressive behavior, high plasma norepinephrine levels, and serum triglyceride levels in
chronic neuroleptic-resistant schizophrenic patients maintained on clozapine. Clin Neuropharmacol 21:245250, 1998
Spivak B, Lamschtein C, Talmon Y, et al: The impact of clozapine treatment on
serum lipids in chronic schizophrenic patients. Clin Neuropharmacol
22:98101, 1999
Spurling RD, Lamberti JS, Olsen D, et al: Changes in metabolic parameters with
switching to aripiprazole from another second-generation antipsychotic: a
retrospective chart review. J Clin Psychiatry 68:406409, 2007
Stamler J, Wentworth D, Neaton JD: Is relationship between serum cholesterol
and risk of premature death from coronary heart disease continuous and
graded? Findings in 356,222 primary screenees of the Multiple Risk Factor
Intervention Trial (MRFIT). JAMA 256:28232828, 1986
Stigler KA, Potenza MN, Posey DJ, et al: Weight gain associated with atypical
antipsychotic use in children and adolescents: prevalence, clinical relevance, and management. Paediatr Drugs 6:3344, 2004

Effects of Antipsychotics on Serum Lipids


Stone NJ, Bilek S, Rosenbaum S: Recent National Cholesterol Education Program Adult Treatment Panel III update: adjustments and options. Am J
Cardiol 96:53E59E, 2005
Stoner SC, Dubisar BM, Khan R, et al: Severe hypertriglyceridemia associated
with olanzapine. J Clin Psychiatry 63:948949, 2002
Stroup TS, Lieberman JA, McEvoy JP, et al: Effectiveness of olanzapine, quetiapine, risperidone, and ziprasidone in patients with chronic schizophrenia following discontinuation of a previous atypical antipsychotic. Am J
Psychiatry 163:611622, 2006
Su KP, Wu PL, Pariante CM: A crossover study on lipid and weight changes associated with olanzapine and risperidone. Psychopharmacology 183:383
386, 2005
Vaisanen K, Rimon R, Raisanen P, et al: A controlled double-blind study of haloperidol versus thioridazine in the treatment of restless mentally subnormal patients: serum levels and clinical effects. Acta Psychiatr Belg 79:673
685, 1979
Vampini C, Steinmayr M, Bilone F, et al: The increase of plasma levels of triglyceride during clozapine treatment: a case report (abstract). Neuropsychopharmacology 10:249S, 1994
Virkkunen M, Wahlbeck K, Rissanen A, et al: Decrease of energy expenditure
causes weight increase in olanzapine treatment: a case study. Pharmacopsychiatry 35:124126, 2002
Waage C, Carlsson H, Nielsen EW: Olanzapine-induced pancreatitis: a case report. Journal of the Pancreas 5:388-391, 2004
Weiden PJ, Daniel DG, Simpson GM, et al: Improvement in indices of health
status in outpatients with schizophrenia switched to ziprasidone. J Clin
Psychopharmacol 23:16, 2003
Weiden PJ, Newcomer JW, Loebel AD, et al: Long-term changes in weight and
plasma lipids during maintenance treatment with ziprasidone. Neuropsychopharmacology 33:985994, 2007
Wetterling T: Bodyweight gain with atypical antipsychotics: a comparative review. Drug Saf 24:5973, 2001
Wetterling T: [Hyperlipidemia: side-effect of the treatment with an atypical
antipsychotic (zotepine)]? (in German). Psychiatr Prax 29:438440, 2002
Wilson PW, Abbott RD, Castelli WP: High density lipoprotein cholesterol
and mortality. The Framingham Heart Study. Arteriosclerosis 8:737741,
Wirshing DA, Boyd JA, Meng LR, et al: The effects of novel antipsychotics on
glucose and lipid levels. J Clin Psychiatry 63:856865, 2002
Wu G, Dias P, Chun W, et al: Hyperglycemia, hyperlipemia, and periodic paralysis: a case report of new side effects of clozapine. Prog Neuropsychopharmacol Biol Psychiatry 24:13951400, 2000

This page intentionally left blank

The Spectrum of
Disease in Patients
With Schizophrenia
Jimmi Nielsen, M.D.
Egon Toft, M.D., F.E.S.C.

Life expectancy for patients with schizophrenia in the United

States is 61 years, or 15 years shorter than for the general population
(Hennekens et al. 2005), even after controlling for unnatural causes of
death (e.g., accidents and suicide). Cardiovascular disease (CVD) is the
major cause of excess mortality in patients with schizophrenia, and this
patient cohort is twice as likely to die of coronary heart disease (CHD)
as the general population (Brown et al. 2000). There are multiple
sources of increased CVD risk, including sedentary lifestyle, smoking,
and the metabolic side effects of antipsychotic treatment. Although
seemingly of less clinical importance, much attention has also been paid
to the increased risk of sudden cardiac death due to tachyarrhythmias,
such as torsade de pointes (TdP), that are induced by antipsychotics. Interestingly, although several antipsychotics have been withdrawn from
the market due to QTc interval prolongation and concern over torsade
de pointes risk, among the antipsychotics associated with increased risk


Medical Illness and Schizophrenia

for CVD mortality mediated by substantial weight gain and dysmetabolic effects, none have been suspended. Although the evidence base is
sparse, the accumulated literature suggests that antipsychotic exposure
accounts for relatively few arrhythmia deaths compared to the expected greater numbers of deaths from myocardial and cerebral infarctions occurring secondary to antipsychotic-induced metabolic side
effects and patient lifestyle factors.
Our purpose in this chapter is to describe the spectrum of CVD, with
a particular focus on those specific issues related to schizophrenia and
the medications commonly used for schizophrenia treatment. The
chapter is not a complete guide to CVD, but we hope to equip psychiatrists and other physicians with the basic tools required to meet the
challenge of treating and preventing CVD in patients with schizophrenia. In this chapter, we first provide the reader with a general background for understanding the relationship between schizophrenia,
medication treatment, and TdP, as this information is necessary for the
understanding of arrhythmia risk related to medication effects, and
later provide an overview of atherosclerotic disease and basic CHD risk

Sudden Cardiac Death

The World Health Organization defines sudden cardiac death (SCD) as
unexpected death within 1 hour of symptom onset if witnessed, or
within 24 hours of the person having being alive and symptom free, if
unwitnessed (Turakhia and Tseng 2007). CHD is involved in up to 70%
of sudden death cases, and in many instances the first symptom of CHD
is sudden death (Turakhia and Tseng 2007); however, establishing the
exact cause of death can be difficult because several causes (overdose,
suicide, cerebrovascular events, seizures, etc.) must be excluded. In
most instances, an autopsy can determine the exact cause of death. If no
structural disease is found, cardiac arrhythmia is the most likely etiology.
Retrospective data suggest that the overall risk of SCD is increased
4.9 times in patients with schizophrenia compared with the general
population (Ruschena et al. 2003), with multiple contributing etiologies.
Patients with schizophrenia have more prevalent cardiovascular risk
factors compared to the background population due to lack of exercise,
obesity, smoking, diabetes, and dyslipidemia (Leucht et al. 2007). In
schizophrenia patients, the prevalence of type 2 diabetes mellitus is increased twofold, smoking threefold, obesity twofold, and dyslipidemia

Cardiovascular Disease


up to fivefold. Substance abuse and dependence, especially alcohol and

cocaine, are also more common in patients with schizophrenia and are
linked to risk for sudden death. Furthermore, patients with schizophrenia are less likely to have proper preventive medical care and monitoring for metabolic disorders than the general population (Leucht et al.
2007) and, when a primary care provider is consulted, are less likely to
follow instructions on diet or exercise.
Antipsychotic treatment has also been linked to increased risk of
SCD. A 2.4-fold increased risk of SCD was found in users of antipsychotics versus nonusers in a retrospective cohort study from Tennessee
(Ray et al. 2001). Cardiac arrhythmia from previously undiagnosed
CHD and risk from antipsychotic medication exposure are the two
most plausible explanations, but other suggestions that have been put
forth include respiratory dyskinesia, laryngeal-pharyngeal dystonia,
and peripheral vasodilation leading to cardiovascular collapse and
oversedation (Haddad and Anderson 2002; Titier et al. 2005).
Although cardiac arrhythmia is the most common etiology for SCD,
it is worthwhile distinguishing whether the arrhythmia is primary or
whether it is secondary to structural changes related to cardiomyopathy, myocarditis, or acute myocardial infarction. Secondary ventricular
arrhythmia is probably the most frequent form of fatal tachyarrhythmia, but the exact distribution of deaths due to structural changes
(acute myocardial infarction or cardiomyopathy) or nonstructural
changes (cardiac arrhythmia) is unclear because no large-scale mortality study with autopsy results has been carried out (Michelsen and
Meyer 2007). In general, the incidence of drug-induced ventricular arrhythmia is largely unknown, partly due to underreporting of spontaneous adverse effects. Drug-induced TdP is a rare event, so extremely
large sample sizes would be needed to detect a small increase in SCD
risk. Further complicating the assignment of TdP as the cause of any arrhythmic death is the need to confirm the diagnosis with electrocardiographic (ECG) recordings obtained at the time of the event.

Electrophysiology of the Heart

The electrophysiology of the heart is complex, and several ion channels
are involved in the repolarization and depolarization of ventricular
cells. The depolarization event, which includes the action potential, is
primarily mediated by the rapid influx of sodium and is reflected on the
electrocardiogram as the QRS complex (Figure 71). Blocking of the
sodium channel is called a quinidine-like effect, named after the class I

Medical Illness and Schizophrenia


antiarrhythmic drug quinidine. This effect can be seen on the electrocardiogram as a widening of the QRS complex and increased PR interval,
and drugs with quinidine-like properties are associated with increased
risk of SCD due to ventricular arrhythmia. Some antipsychotics, such as
thioridazine, have been shown to possess quinidine-like properties at
high dosages, but among commonly used psychotropic medications,
only tricyclic antidepressants affect depolarization to any significant












QT interval

FIGURE 71. The role of ion channels during depolarization and

A. The monophasic action potential with notations for involved ion channels.
B. The corresponding surface electrocardiogram.
Source. Reprinted from Titier K, Girodet PO, Verdoux H, et al: Atypical
Antipsychotics: From Potassium Channels to Torsade de Pointes and Sudden
Death. Drug Safety 28:3551, 2005. Used with permission.

Cardiovascular Disease


Repolarization is mediated primarily by the efflux of potassium via

two families of potassium channels: the rapid IKr and the slow IKs channels. The IKr channel is encoded by the human ether-a-go-go related
gene (HERG), polymorphisms of which are involved in the congenital
long QT syndrome, type II. Patients with congenital long QT syndrome
have one or more mutations in the genes encoding for ion channels involved in the repolarization and depolarization of the heart, with nine
known genetic variants involved in this disease. The cardinal symptoms of congenital long QT syndrome are syncope or SCD, and in some
cases the first symptom is sudden death (Priori et al. 2003). Many antipsychotic drugs block the IKr channel to an extent comparable with that
seen in congenital long QT syndrome. Antagonism of IKr channels is the
mechanism mostly responsible for instances of drug-induced QT prolongation and is the suspected mechanism for the majority of antipsychotic-induced sudden cardiac deaths.
The prolongation of the repolarization process, and thereby of the
QT interval, is thought to allow the development of spontaneous contractions of non-pacemaker cells called early after-depolarizations
(EADs). The induction of EADs, whether drug-induced or genetic as in
the congenital long QT syndrome, can also increase risk for ventricular
arrhythmia, as the development of EADs is linked to an increased inward current and/or decreased outward current during the repolarization phase of the action potential (Glassman and Bigger 2001;
Haverkamp et al. 2000).

Torsade de Pointes
Torsade de pointes (TdP), which literally means twisting points in
French, is a polymorphic ventricular arrhythmia and the most common
drug-induced ventricular arrhythmia. The name is derived from the
twisting of the QRS complex around the isoelectric line, as shown in
Figure 72A. TdP can be asymptomatic or cause self-limiting palpitations, dizziness, syncope, and death. In approximately one-third of patients, TdP will lead to ventricular fibrillation, and in 10% result in SCD
(Figure 72B) (Abdelmawla and Mitchell 2006). In one Swedish study,
the annual incidence of drug-induced TdP was 4 per 100,000 patients,
although the real incidence has been estimated to be at least 10 times
higher (Wysowski and Bacsanyi 1996).


Medical Illness and Schizophrenia


Examples of torsade de points (TdP).

A. Electrocardiogram shows a self-limited case of TdP. Notice the typical shift

between an upward and a downward pointing of the complexes.
B. A case of TdP deteriorating into ventricular fibrillation.
Source. Reprinted from Yap YG, Camm AJ: Drug Induced QT Prolongation
and Torsades de Pointes. Heart 89:13631372, 2003. Used with permission.

Risk Factors for Torsade de Pointes

Risk factors for developing TdP are shown in Table 71 (Yap and Camm
2003). An important issue to consider when exploring drug-related
causes is the potential for pharmacokinetic drug-drug interactions that
increase the plasma levels of an offending drug and result in dosedependent QTc prolongation. Pharmacodynamic interactions can occur
when two or more drugs are prescribed that prolong the QTc interval,
with the combined effects resulting in significant QT prolongation.
Medications with quinidine-like properties (e.g., tricyclic antidepressants) given together with QT prolonging antipsychotics is just a single
example of a combination of psychiatric medications that may increase
the risk of sudden death.
An additional risk is that induced by biological variability. In 10%
15% of patients with drug-induced TdP, mutations or polymorphisms
were found in one of the congenital long QT genes (Yang et al. 2002).
These patients have smaller repolarization reserve and are therefore
more susceptible to drug-induced changes. Although genetic causes
provide an obvious source of risk, most cases of TdP occur in patients
with structural heart disease or other risk factors (Table 71). For this
reason, drug-induced TdP is comparably rarer in young, healthy pa-

Cardiovascular Disease



Risk factors for developing drug-induced torsade de pointes


Female gender
Hypokalemia (e.g., vomiting and diarrhea)
Structural heart disease (e.g., cardiomyopathy or infarction)
Atrioventricular block
Prolonged QTc (>450 milliseconds)
Significant T-wave abnormalities
Genetic disposition/congenital long QT syndrome
Prior drug-induced TdP
Multiple QT-prolonging drugs or other drugs that interfere with
their metabolism
Hepatic impairment
Drug-drug interactions


Adapted from Yap and Camm 2003.

tients than in older individuals, who have a greater likelihood of having

established cardiac disease.

Estimating the Risk: Surrogate Markers

Great interest has been shown in developing surrogate markers for the
clinical evaluation of medication proarrhythmic effects, but so far no
valid marker has been elucidated. The QTc interval is the most widely
used marker for determining the proarrhythmic effects of any drug, but
an exact estimate of risk cannot be derived solely from changes in the
QTc interval. Even among patients with prolonged QTc intervals, the
actual risk of TdP is often quite low. This fact alone poses great difficulties for the identification of patients at risk for TdP and SCD.
The overall safety of an antipsychotic drug rests on the balance between protective and causative factors for SCD. QTc-prolonging drugs
might also have actions that diminish the proarrhythmic risk, such as
increasing the heart rate or altering other ion currents, which counteract
the proarrhythmic effects (Taylor 2003a). Diminished long-term SCD
risk can also result if the medication is associated with limited weight
gain (or weight loss) and thereby lowers mortality by decreasing traditional CHD risk associated with obesity. For these reasons, the relationship between drug-induced QTc prolongation and the risk of sudden
death is not simple, and false positives are seen with drugs that prolong


Medical Illness and Schizophrenia

the QTc interval but are only rarely associated with TdP. Amiodarone
is a class III antiarrhythmic that is rarely associated with TdP but is
known to significantly prolong the QTc interval (Hohnloser et al. 1994).
Other examples include verapamil, ziprasidone, and sertindole, all of
which prolong the QT interval significantly, but whose association with
TdP is questionable (De Cicco et al. 1999; Haddad and Anderson 2002).
Because QT effects by themselves do not always predict TdP risk, surrogate markers for drug-induced arrhythmia risk have been studied, as
presented in the next section, including some recent and technologically more advanced parameters developed from electrocardiograms in
patients with long QT syndrome. Although these studies are not commonly used in routine clinical practice, an understanding of changes in
T-wave morphology may be very useful in the near future, and parameters based on T-wave morphological changes have evolved to the extent that they are expected to be employed in drug studies in the next
few years (Couderc et al. 2008; Kanters et al. 2004; Struijk et al. 2006).

QT Interval Estimation
The QT interval comprises the entire interval from the beginning of the
QRS complex to the termination of the T-wave (Figure 73). The end of
the T-wave is defined as the intercept between the isoelectric line and
the tangent of the steepest part of the descending T-wave. Determination of the exact point of T-wave termination is critical to the estimation
of QT interval duration and is subject to certain pitfalls. A U-wave incorporated in the T-wave often complicates QT measurement and can
lead to the erroneous assessment of QTc prolongation because the Uwave is included in the interval. T-wave alterations, such as biphasic or
flat T-waves (e.g., due to IKr blockade), can cause similar problems. A
U-wave is most often present in lateral precordial leads, whereas biphasic T-waves are present in multiple leads. For cases where any of
these features are present, manual estimation of the QT interval is a
better option, with possible referral to a cardiologist for interpretation.
Manual QT interval measurement should be derived from at least
three to five cycles, and a mean value calculated across these cycles. Although the accuracy level of manual QT interval determination is 2040
milliseconds (Goldenberg et al. 2006), the accuracy of automatic measurement of the QT interval varies greatly, because manufacturers have
chosen varying algorithms for QT calculations (Goldenberg et al. 2006).
The QT interval can be measured in any lead, yet the length of the interval can differ between leads. Although there is no consensus on which

Cardiovascular Disease


QRS complex

T peak

T end


PR interval

T asc.

T desc.



QT interval


As shown, finding the end of the T-wave can be difficult, particularly in cases
where a U-wave is interfering.
asc. =ascending; desc. =descending; TPTE=time from peak to end of T-wave.
Source. Reprinted from Pater C: Methodological Considerations in the Design of Trials for Safety Assessment of New Drugs and Chemical Entities.
Current Controlled Trials in Cardiovascular Medicine 6:113, 2005. Used with permission of Biomed Central Ltd.

leads must be used, V3 and V4 have generally been suggested as the

preferred leads for QT estimation (Malik 2004; Sadanaga et al. 2006).
Another source of potential confusion for QT estimation lies in the
fact that the QT interval shortens with increasing heart rate, making
correction necessary. Several correction formulas have been suggested
(Figure 74), of which the Bazett formula is the most commonly used;
however, it is less accurate at heart rates above 70 beats per minute
(bpm) (Dogan et al. 2005). The Fridericia correction formula performs

Medical Illness and Schizophrenia


QTcB 2


QTcF = 3

The most commonly used correction formulas.

The Fridericia formula is recommended and performs better than the Bazett formula in cases of tachycardia or bradycardia. QT is measured in milliseconds
and RR in seconds. The RR interval used should be the preceding beat to the QT
interval measured. The notation QTcB or QTcF can be used to indicate whether
the method used was the Bazett or the Fridericia method.

better than Bazett at heart rates above 70 bpm and is often used in
newer ECG machines and by the pharmaceutical industry for drug
evaluation. Heart rate correction is especially important in patients receiving antipsychotic drugs, because most antipsychotics increase resting heart rate.
The reference interval for the QTc interval associated with greater
TdP risk is not clearly defined, but most agree that a QTc interval
greater than 500 milliseconds indicates an area of increased risk. Females tend to have longer QTc intervals, in part related to the fact that
testosterone decreases the QT interval (Gupta et al. 2007; Rautaharju et
al. 1992), and 70% of TdP cases occur in women (Vieweg 2002). The importance of QT interval prolongation has been suggested to be 1.052X
where X is prolongation above 440 milliseconds divided by 10. For example, prolongation from 440 milliseconds to 500 and 640 milliseconds
would increase the risk for TdP by 1.0526 = 1.4 and 1.05220 =2.8 times, respectively (Moss 1993). In the general population, the risk for TdP is
rare, and therefore a doubling would still mean a small increase in the
number of cases. Furthermore, these data are based completely on findings from individuals with congenital long QT syndrome, and whether
these calculations can be extrapolated to drug-induced TdP risk is unclear.
The QT interval length fluctuates throughout the day related to several factors, including inherent diurnal variation, level of activity, and
body weight. The QTc interval is longest during nights and shortest after awakening (Molnar et al. 1996); moreover, drug-induced prolongation of the QTc interval is greatest during the first half of the menstrual
cycle (Rodriguez et al. 2001). The effect of weight gain on QTc prolon-

Cardiovascular Disease


gation has been quantified, with a 10-kg weight increase associated

with an increase of over 5-milliseconds in QTc length (el-Gamal et al.
1995). Besides these biological factors, inaccurate measurement of and
correction for heart rate add to the variation in reported QT intervals
and complicate the assessment of drug-related effects on QTc and level
of TdP risk. As alluded to earlier, some proposed surrogate markers for
drug-related TdP risk have been tested and show promise. These include QT dispersion, time from the peak to the end of the T-wave
(TpeakTend, also called TPTE), T-wave morphology, and heart-rate variability (HRV).

QT Dispersion
That the QT interval differs between different leads is partly due to differences in repolarization time between the various myocardial layers.
The QT dispersion is the difference in time between the lead with the
shortest and the lead with the longest QT interval and is thus a measure
of myocardial heterogeneity during the repolarization process. A QT
dispersion of more than 100 milliseconds, or an incremental difference
of more than 100%, has been associated with proarrhythmic effects of
medications (Haddad and Anderson 2002). Amiodarone, a class III
antiarrhythmic drug that prolongs the QT interval but rarely induces
TdP, was found to decrease repolarization heterogeneity and thereby
decrease QT dispersion (Hii et al. 1992; Hohnloser 1997). Risperidone
has been shown to increase the QT interval but without an incremental
increase in QT dispersion, and thus would be not associated with TdP
risk, a fact borne out by clinical experience (Yerrabolu et al. 2000). Arguments against the use of QT dispersion as a surrogate marker relate
to the fact that the QT interval itself varies widely and in some leads can
be difficult to estimate. Because QT dispersion relies on QT measurements in all 12 leads, there is added variability, leading some to argue
that further studies are needed to determine the validity of QT dispersion as a surrogate marker for TdP risk.

T-Wave-Related Measurements
Other proposed surrogate markers rely solely on T-wave-related
measurements. One of these is the time from the peak to the end of the
T-wave (see Figure 73). TPTE exists as an absolute value and also as
a measure of dispersion as calculated in the QT dispersion equation.
Although TPTE purportedly reflects transmural dispersion of repolarization (Antzelevitch et al. 2007), changes in T-wave morphology
are another surrogate marker for TdP risk derived from changes seen in


Medical Illness and Schizophrenia

patients with congenital long QT syndrome (Malik and Camm 2001;

Struijk et al. 2006). The congenital long QT syndrome type II is due to
mutations in the HERG potassium channel gene, and the electrocardiogram in these individuals demonstrates T-wave flattening and
notches. Unfortunately, the widespread use of T-wave morphology as
a TdP risk marker has been complicated by difficulties in quantifying
these morphological changes (Taylor 2003a), but newer data, derived
from more intensive study of long QT patients, has overcome many of
the technical hurdles to the extent that this parameter will be used to
monitor TdP risk in future drug studies (Couderc et al. 2008; Kanters et
al. 2004; Struijk et al. 2006).

Heart Rate Variability

During inspiration and expiration, the heart rate accelerates and decelerates, creating an innate source of variability that is a measure of heart
responsiveness. HRV thus relates to these beat-to-beat alterations in the
heart rate and is usually derived from a 24-hour Holter recording. HRV
reflects peripheral autonomic tone, and drugs with anticholinergic
properties reduce HRV (Rechlin et al. 1998). The importance of reduced
HRV for arrhythmia risk is based on findings from post-myocardial infarction studies indicating that decreased HRV is associated with increased mortality (Zuanetti et al. 1996).
Olanzapine and clozapine have been found to reduce HRV, but the
importance of this finding is unclear (Mueck-Weymann et al. 2002;
Rechlin et al. 1994). On the other hand, thioridazine, which is associated
with significant QTc prolongation and sudden death, was found to improve HRV (Silke et al. 2002). Interestingly, reduced HRV is seen not
only in antipsychotic-treated individuals but also in drug-naive patients with schizophrenia (Malaspina et al. 2002), suggesting that decreased HRV might be related to the disease itself, through inherent
biological mechanisms or an inactive lifestyle. As with other markers,
more studies are needed to clarify the importance of HRV in studying
the impact of antipsychotics on SCD risk.

Antipsychotic Drugs
Virtually all antipsychotic drugs prolong the QT interval to some measurable extent (Figure 75), and patients treated with antipsychotic
drugs typically have longer QTc intervals and higher heart rates than
do controls (Cohen et al. 2001). A critical issue when estimating the extent of drug-induced QT prolongation is the baseline QT value. Because

Cardiovascular Disease


patients are rarely drug-free at baseline for most clinical trials, the true
extent of QT prolongation from the drug under study can be underestimated if the baseline medication also prolongs the QT interval. Furthermore, kinetic or dynamic drug-drug interactions can increase measured
QT prolongation.
In general, atypical antipsychotics are associated with a lower risk of
arrhythmia than conventional antipsychotics (Sicouri and Antzelevitch
2008), yet conclusions drawn from such findings often fail to account
for the long-term impact of metabolic side effects on cardiac risk. One
paper estimated that 416 patients per 100,000 treated will die from the
sequelae of clozapine-induced weight gain within a 10-year period
(Fontaine et al. 2001), a figure that is much higher than mortality rates
from directly induced drug-related cardiac arrhythmia. Moreover, in
this same 10-year time frame, only 1015 clozapine-treated patients
would die from agranulocytosis.

Conventional Antipsychotic Drugs

Conventional antipsychotic drugs were approved at a time when the
demand for cardiac safety data was less strict. As a consequence, many
widely used conventional antipsychotic drugs have not been thoroughly investigated, although some have gained notoriety based on
newer safety data. Thioridazine and the related compound meso-

QTc change (milliseconds)



QTc prolongation with antipsychotics.

Adapted from Glassman and Bigger 2001.





t ia












Medical Illness and Schizophrenia

ridazine (which is metabolized to thioridazine) received black-box

warnings in the U.S. market in 2001 due to QTc data that emerged from
premarketing studies for ziprasidone. In these studies, the mean QTc
prolongation during treatment with thioridazine was 35 milliseconds
(Salih et al. 2007). The Pfizer data were not the sole source for the warning; clinical data indicated an increased risk of developing TdP and
SCD with thioridazine (Reilly et al. 2002). In one Finnish study analyzing 49 sudden deaths during treatment with antipsychotics, thioridazine was involved in more than half of the cases, a finding that
could not be explained by its relative market share (Mehtonen et al.
1991). Other drugs under suspicion or withdrawn from various markets include droperidol and pimozide. In some instances, the suspicion
or withdrawal was based solely on measured QTc effects, despite a
paucity of clinical cases of drug-induced TdP or SCD.
Haloperidol has been widely used for many years and is in general
considered to have a wide margin of cardiac safety based on overdose
and other data accumulated over decades of use. Although haloperidol
is associated with an average 5-millisecond QTc prolongation, more
than 20 cases of haloperidol-induced TdP have been published (Yap
and Camm 2003). This impression of safety might explain the cases of
TdP, because haloperidol is often prescribed in intensive care units and
other critical care hospital settings among a cohort of patients with
overt cardiac comorbidity. Patients with cardiac complications are
more vulnerable and often monitored more intensely with electrocardiography, a fact that might increase the number of detected TdPs,
some of which may be incidental to the use of haloperidol and are the
result of underlying comorbid conditions (e.g., electrolyte disturbances,
use of other medications).

Atypical Antipsychotics
Sertindole was introduced in 1997 as an atypical antipsychotic but was
voluntarily withdrawn from the market a year later due to a possible
link to sudden death. Despite the early concerns, subsequent postmarketing surveillance studies failed to show an increased risk of sudden
death or arrhythmia (Peuskens et al. 2007). The mean QTc prolongation
from sertindole is 20 milliseconds at therapeutic dosages, but so far no
TdP cases with sertindole as the offending agent have been published.
Based on further surveillance data documenting a general pattern of
low SCD risk, sertindole was reintroduced in several European countries in 2006 but with restricted use as a second-line agent and with
mandatory baseline and follow-up ECG monitoring.

Cardiovascular Disease


Ziprasidone also prolongs the QTc interval about 20 milliseconds but

has not been associated with TdP (Glassman and Bigger 2001). Maximal
QTc prolongation is reached at low therapeutic dosages. Reported exposures to ziprasidone dosages above 12 g have not resulted in TdP,
confirming the medications lack of TdP risk; moreover, the majority of
ziprasidones metabolism is through the aldehyde oxidase pathway,
one that is neither saturable nor inhibitable by medications, a feature
that decreases the risk for drug interactions (Taylor 2003b).

Clinical Implications
Basic monitoring for elements of cardiovascular risk should include
body weight, a measure of central adiposity (e.g., waist circumference
or hip:waist ratio), blood pressure, fasting lipids, and fasting glucose.
Whether routine ECG monitoring will decrease morbidity or mortality
in patients with schizophrenia has not been extensively studied; however, electrocardiography is a cheap and noninvasive procedure that
provides important information about cardiac status. Experts at the
2004 Mount Sinai Conference on the Pharmacotherapy of Schizophrenia concluded that electrocardiography is recommended when prescribing thioridazine, pimozide, or ziprasidone only when cardiac
complications, such as congenital long QT syndrome, known heart disease, history of syncope, or genetic disposition to sudden death (for individuals under age 40), are present (Marder et al. 2004). In contrast, the
Danish Health Board recommends obtaining an electrocardiogram at
baseline, after 12 weeks, and annually in all patients treated with antipsychotic drugs (Sundhedsstyrelsen 2007). The Maudsley Prescribing
Guidelines (Taylor et al. 2007) suggest monitoring with electrocardiography when patients are treated with high-dose antipsychotics or
combinations of drugs that have the propensity to prolong the QT interval. In conclusion, the evidence for routine ECG monitoring in patients
treated with antipsychotics is sparse, but we recommend ECG monitoring in patients with vulnerability to arrhythmia (e.g., electrolyte abnormalities, bradycardia, structural heart disease) and in those receiving
clozapine or drugs that significantly prolong the QT interval, have the
propensity to cause TdP, or have such monitoring mandated in the text
of the package insert.
When electrocardiograms are obtained, we recommend using automated calculations of the QTc interval, because most psychiatric clinics
and mental health professionals have less experience with manual
reading. Machine-read electrocardiograms tend to overestimate the QT
duration compared to manual readings, so in cases of significant QT


Medical Illness and Schizophrenia

prolongation based on machine interpretation, one should seek cardiologist confirmation. Many ECG machines still use the Bazett correction
formula; when that formula indicates that the heart rate is >70 bpm,
manual reading and correction using Fridericias method are recommended. The correction formula used by the machine can be found in
the manual, so one can state in the patient record which formula was
used for QT correction (e.g., QTcF indicates that Fridericias formula
has been used).
The following QTc parameters should lead to a discussion with the
patient about discontinuation of potentially offending medications or,
at the minimum, a reevaluation of benefits and risks of the current antipsychotic treatment (Sicouri and Antzelevitch 2008):
QTc prolongation >60 milliseconds
QTc interval >500 milliseconds
Drugs that significantly prolong the QTc interval (ziprasidone, sertindole) or are associated with TdP (e.g., thioridazine or pimozide) should
be avoided in patients with prolonged QTc interval and electrolyte abnormalities. When significant prolongations of QTc interval are detected in patients, serum potassium should be monitored and corrected.
Finally, patients experiencing syncope or palpitations should be referred for thorough examination to exclude other physical causes, including long QT syndrome.

Sudden Death and the Combination of

Benzodiazepines and Antipsychotics
Because the combination of antipsychotics and benzodiazepines has
been associated with sudden death, a black-box warning is provided regarding the use of parenteral olanzapine in combination with parenteral benzodiazepines (American Psychiatric Association 2005) based
on eight fatal events. Benzodiazepines and sedating antipsychotics
have respiratory depressant effects, and intramuscular administration
increases the risk for respiratory collapse due to rapid systemic absorption with no first-pass metabolism. Concomitant use of parenteral benzodiazepines and sedating antipsychotics (especially clozapine and
parenteral olanzapine) should prompt careful monitoring of blood
pressure, respiratory rate, and oxygen saturation using an oximeter
with an alarm (if available). In cases of respiratory depression, flumazenil can reverse the effect of benzodiazepines. The risk of sudden death
associated with the combination of benzodiazepines and an antipsychotic has been best described for clozapine and benzodiazepines (Rup-

Cardiovascular Disease


precht et al. 2004), and this risk is probably due to the sedative and
hypotensive effects of clozapine. Especially during the initial titration
of clozapine, when patients are not yet tolerant to its sedative properties, precautions should be taken with the concomitant use of benzodiazepines (Borentain et al. 2002).

Cardiac Aspects of Clozapine Treatment

Clozapine remains the drug of choice for patients with treatment-resistant
schizophrenia and for schizophrenia patients with suicidal ideation.
According to most treatment guidelines, a trial of clozapine should be
initiated after failed trials of two or three other antipsychotics. The superior efficacy of clozapine compared with other antipsychotics has
been established in several trials, but clozapine is not considered a firstline drug due to the risk of agranulocytosis. Clozapine is also associated
with several cardiac complications, and the estimated number of deaths
due to CVD is actually higher than the actual number of deaths due to
hematological causes (Fontaine et al. 2001; A.M. Walker et al. 1997). The
hematological monitoring system required for all clozapine patients
prevents the vast majority of agranulocytosis cases from being fatal, but
unrecognized cardiac complications remain a potentially serious source
of morbidity and mortality.
One early retrospective study found that clozapine exposure was associated with 3.8 times increased risk of SCD compared with use of
nonclozapine antipsychotics, based on six deaths in the clozapine
group; however, the control group was not matched for basic demographic variables such as severity of disease and duration of illness
(Modai et al. 2000). The absence of matching is a serious methodological
issue related to prescription bias, because clozapine is typically prescribed for more chronic patients who, by virtue of age and prior medication exposure, might possess more SCD risk factors. Furthermore, an
autopsy was performed in only one SCD patient, so the etiology of SCD
in some cases might be related to other etiologies (e.g., structural diseases). Given the small volume of published data, it is not clear whether
clozapine treatment is associated with an increased mortality risk from
SCD compared to treatment with other antipsychotics. Any added mortality risk associated with clozapine is ascribable primarily to other cardiac diseases, such as cardiomyopathy or myocarditis, or to the
secondary cardiovascular effects of clozapine-induced weight gain and
metabolic changes (Fontaine et al. 2001).
Myocarditis and Pericarditis. Although clozapine is unlikely to directly induce SCD, clozapine has been associated with some unique


Medical Illness and Schizophrenia

cardiac complications, including myocarditis and cardiomyopathy.

Myocarditis is an inflammation of the myocardial cells reported to occur within the first months of clozapine initiation. The development of
this condition is probably due to immunoglobulin Emediated hypersensitivity, a type 3 allergic reaction, but direct toxic effects on the myocardium (Killian et al. 1999) and even low selenium induced by
clozapine have also been proposed as mechanisms (Vaddadi et al.
2003). The incidence of clozapine-induced myocarditis is estimated to
be 1 in 500 and is fatal in up to 50% of cases (Killian et al. 1999). The presentation of myocarditis includes flulike symptoms, fever, dyspnea, sinus tachycardia, palpitations, fatigue, respiration-dependent pain, and
chest discomfort. Pain is usually present only if the pericardium is involved. The primary ECG findings are ST elevations and T-wave inversions, and common laboratory abnormalities include elevated serum
creatine kinase, leukocytosis, and eosinophilia. Eosinophilia is common
in patients developing myocarditis but also occurs without myocarditis. Usually the eosinophilia is transient.
Within the first months of treatment with clozapine, clinicians
should carefully examine any patients with flulike symptoms, fever, sinus tachycardia, and other cardiac symptoms to exclude myocarditis,
because the presentation may be subtle. The optimal way to diagnose
myocarditis is to measure blood troponin levels, because troponin reflects myocardial damage (Kay et al. 2002; Merrill et al. 2006). Some clinicians have suggested regular troponin testing during the initial
treatment with clozapine to increase the chances for prompt detection
of myocarditis, but no prospective data exist to endorse this as a routine
measure (Kay et al. 2002). In cases of myocarditis, clozapine must be
discontinued immediately and treatment with corticosteroids initiated.
Rechallenge with clozapine should be performed only in severe refractory schizophrenia cases without other viable treatment options, if the
benefits are considered to outweigh the possible serious risks, and with
close cardiac monitoring (Reid 2001).
Complicating the diagnosis of myocarditis is the finding that approximately 20% of patients treated with clozapine develop benign
hyperthermia within the first 3 weeks of treatment (Tham and Dickson
2002). Benign hyperthermia is often misdiagnosed as neuroleptic malignant syndrome, flu, or infection; however, the symptoms of benign
hyperthermia are similar to those of myocarditis, so in these instances a
careful examination must be performed. When benign hyperthermia
occurs in the beginning of clozapine treatment, a white blood count,
electrocardiogram, and troponin levels should be obtained to exclude

Cardiovascular Disease


Cardiomyopathy. The three types of cardiomyopathy are dilated (congestive), hypertrophic, and restrictive. Although dilated cardiomyopathy is a rare occurrence in the general population, in patients treated
with clozapine, it accounts for more than 60% of cardiomyopathy cases
(Meyer et al. 2007). Dilated cardiomyopathy typically starts in the left
ventricle where myocardial muscle fibers stretch and become thinner,
leading to chamber enlargement, and later spread to the right ventricle
and to the atria. As with myocarditis, the underlying mechanism is unclear, but one explanation is that in some patients, clozapine might have
a direct toxic effect on myocytes mediated by free radicals that induce
myocardial injury and resultant myopathy. Another explanation relies
on the fact that some cases of clozapine-induced cardiomyopathy have
evolved from myocarditis. The hypothesis is that prior myocarditis may
initiate an autoimmune process that further injures myocardial cells.
The myocarditis episode might have been subclinical and resolved without treatment, but the latent autoimmune process explains the relatively
late onset of cardiomyopathy during treatment compared with the earlier presentation of myocarditis. Clozapine-induced cardiomyopathy
has also been related to low selenium levels, but it is unknown whether
selenium supplementation can reduce the risk of cardiomyopathy or
Although case reports have been published, there is some debate
whether clozapine increases the risk for cardiomyopathy. Novartis reported that patients treated with clozapine had a cardiomyopathy incidence comparable to that of the U.S. population (Meyer et al. 2007);
however, failure to show a difference may be related to type II error
(given the infrequent occurrence in the general population) or might be
due to underreporting of spontaneous adverse events. Killian et al.
(1999) estimated the risk of clozapine-induced cardiomyopathy to be
51.5 per 100,000 patient-years, a fivefold increased risk over that seen in
the general population. The other issue in ascribing causality relates to
the fact that patients treated with clozapine have several known cardiomyopathy risk factors, including sinus tachycardia (a feature of clozapine treatment discussed below), alcohol abuse or dependence, CHD,
smoking, obesity, and metabolic disorders.
In more than 65% of reported cases, cardiomyopathy occurs after
6 months of clozapine treatment, with a 20% fatality rate (Merrill et al.
2005). Because cardiomyopathy is likely underreported, the incidence
may be higher. With cardiomyopathy, as with myocarditis, no dosage
dependency has been found. The physical symptoms are those related
to heart failure: exertional dyspnea, fatigue, peripheral edema, orthopnea, and chest pain. Diagnosing cardiomyopathy can be difficult in


Medical Illness and Schizophrenia

patients with schizophrenia because many patients already experience

dyspnea, peripheral edema, and fatigue due to poor physical conditioning. The electrocardiographic findings include sinus tachycardia, atrial
and ventricular arrhythmias, and ST-segment, T-wave, and P-wave abnormalities, but the electrocardiogram can remain normal in earlier, less
severe stages. Echocardiography is the most sensitive test for diagnosing cardiomyopathy, but it is unclear whether routine echocardiography is feasible or cost effective. Nonetheless, it should be performed
early and with short notice in cases where cardiomyopathy is suspected.
When cardiomyopathy is diagnosed, clozapine must be discontinued,
and the focus of medical treatment should be on treating the complications of heart failure. Clozapine-induced cardiomyopathy can be reversible if it is detected early and clozapine is promptly discontinued.
Sinus Tachycardia. One-fourth of patients treated with clozapine will
develop a 10- to 15-bpm increase in pulse rate. This effect is due to the
combination of anticholinergic vagal inhibition and an increase in circulation of catecholamines due to 1-adrenergic blockade (Safferman et
al. 1991). Clozapine-induced tachycardia is dosage dependent but can
occur at low dosages in sensitive individuals. Sinus tachycardia is particularly prominent during the initial titration, so a slower titration
schedule might resolve the problem in certain individuals.
Sinus tachycardia is often asymptomatic, but some patients complain
of increased heart rate. As tachyphylaxis can occur during the first 46
weeks of treatment, -adrenergic blocker treatment need not be initiated
during this period. Clinicians need to remember that tachycardia is a
symptom of both myocarditis and cardiomyopathy, so treatment with a
-adrenergic blocker might complicate the diagnosis of these conditions.
Whether asymptomatic clozapine-induced sinus tachycardia should
be corrected is unclear, but in general, persistent tachycardia (>100 bpm)
is a known cardiomyopathy risk factor, so -adrenergic blockers are often used in those patients with chronic tachycardia, although there is no
compelling data that this treatment reduces myopathy risk. Symptomatic sinus tachycardia should be treated with a cardioselective -adrenergic blocker (e.g., metoprolol) to avoid the possible effects of nonspecific
beta blockade (e.g., exacerbation of restrictive airway disease) (Young et
al. 1998); monitoring of blood pressure is also recommended.
Nonspecific T-Wave Changes. T-wave changes are common during
treatment with clozapine, and up to 20% of patients will develop ECG
abnormalities when switched to clozapine (Kang et al. 2000). The typical
T-wave changes that occur during clozapine therapy include flattening,

Cardiovascular Disease


inversion, and depression. The significance of these changes is unknown,

and they are most likely benign; however, it is important to exclude other
conditions causing T-wave changes such as myocarditis, myocardial infarction, or cardiomyopathy. Asymptomatic ST-segment elevation with
no relation to ischemia can also occur during treatment with clozapine
and probably has no clinical significance. As with certain T-wave
changes, the clinician should thoroughly exclude other potentially serious diagnoses related to ST-segment elevations, such as myocardial infarction, ischemia, myocarditis, and cardiomyopathy (Kang et al. 2000). In
addition to electrocardiogram, other testsechocardiography, serum
troponin levels, or stress testsmight be indicated on the basis of ECG
changes and the clinical picture so as to rule out other diagnoses.
QTc Prolongation. Clozapine has been associated with QTc prolongation (Kang et al. 2000; Lin et al. 2004), but the ability of clozapine to prolong the QT interval is questionable, in part due to the common
occurrence of clozapine-induced sinus tachycardia. Because many ECG
machines use the Bazett formula, the tendency is to overestimate the
corrected QT interval due to Bazetts limitations at heart rates >70 bpm
(Dogan et al. 2005); moreover, the frequent presence of T-wave abnormalities with clozapine exposure creates further difficulties for the exact estimation of the QT interval.
Orthostatic Hypotension. Orthostatic hypotension is defined as a
sustained decrease in blood pressure exceeding 20 mmHg systolic
or 10 mmHg diastolic occurring within 3 minutes of upright posture.
Several antipsychotic drugs have the propensity to induce orthostatic
hypotension due to potent 1-adrenergic antagonism. Among the atypical antipsychotics, clozapine and quetiapine are more often associated
with orthostatic hypotension, which generally seems to be dosage dependent, with dizziness corresponding roughly to the extent of postural
change. Orthostatic hypotension is often the limiting factor during the
up-titration of clozapine, but tachyphylaxis will occur within weeks, so
slowing the pace of titration may resolve the problem.

Cardiac Complications With Nonclozapine

Myocarditis and cardiomyopathy have been reported with antipsychotic drugs other than clozapine, but whether these cases are clearly
medication related is unknown. The most common cardiac-related
complications from nonclozapine antipsychotic drugs are due to ortho-


Medical Illness and Schizophrenia

static hypotension and QT prolongation, but one should not discount

the long-term effects of dyslipidemia and weight gain on cardiovascular risk. Orthostatic hypotension is especially problematic for elderly
patients due to poor vasomotor tone and can be troublesome due to the
increased risk of falls. Slower titration schedules may permit tolerance
to medication-induced postural changes and thereby diminish the risk
of falls and syncope.

Coronary Heart Disease

Coronary heart disease refers to the failure of coronary circulation,
whereas cardiovascular disease reflects all consequences of atherosclerotic diseases, including stroke and diseases of the heart. In CHD patients, inadequate circulation to the myocardium leads to ischemia and
eventually to angina or acute myocardial infarction. The typical cause
of CHD is atherosclerosis due to an accumulation of cholesterol-laden
atheromatous plaques in vessel walls that are prone to inflammation
and eventual rupture with vessel thrombosis. CHD is the most common
cause of death in Western societies and is responsible for about 20% of
all deaths (British Heart Foundation 2007).
Using baseline data from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) schizophrenia trial, Goff et al. (2005) estimated 10-year CHD risk with the algorithm derived from the
Framingham heart study (see Figures 76A and 76B) and found that
patients with schizophrenia had an increased risk for CVD compared
with matched subjects in the general population: 9.4% versus 7.0% for
males and 6.3% versus 4.2% for females. Several factors are involved in
this level of risk, including high smoking prevalence, sedentary lifestyle, obesity, medication effects, and perhaps the disease itself. Activity in the hypothalamic-pituitary-adrenal axis is increased in patients
with schizophrenia and leads to increased levels of catecholamines
and cortisol (E. Walker et al. 2008). Increased hypothalamic-pituitaryadrenal axis activity increases the risk for CVD, dyslipidemia, hypertension, vasoconstriction, and platelet activation, and might be another
mechanism that increases CHD risk in schizophrenia patients.
The CHD risk for any individual represents the sum of both nonmodifiable and modifiable factors. Nonmodifiable factors include age,
gender, history of CVD, and genetic predisposition. Modifiable factors
have been identified by large prospective trials, such as the INTERHEART study (Yusuf et al. 2004), and include the following:

Cardiovascular Disease


Abdominal obesity
Inadequate daily consumption of fruits and vegetables
Psychosocial factors, stress, and depression
Insufficient exercise
Alcohol intake

Risk factors are additive, and several lifestyle risk factors are known
to be increased in patients with schizophrenia. Dyslipidemia is one
modifiable CVD risk factor that is more prevalent in patients with
schizophrenia. Certain antipsychotic drugs are associated with deleterious changes in serum cholesterol and triglyceride levels; moreover,
the effects of these changes are compounded by the fact that patients
with schizophrenia are less likely to receive treatment with statins
(Hennekens et al. 2005). (See Chapter 6, Effects of Antipsychotics on
Serum Lipids, for more information about dyslipidemia and antipsychotics.) Smoking increases CVD mortality risk by 60%80%, and up to
80% of patients with schizophrenia smoke cigarettes (Hughes et al.
1986). Not only is smoking prevalence increased in this patient population, but smokers with schizophrenia inhale more deeply and extract
more compounds from the cigarettes than do other smokers (Olincy et
al. 1997). (For further information about smoking and schizophrenia,
see Chapter 9, Nicotine and Tobacco Use in Patients With Schizophrenia.) Diabetes mellitus increases CHD risk by two- to threefold in men
and three- to sixfold in women (Hennekens et al. 2005), and the prevalence of diabetes mellitus and prediabetic states such as metabolic syndrome are 1.5 to 2 times greater in patients with schizophrenia
(Mukherjee et al. 1996) (see Chapter 4, Obesity and Schizophrenia,
and Chapter 5, Glucose Intolerance and Diabetes in Patients With
Schizophrenia). Alcohol and substance use are also more common in
patients with schizophrenia (see Chapter 11, Substance Abuse and
Schizophrenia). Lastly, a direct effect of antipsychotic drugs on myocardial infarction risk has been hypothesized as being mediated by vasospasm or induction of thrombosis. This suggestion emanates from
epidemiological studies showing that patients receiving psychotropic
medications have an increased risk of cardiovascular events; however,
as discussed previously, this might be a confounding by indication
more than a direct effect of antipsychotics.


Medical Illness and Schizophrenia

Hypertension is defined as blood pressure >140/90 mmHg and is a
well-established risk factor for CVD, including stroke and myocardial
infarction (Wang et al. 2007). Patients with schizophrenia do not seem
to be more prone to hypertension, but conflicting data have been reported on this subject. Clozapine has been shown in some studies to increase blood pressure during long-term treatment, but not all studies
have shown similar findings (Henderson et al. 2004; Lund et al. 2001).
However, the increase in weight associated with clozapine and olanzapine increases future hypertension risk, although this effect is often not
seen for years.

Managing Coronary Heart Disease in

Patients With Schizophrenia
Treatment of CVD in schizophrenia is similar to that for the general
population, although patient factors, including lack of motivation, poor
communication skills, and poor medication compliance, make it difficult to achieve the same results. The first step in managing CVD risk is
to obtain baseline data, including history of CVD, family CVD history,
level of physical activity, tobacco and alcohol use, presence of other
physical diseases conferring risk (especially diabetes and hypertension), and current medications. Obtaining a good medical history takes
time, and more than one session might be needed for more severely ill
patients. The clinician should talk to the patients about the nature and
history of prior physical diseases, because some patients will deny or be
reluctant to discuss these issues. Furthermore, a thorough physical examination is needed, including weight, central adiposity (waist circumference or hip:waist ratio), body mass index, and blood pressure.
Laboratory values are also important and must include a fasting glucose and lipid panel, as well as an electrocardiogram in certain circumstances.
The most important aspect of preventive CVD treatment is to reduce
or remove modifiable risk factors by promoting smoking cessation,
increased exercise, and healthier eating habits. The extent to which any
interventions are successful will always depend on the patients motivation; however, one should bear in mind that patients with schizophrenia can achieve meaningful results but may need more motivational support. Ten-year risk of CHD can be estimated from the
Framingham heart study algorithm (see Figure 76). For patients with

Cardiovascular Disease


two or more cardiovascular risk factors, the clinician should estimate

their 10-year risk and provide appropriate interventions.
In addition to lifestyle interventions that require active efforts on the
part of the patient, other interventions require activity by the physician.
These include treating the patients hypertension, recommending aspirin therapy, and carefully considering appropriate antipsychotic medication.

Treating Hypertension
Before treating hypertension, the clinician should ensure that the patient has verifiable hypertension and not hypertension due to whitecoat syndrome. Blood pressure should be measured after having the
patient sit in a chair or lie on a bed for 10 minutes. An end-of-visit blood
pressure reading is likely to be more reflective of the patients true resting blood pressure than that obtained when the anxious patient first enters the office or clinic. The width of the cuff selected should be at least
40% of the circumference of the limb to be used. A common mistake is
to use the wrong cuff size: a small cuff tends to overestimate blood pressure. Also, blood pressure measures should be repeated at least two
times at two different occasions before hypertension is diagnosed.
According to consensus guidelines, blood pressure should be monitored before antipsychotic treatment is initiated, 12 weeks after initiation, and then annually (American Diabetes Association et al. 2004). No
formal guidelines exist for treating hypertension in patients with
schizophrenia, but the clinician should inform patients about nonpharmacological interventions (e.g., exercise, weight loss, restriction of salt
intake) and allow a period of up to 3 months to assess the effects of lifestyle modification. When medications are required, possible additive
hypotensive effects with other medications should be taken into consideration, but otherwise a reasonable approach is to adapt general guidelines for hypertension treatment.

Aspirin reduces myocardial infarction and stroke risk but at the expense of increased risk for gastrointestinal bleeding. The anticoagulant
effect of aspirin is mediated through irreversible blockade of thromboxane A2 formation in platelets. Daily low-dose aspirin (81 mg) should be
considered for patients with a history of CVD or CVD-equivalent disorders (e.g., diabetes mellitus, severe peripheral vascular disease) or for
patients with moderate to high risk. Patients with schizophrenia should
be educated about symptoms of gastritis and gastrointestinal bleeding

Medical Illness and Schizophrenia





















Total cholesterol
mg/dL (mmol/L)





< 160 (< 4.1)

160199 (4.15.2)

200239 (5.36.2)

240279 (6.37.2)


280 ( 7.3)








HDL mg/dL (mmol/L)


60 (1.6)

5059 (1.31.5)

4049 (1.01.2)

< 40 (< 1.0)

Point total

If untreated

If treated

< 120

Systolic BP (mmHg)





10-Year cardiovascular risk (%)

Point total

10-Year cardiovascular risk (%)

















for males.

Framingham 10-year cardiovascular risk estimation charts

Cardiovascular Disease





















Total cholesterol
mg/dL (mmol/L)





< 160 (< 4.1)

160199 (4.15.2)

200239 (5.36.2)

240279 (6.37.2)


280 ( 7.3)











HDL mg/dL (mmol/L)


60 (1.6)

5059 (1.31.5)

4049 (1.01.2)

< 40 (< 1.0)

Point total

Systolic BP (mmHg)

If untreated

If treated

< 120





10-Year cardiovascular risk (%)

Point total






10-Year cardiovascular risk (%)
















for females.

Framingham 10-year cardiovascular risk estimation charts


Medical Illness and Schizophrenia

(e.g., tarry stools, midepigastric pain between meals), and those taking
aspirin should be questioned about gastrointestinal side effects because
their self-report rates are lower than for the general population.

Type of Antipsychotic Medication

The metabolic side effects of atypical antipsychotic drugs differ greatly.
Although olanzapine and clozapine are associated with greater risk for
weight gain than the other atypical antipsychotics, in practice almost
any of these drugs is associated with greater weight gain than placebo.
Switching from a drug with high metabolic liabilities to a lower-risk
drug can lead to substantial weight loss and other metabolic improvement, but in some cases switching to a lower-risk drug is not possible
due to risk of psychotic relapse, especially in patients treated with clozapine. For clozapine-treated patients, standard treatment for dyslipidemia and glucose intolerance should be offered, and some data
indicate that adding a metabolically low-risk atypical antipsychotic
(e.g., aripiprazole) combined with judicious reduction of the clozapine
dosage can be beneficial, although evidence for this combination remains sparse. Treating patients with schizophrenia always requires a
balance between efficacy and side effects. Causing diabetes due to direct drug-related mechanisms or secondary to antipsychotic-induced
weight gain is devastating, but not treating schizophrenia can be fatal.
Any choice of an agent should be balanced by the knowledge that this
is a patient population at high risk for cardiovascular disease and related mortality; therefore, every effort should be made to minimize the
iatrogenic contributions to this risk equation.

Access to Somatic Treatment

Patients with schizophrenia are less likely than the general population
to receive proper treatment for their medical comorbidities (Goldman
1999), a fact related to patient lifestyle variables, lack of health insurance, and systemic health care issues. (See Chapter 1, Improving Physical Health Care for Patients With Serious Mental Illness, for more
discussion of these topics.) Patients with schizophrenia often have negative symptoms that result in less motivation to seek care and decreased
drive to maintain good physical health. Poor communication skills and
decreased pain sensitivity can lead to underreporting of symptoms,
making diagnosis of health conditions more difficult (Singh et al. 2006).
Marchand (1955) reported that up to 85% of psychotic patients were
noted to have experienced an acute myocardial infarction without having any pain, compared with 29% of acute myocardial infarction pa-

Cardiovascular Disease


tients in the Framingham heart study. In acute myocardial infarction

cases presenting without pain, the infarction is often untreated and first
recognized on a subsequent routine electrocardiogram or postmortem
during an autopsy. This lack of pain sensitivity emphasizes the need for
monitoring of cardiovascular risk elements in all patients with schizophrenia, because the actual myocardial infarction event may go unrecognized and untreated.
An important theme throughout this book is the need for psychiatric
providers to assume responsibility for physical health monitoring of
patients with schizophrenia. Many patients with schizophrenia do not
regularly receive primary care, and primary care providers might be reluctant to monitor and treat physical disease due to the barriers posed
by presence of a severe mental illness. As discussed in Chapter 1, Improving Physical Health Care for Patients With Serious Mental Illness,
several possible models have been proposed for integrating psychiatric
care and medical care for this patient population, but the fact remains
that all doctors caring for schizophrenia patients must be alert for signs
of physical disease and be prepared to perform basic physical health
monitoring (Brown 1997).

Key Clinical Points

Cardiovascular heart disease is the most common cause of death in patients with schizophrenia.

Drug-induced torsade de pointes is rare and accounts for a fraction of

sudden-death cases.

Reducing risk factors and treating comorbidities (e.g., obesity, hypertension, dyslipidemia, smoking) are the cornerstones of the treatment
and prevention of cardiovascular disease.

Clozapine is associated with both myocarditis and cardiomyopathy.

Cardiovascular disease attributable to clozapines metabolic properties
is estimated to cause more deaths than agranulocytosis.

Switching antipsychotic medications to a metabolically lower-risk drug

can lead to substantial weight loss and metabolic improvements.


Medical Illness and Schizophrenia

Abdelmawla N, Mitchell AJ: Sudden cardiac death and antipsychotics, part 1: risk
factors and mechanisms. Advances in Psychiatric Treatment 12:3544, 2006
American Diabetes Association, American Psychiatric Association, American
Association of Clinical Endocrinologists, et al: Consensus Development
Conference on Antipsychotic Drugs and Obesity and Diabetes. Diabetes
Care 27:596601, 2004
American Psychiatric Association: Letters outline adverse events. Psychiatric
News 40(8), 2005, p 66
Antzelevitch C, Sicouri S, Di Diego JM: Does Tpeak-Tend provide an index of
transmural dispersion of repolarization? Heart Rhythm 4:11141116, 2007
Borentain S, Millet B, Olie JP, et al: Cardiac risk at the onset of treatment in patients
treated with benzodiazepines and clozapine. Eur Psychiatry 17:419420, 2002
British Heart Foundation: Coronary heart disease statistics 2007. Available at: Accessed October 20, 2008.
Brown S: Excess mortality of schizophrenia: a meta-analysis. Br J Psychiatry
171:502508, 1997
Brown S, Inskip H, Barraclough B: Causes of the excess mortality of schizophrenia. Br J Psychiatry 177:212217, 2000
Cohen H, Loewenthal U, Matar M, et al: [Heart rate variability in schizophrenic
patients treated with antipsychotic agents] (in Hebrew). Harefuah
140:11421147, 2001
Couderc JP, McNitt S, Hyrien O, et al: Improving the detection of subtle I Krinhibition: assessing electrocardiographic abnormalities of repolarization
induced by moxifloxacin. Drug Saf 31:249260, 2008
Danish Secretariat for Clinical Guidelines: Clinical Guideline for Schizophrenia.
Sunhedsstyrelsen (Danish National Board of Health), 2007
De Cicco M, Macor F, Robieux I, et al: Pharmacokinetic and pharmacodynamic
effects of high-dose continuous intravenous verapamil infusion: clinical experience in the intensive care unit. Crit Care Med 27:332339, 1999
Dogan A, Tunc E, Varol E, et al: Comparison of the four formulas of adjusting
QT interval for the heart rate in the middle-aged healthy Turkish men. Ann
Noninvasive Electrocardiol 10:134141, 2005
el-Gamal A, Gallagher D, Nawras A, et al: Effects of obesity on QT, RR, and QTc
intervals. Am J Cardiol 75:956959, 1995
Fontaine KR, Heo M, Harrigan EP, et al: Estimating the consequences of antipsychotic induced weight gain on health and mortality rate. Psychiatry Res
101:277288, 2001
Glassman AH, Bigger JT Jr: Antipsychotic drugs: prolonged QTc interval, torsade de pointes, and sudden death. Am J Psychiatry 158:17741782, 2001
Goff DC, Sullivan LM, McEvoy JP, et al: A comparison of ten-year cardiac risk
estimates in schizophrenia patients from the CATIE study and matched
controls. Schizophr Res 80:4553, 2005

Cardiovascular Disease


Goldenberg I, Moss AJ, Zareba W: QT interval: how to measure it and what is

normal. J Cardiovasc Electrophysiol 17:333336, 2006
Goldman LS: Medical illness in patients with schizophrenia. J Clin Psychiatry
60 (suppl 21):1015, 1999
Gupta A, Lawrence AT, Krishnan K, et al: Current concepts in the mechanisms
and management of drug-induced QT prolongation and torsade de
pointes. Am Heart J 153:891899, 2007
Haddad PM, Anderson IM: Antipsychotic-related QTc prolongation, torsade
de pointes and sudden death. Drugs 62:16491671, 2002
Haverkamp W, Breithardt G, Camm AJ, et al: The potential for QT prolongation
and pro-arrhythmia by non-anti-arrhythmic drugs: clinical and regulatory
implications. Report on a Policy Conference of the European Society of
Cardiology. Cardiovasc Res 47:219233, 2000
Henderson DC, Daley TB, Kunkel L, et al: Clozapine and hypertension: a chart
review of 82 patients. J Clin Psychiatry 65:686689, 2004
Hennekens CH, Hennekens AR, Hollar D, et al: Schizophrenia and increased
risks of cardiovascular disease. Am Heart J 150:11151121, 2005
Hii JT, Wyse DG, Gillis AM, et al: Precordial QT interval dispersion as a marker
of torsade de pointes. Disparate effects of class Ia antiarrhythmic drugs and
amiodarone. Circulation 86:13761382, 1992
Hohnloser SH: Proarrhythmia with class III antiarrhythmic drugs: types, risks,
and management. Am J Cardiol 80:82G89G, 1997
Hohnloser SH, Klingenheben T, Singh BN: Amiodarone-associated proarrhythmic effects: a review with special reference to torsade de pointes tachycardia. Ann Intern Med 121:529535, 1994
Hughes JR, Hatsukami DK, Mitchell JE, et al: Prevalence of smoking among
psychiatric outpatients. Am J Psychiatry 143:993997, 1986
Kang UG, Kwon JS, Ahn YM, et al: Electrocardiographic abnormalities in patients treated with clozapine. J Clin Psychiatry 61:441446, 2000
Kanters JK, Fanoe S, Larsen LA, et al: T wave morphology analysis distinguishes between KvLQT1 and HERG mutations in long QT syndrome.
Heart Rhythm 1:285292, 2004
Kay SE, Doery J, Sholl D: Clozapine associated pericarditis and elevated troponin I. Aust N Z J Psychiatry 36:143144, 2002
Killian JG, Kerr K, Lawrence C, et al: Myocarditis and cardiomyopathy associated with clozapine. Lancet 354:18411845, 1999
Leucht S, Burkard T, Henderson J, et al: Physical illness and schizophrenia: a review of the literature. Acta Psychiatr Scand 116:317333, 2007
Lin CH, Chen MC, Wang SY, et al: Predictive factors for QTc prolongation in
schizophrenic patients taking antipsychotics. J Formos Med Assoc 103:437
441, 2004
Lund BC, Perry PJ, Brooks JM, et al: Clozapine use in patients with schizophrenia and the risk of diabetes, hyperlipidemia, and hypertension: a claimsbased approach. Arch Gen Psychiatry 58:11721176, 2001


Medical Illness and Schizophrenia

Malaspina D, Dalack G, Leitman D, et al: Low heart rate variability is not caused
by typical neuroleptics in schizophrenia patients. CNS Spectr 7:5357, 2002
Malik M: Errors and misconceptions in ECG measurement used for the detection of drug induced QT interval prolongation. J Electrocardiol 37(suppl):
2533, 2004
Malik M, Camm AJ: Evaluation of drug-induced QT interval prolongation: implications for drug approval and labelling. Drug Saf 24:323351, 2001
Marchand WE: Occurrence of painless myocardial infarction in psychotic patients. N Engl J Med 253:5155, 1955
Marder SR, Essock SM, Miller AL, et al: Physical health monitoring of patients
with schizophrenia. Am J Psychiatry 161:13341349, 2004
Mehtonen OP, Aranko K, Malkonen L, et al: A survey of sudden death associated with the use of antipsychotic or antidepressant drugs: 49 cases in Finland. Acta Psychiatr Scand 84:5864, 1991
Merrill DB, Dec GW, Goff DC: Adverse cardiac effects associated with clozapine. J Clin Psychopharmacol 25:3241, 2005
Merrill DB, Ahmari SE, Bradford JM, et al: Myocarditis during clozapine treatment. Am J Psychiatry 163:204208, 2006
Meyer JM, Rao SD, Nielsen JR: Clozapine and dilated cardiomyopathy. Clinical
Schizophrenia and Related Psychoses 1:175180, 2007
Michelsen JW, Meyer JM: Cardiovascular effects of antipsychotics. Expert Rev
Neurother 7:829839, 2007
Modai I, Hirschmann S, Rava A, et al: Sudden death in patients receiving clozapine treatment: a preliminary investigation. J Clin Psychopharmacol
20:325327, 2000
Molnar J, Zhang F, Weiss J, et al: Diurnal pattern of QTc interval: how long is
prolonged? Possible relation to circadian triggers of cardiovascular events.
J Am Coll Cardiol 27:7683, 1996
Moss AJ: Measurement of the QT interval and the risk associated with QTc interval prolongation: a review. Am J Cardiol 72:23B25B, 1993
Mueck-Weymann M, Rechlin T, Ehrengut F, et al: Effects of olanzapine and
clozapine upon pulse rate variability. Depress Anxiety 16:9399, 2002
Mukherjee S, Decina P, Bocola V, et al: Diabetes mellitus in schizophrenic patients. Compr Psychiatry 37:6873, 1996
Olincy A, Young DA, Freedman R: Increased levels of the nicotine metabolite
cotinine in schizophrenic smokers compared to other smokers. Biol Psychiatry 42:15, 1997
Peuskens J, Moore N, Azorin JM, et al: The European sertindole safety and exposure survey: a follow-up study of 8600 patients. Pharmacoepidemiol
Drug Saf 16:804811, 2007
Priori SG, Schwartz PJ, Napolitano C, et al: Risk stratification in the long-QT
syndrome. N Engl J Med 348:18661874, 2003
Rautaharju PM, Zhou SH, Wong S, et al: Sex differences in the evolution of the
electrocardiographic QT interval with age. Can J Cardiol 8:690695, 1992

Cardiovascular Disease


Ray WA, Meredith S, Thapa PB, et al: Antipsychotics and the risk of sudden cardiac death. Arch Gen Psychiatry 58:11611167, 2001
Rechlin T, Claus D, Weis M: Heart rate variability in schizophrenic patients and
changes of autonomic heart rate parameters during treatment with clozapine. Biol Psychiatry 35:888892, 1994
Rechlin T, Beck G, Weis M, et al: Correlation between plasma clozapine concentration and heart rate variability in schizophrenic patients. Psychopharmacology (Berl) 135:338341, 1998
Reid P: Clozapine rechallenge after myocarditis. Aust N Z J Psychiatry 35:249,
Reilly JG, Ayis SA, Ferrier IN, et al: Thioridazine and sudden unexplained
death in psychiatric in-patients. Br J Psychiatry 180:515522, 2002
Rodriguez I, Kilborn MJ, Liu XK, et al: Drug-induced QT prolongation in
women during the menstrual cycle. JAMA 285:13221326, 2001
Rupprecht R, Soyka M, Grohmann R, et al: Considerations in the combination
of clozapine and benzodiazepines [in German]. Nervenarzt 75:857860,
Ruschena D, Mullen PE, Palmer S, et al: Choking deaths: the role of antipsychotic medication. Br J Psychiatry 183:446450, 2003
Sadanaga T, Sadanaga F, Yao H, et al: An evaluation of ECG leads used to assess
QT prolongation. Cardiology 105:149154, 2006
Safferman A, Lieberman JA, Kane JM, et al: Update on the clinical efficacy and
side effects of clozapine. Schizophr Bull 17:247261, 1991
Salih IS, Thanacoody RH, McKay GA, et al: Comparison of the effects of thioridazine and mesoridazine on the QT interval in healthy adults after single
oral doses. Clin Pharmacol Ther 82:548554, 2007
Sicouri S, Antzelevitch C: Sudden cardiac death secondary to antidepressant
and antipsychotic drugs. Expert Opin Drug Saf 7:181194, 2008
Silke B, Campbell C, King DJ: The potential cardiotoxicity of antipsychotic
drugs as assessed by heart rate variability. J Psychopharmacol 16:355360,
Singh MK, Giles LL, Nasrallah HA: Pain insensitivity in schizophrenia: trait or
state marker? J Psychiatr Pract 12:90102, 2006
Struijk JJ, Kanters JK, Andersen MP, et al: Classification of the long-QT syndrome based on discriminant analysis of T-wave morphology. Med Biol
Eng Comput 44:543549, 2006
Taylor DM: Antipsychotics and QT prolongation. Acta Psychiatr Scand 107:85
95, 2003a
Taylor D: Ziprasidone in the management of schizophrenia: the QT interval issue in context. CNS Drugs 17:423430, 2003b
Taylor D, Paton C, Kerwin R: The Maudsley Prescribing Guidelines, 9th Edition. London, Informa Healthcare, 2007
Tham JC, Dickson RA: Clozapine-induced fevers and 1-year clozapine discontinuation rate. J Clin Psychiatry 63:880884, 2002


Medical Illness and Schizophrenia

Titier K, Girodet PO, Verdoux H, et al: Atypical antipsychotics: from potassium

channels to torsade de pointes and sudden death. Drug Saf 28:3551, 2005
Turakhia M, Tseng ZH: Sudden cardiac death: epidemiology, mechanisms, and
therapy. Curr Probl Cardiol 32:501546, 2007
Vaddadi KS, Soosai E, Vaddadi G: Low blood selenium concentrations in
schizophrenic patients on clozapine. Br J Clin Pharmacol 55:307309, 2003
Vieweg WV: Mechanisms and risks of electrocardiographic QT interval prolongation when using antipsychotic drugs. J Clin Psychiatry 63 (suppl 9):18
24, 2002
Walker AM, Lanza LL, Arellano F, Rothman KJ: Mortality in current and former
users of clozapine. Epidemiology 8:671677, 1997
Walker E, Mittal V, Tessner K: Stress and the hypothalamic pituitary adrenal
axis in the developmental course of schizophrenia. Ann Rev Clin Psychol
4:189216, 2008
Wang JG, Li Y, Franklin SS, et al: Prevention of stroke and myocardial infarction
by amlodipine and angiotensin receptor blockers: a quantitative overview.
Hypertension 50:181188, 2007
Wysowski DK, Bacsanyi J: Cisapride and fatal arrhythmia. N Engl J Med 335:290
291, 1996
Yang P, Kanki H, Drolet B, et al: Allelic variants in long-QT disease genes in patients with drug-associated torsades de pointes. Circulation 105:19431948,
Yap YG, Camm AJ: Drug induced QT prolongation and torsades de pointes.
Heart 89:13631372, 2003
Yerrabolu M, Prabhudesai S, Tawam M, et al: Effect of risperidone on QT interval and QT dispersion in the elderly. Heart Dis 2:1012, 2000
Young CR, Bowers MB Jr, Mazure CM: Management of the adverse effects of
clozapine. Schizophr Bull 24:381390, 1998
Yusuf S, Hawken S, Ounpuu S, et al: Effect of potentially modifiable risk factors
associated with myocardial infarction in 52 countries (the INTERHEART
study): case-control study. Lancet 364:937952, 2004
Zuanetti G, Neilson JM, Latini R, et al: Prognostic significance of heart rate variability in post-myocardial infarction patients in the fibrinolytic era: the
GISSI-2 results. Gruppo Italiano per lo Studio della Sopravvivenza nell Infarto Miocardico. Circulation 94:432436, 1996

Treatments for
Weight Management
of Patients With
Rohan Ganguli, M.D., F.R.C.P.C.
Tony Cohn, M.B.Ch.B., M.Sc., F.R.C.P.C.
Guy Faulkner, B.Ed., M.Sc., Ph.D.

An explosion of articles in the last decade has called attention to the

high prevalence of metabolic abnormalities, such as obesity, diabetes,
dyslipidemias, and associated problems, in persons with severe mental
illness. As pointed out in Chapter 2, Excessive Mortality and Morbidity Associated With Schizophrenia, increasingly robust evidence also
indicates that in Europe and North America, people with schizophrenia
and other serious mental illnesses die 2025 years earlier on average
than comparable persons in the general population (Hennekens et al.
2005; McGrath et al. 2008; Osby et al. 2000; Saha et al. 2007). Data also
suggest that the increased prevalence of some risk factors for early mortality from cardiovascular disease and diabetes may have been present
in individuals with psychotic illnesses as long as 100 years ago. For


Medical Illness and Schizophrenia

example, long before the advent of modern antipsychotic medications,

astute clinicians had noted that recovery in psychotic illness is often accompanied by weight gain (Jaspers 1923; Kraepelin 1919).
Much of the recent attention to metabolic issues in serious mental illness is linked to the widespread use of novel antipsychotics. However,
patients and some observant practitioners had already highlighted
medication-associated weight gain as a common side effect of most antipsychotics from the class now called first-generation antipsychotics
(see Ganguli 1999 for a review). For example, Buis (1992) reported that
weight gain was one of patients most frequent complaints about the
side effects of conventional depot antipsychotics. Although some novel
antipsychotics, notably clozapine and olanzapine, are associated with
especially high risk of weight gain (and insulin resistance and diabetes),
all antipsychotics except molindone result in more clinically significant
weight gain than placebo in randomized clinical trials (Casey et al.
Despite the long-standing nature of the problem and its importance
to consumers and to their general health status, few intervention studies have addressed weight gain and its associated risks for people with
schizophrenia. From a practical perspective, the interventions that
have been used for persons with schizophrenia have focused almost
exclusively on weight reduction. Focusing on weight reduction as a
strategy to reduce the risk of cardiovascular disease and diabetes is
completely consistent with the approach being taken in studies funded
by the National Institutes of Health, such as the LOOK-AHEAD study
(Ryan et al. 2003; see also Look AHEAD Research Group 2007), and
with National Heart, Lung, and Blood Institute (1998) recommendations. Thus, in this chapter we focus on studies aimed at weight reduction. With respect to nonpharmacological approaches to weight
reduction, the majority of studies have used techniques based on behavioral therapy principles, so before discussing the approaches, we
briefly review the behavioral techniques common to most weight loss

Principles of Behavioral Approaches

to Weight Loss
As reviewed by Wing (2004), the systematic application of behavioral
therapy techniques to induce weight loss started in the late 1960s. The
early studies tended to treat milder forms of overweightness and obesity, through a focus on stimulus control rather than on specific calorie-

Behavioral Treatments for Weight Management


intake goals or exercise and activity. In reviewing these early programs,

Wadden et al. (2004) noted that weight loss of 45 kg was common in
programs typically lasting 10 weeks. During the 1970s and 1980s, as the
prevalence of obesity grew, treatments became more sophisticated and
comprehensive, and tended to last for longer. By the mid-1990s, average weight loss in behavioral programs, which typically lasted for
6 months, had risen to about 9 kg, or about double what had been reported in the 1960s (Wing 2004). Probably one of the most persuasive
demonstrations of the efficacy of behavioral methods for weight loss
was the Diabetes Prevention Trial, in which over 3,000 overweight or
obese individuals with impaired glucose tolerance were randomized to
a lifestyle intervention aimed at weight loss, metformin, or placebo. Not
only was significant weight reduction achieved in the lifestyle group,
but progression to diabetes was also reduced (Knowler et al. 2002). Furthermore, not only was behavioral treatment twice as effective as metformin in producing weight loss, but the intervention was so effective
in preventing the progression of prediabetes to diabetes that the trial
was stopped prematurely (Knowler et al. 2002). The key components
for the nonpharmacological management of overweight and obesity are
identified in Table 81.
Most approaches to the treatment of obesity are described as behavioral and are based on learning theory (Wing 2004) and the principles
of classical conditioning (Wadden and Foster 2000). In the last 20 years
or so, cognitive approaches have been added to behavioral therapy to
restructure and correct distorted and irrational thoughts that undermine motivation and progress in treatment (Wadden and Foster 2000).
Common components of most behavioral weight reduction programs
include 1) goal-setting, especially establishing realistic short-term goals
(Bandura 1977); 2) self-monitoring (Kazdin 1974) of nutritional intake
and physical activity; 3) a nutritional focus, with teaching and demonstrating of healthy eating habits (Brownell 2004; Wing 1989); and 4)
strategies to increase exercise and decrease sedentary behavior (Jakicic
2002; Jakicic and Gallagher 2003; Jakicic et al. 2004). Stimulus control, by
changing the environment to alter cues so as to increase appropriate
(and decrease inappropriate) eating behavior, was also an early component of behavioral programs (Ferster et al. 1962; Stuart 1967). Problem
solving (DZurilla and Goldfried 1971) is often included to help individuals develop strategies individualized to their own unique situations
(Wing 2004). Once weight loss is achieved, most programs move participants to relapse prevention or weight maintenance regimens
(Brownell et al. 1986; Jeffery and French 1997; Klem et al. 2000; Perri et
al. 2001; Wadden et al. 2004). Because these strategies are often offered

Medical Illness and Schizophrenia



Key components for the management of overweightness

and obesity


Key points for management


Energy intake should be reduced by 5001,000 kcal/day.

Dietary fat should be restricted to <30% of energy intake.
Optimal intakes of carbohydrate and protein have not
been established.


Significant health benefits will occur with 150 minutes of

moderate exercise (at 55%69% of maximum heart rate)
per week. Overweight and obese individuals should
increase moderate exercise to 200300 minutes per


Training should be given in behavioral concepts (e.g.,

problem solving, goal setting, social support). Such
training is associated with improved long-term


Adapted from Jakicic et al. (2001).

as a package, determining which program components are essential to

the efficacy of the treatments is difficult. Recently, cognitive-behavioral
therapy (CBT) has attempted to distinguish itself from behavioral therapy by pointing out that the former specifically includes restructuring
cognitive processes (Cooper and Fairburn 2002).
In the studies of weight loss in individuals with schizophrenia, the
approaches have included one or more elements of common behavioral
approaches, but in many instances, the precise theoretical underpinnings of the program components have not been specified. A selective
review of interventions for weight loss in schizophrenia follows.

Behavioral and Nutritional Interventions

in Schizophrenia
One of the earliest published attempts to assist psychotic patients with
weight loss was carried out in a state hospital in the United States, more
than 40 years ago, by Harmatz and Lapuc (1968). This pioneering study
involved a rigorous behavioral program using negative reinforcement:

Behavioral Treatments for Weight Management


one group of patients lost money if they failed to lose weight. Two comparison groups included a group that discussed weight loss strategies
and provided peer support and a group that received nutritional counseling. The subjects were randomly assigned to one of the three groups.
The group assigned to contingent negative reinforcement had significantly greater mean weight loss than the other two groups (7% of initial body weight over 10 weeks). In a second historically important
study, Rotatori et al. (1980) recruited patients with psychotic illness residing in community-based group homes. Patients were then randomly
assigned to either a 14-week behavioral weight loss group intervention,
closely resembling most modern weight loss programs, or to no intervention (usual care). The group randomly assigned to behavioral
treatment lost significantly more weight (mean 3.3 kg) than the control
group (mean +0.02 kg). The Rotatori et al. study is notable for the use of
a well-developed manual for the delivery of treatment, and the behavioral techniques employed had already been refined in earlier studies of
patients with Down syndrome. We highlight these early studies because they dealt with populations that are still relevant today: severely
mentally ill persons who are in long-stay hospitals as well as those in
community residential services.
These early (first-generation) studies also deserve more recognition
for the following reasons: 1) they demonstrated that patients with psychotic illnesses could participate successfully in nonpharmacological
weight loss interventions, and 2) even though these were pioneering
studies, they were well-designed randomized trials, in the best traditions of generating evidence for clinical practice. Surprisingly, the
second generation of published studies was, for the most part, uncontrolled, or failed to employ random assignment to the intervention or
comparison conditions. Fortunately, the most recent generation (third
generation) of reports is predominantly from randomized controlled
trials, and thus a good evidence base is growing.
Numerous reviews have been published of behavioral and nutritional interventions for weight loss in patients with schizophrenia (e.g.,
Alvarez-Jimnez et al. 2008; Faulkner and Cohn 2006; Faulkner et al.
2007; Ganguli 2007; Loh et al. 2006; Strassnig and Ganguli 2007;
Werneke et al. 2003). All reviews concluded that modest short-term
weight loss is possible in this population. In a recent meta-analysis that
examined randomized controlled trials only, Alvarez-Jimnez et al.
(2008) reported a statistically significant reduction in mean body
weight for those in the nonpharmacological intervention groups compared with those on treatment as usual (weighted mean difference
[WMD]=2.56 kg, 95% CI 3.20 to 1.92, P<0.001) at the end of treat-


Medical Illness and Schizophrenia

ment. The effect was slightly larger but not significant in studies designed to prevent weight gain. Additionally, no statistically significant
or practically important differences were evident between therapeutic
approaches using individual compared with group interventions or using CBT compared with nutritional counseling (Alvarez-Jimnez et al.
Rather than review each of the included randomized clinical trials in
detail, we summarize them in Table 82. Seven of the 12 included trials
investigated CBT strategies (Alvarez-Jimnez et al. 2006; Brar et al.
2005; Jean-Baptiste et al. 2007; Khazaal et al. 2007; Kwon et al. 2006;
McKibbin et al. 2006; Weber and Wyne 2006); three described nutritional counseling interventions (Evans et al. 2005; Littrell et al. 2003;
Scocco et al. 2006); and two combined nutritional and exercise interventions and compared this type of intervention with standard care (Wu et
al. 2007) or metformin alone or in combination (Wu et al. 2008). Six trials
tested group intervention formats (Brar et al. 2005; Jean-Baptiste et al.
2007; Khazaal et al. 2007; Littrell et al. 2003; McKibbon et al. 2006; Weber
and Wyne 2006), and the remaining six examined individual interventions (Alvarez-Jimnez et al. 2006; Evans et al. 2005; Kwon et al. 2006;
Scocco et al. 2006; Wu et al. 2007, 2008). Eight studies were designed to
treat weight gain, whereas the remaining four studies were designed to
prevent weight gain, typically after patients started taking or were
switched to an atypical antipsychotic (Alvarez-Jimnez et al. 2006;
Evans et al. 2005; Littrell et al. 2003; Scocco et al. 2006). Participants were
generally outpatients, except three studies included inpatients (Alvarez-Jimnez et al. 2006; Wu et al. 2007) or a combination (Khazaal et al.
2007). Interventions lasted from 8 to 24 weeks, with an average of approximately 15 weeks. Five studies reported a follow-up assessment of
8 weeks (Littrell et al. 2003), 24 weeks (Evans et al. 2005; Khazaal et al.
2007; Scocco et al. 2006), or 6 months (McKibbin et al. 2006). In the next
paragraphs, we highlight some of these studies that bring up issues we
address in the discussion, including two studies (Jean-Baptiste et al.
2007; Wu et al. 2008) not incorporated in Alvarez-Jimnez et al.s (2008)
In the first of the randomized controlled clinical trials, Litrell et al.
(2003) provided a 16-week psychoeducational program, focusing on
nutrition, exercise, and healthy lifestyle, to patients who had been
switched to olanzapine from other antipsychotics. That all the patients
in this study were on one drug is notable, because in many other studies, treatment effects are potentially confounded by medication effects
and interactions. Litrell et al. reported little weight change in the intervention subjects as opposed to a statistically significant weight gain in

TABLE 82. Randomized, controlled trials of behavioral interventions for weight gain in schizophrenia

Littrell et al. 2003 (N=70)

Prior conventional
antipsychotics, commencing
olanzapine at study entry

16 weekly 1-hour group sessions Intervention: 0.3 kg; usual care: +4.3 kg. Compliance
rate of 92% to program sessions. No dropout rate
for diet and exercise education
vs. usual care; 2-month follow-up reported.

Brar et al. 2005 (N =72)

BMI> 26, switched from
olanzapine to risperidone

Two sessions per week for

6 weeks then one session per
week for 8 weeks of diet and
exercise education vs. usual
care (encouraged to lose

Intervention: 2.0 kg; usual care: 1.1 kg (NS).

5% weight loss in 32.1% of intervention subjects vs.
10.8% in control group. 15/28 patients attended all
20 sessions. 21% dropout rate in treatment group.

Evans et al. 2005 (N = 51)

Commenced olanzapine
within 12 weeks of study

Six 1-hour individual nutrition

education sessions over
3 months (every 2 weeks) vs.
usual care (plus passive
nutrition information)

Intervention: +2.0 kg; usual care: +9.9 kg. Fewer patients

in experimental group (13%) than in control group
(64%) increased initial body weight by more than 7%.
Compliance not reported. 21% dropout rate in
treatment group.

Behavioral Treatments for Weight Management

Outcome at final assessment (weight change),

compliance, and attrition

Authors and participants



TABLE 82. Randomized, controlled trials of behavioral interventions for weight gain in schizophrenia (continued)
Outcome at final assessment (weight change),
compliance, and attrition


Alvarez-Jimnez et al. 2006

(N =61)
First-episode psychosis and
<6 weeks antipsychotic
Treated with: risperidone,
olanzapine, haloperidol

1014 individual sessions

Intervention: +4.1 kg; usual care: +6.9 kg. Fewer patients
(psychoeducation, behavioral
in experimental group (39.3%) than in control group
therapy, dietary counseling,
(78.8%) increased initial body weight by more than 7%.
exercise program) for 3 months Compliance not reported. No dropouts, and all patients
vs. usual care
completed study.

Kwon et al. 2006 (N= 48)

More than 7% body weight
gain on olanzapine

Weekly individual sessions

with dietitian and exercise
coordinator over 12 weeks
(weekly for first 4 weeks, then
every 2 weeks) vs. usual care

McKibbin et al. 2006 (N =64)

Schizophrenia and diabetes

Weekly 90-minute group sessions Intervention: 2.3 kg; usual care: +3.1 kg. 5% weight loss
in 38% of intervention subjects vs. 12% of control group.
for 24 weeks focused on
80% of treatment group attended at least half of
diabetes education, nutrition,
and exercise vs. usual care (plus intervention sessions. No difference in dropout rates
between intervention and usual care groups.
passive information)

Intervention: 3.9 kg; usual care: 1.5 kg. Diet group: all
were over 80% compliant; exercise group: 36% were
over 80% compliant. 33% dropout rate in treatment

Medical Illness and Schizophrenia

Authors and participants

TABLE 82. Randomized, controlled trials of behavioral interventions for weight gain in schizophrenia (continued)

Scocco et al. 2006 (N = 20)

8-week individual dietary
Switched to olanzapine from intervention provided by a
conventional antipsychotics

Outcome at final assessment (weight change),

compliance, and attrition
Intervention: +0.99 kg; control: +2.96 kg. Compliance not
clearly reported. Dropouts: intervention 0/10, control

Weber and Wyne 2006 (N =17) 16 weekly 1-hour group sessions Intervention: 2.5 kg; usual care: 0.6 kg (NS).
BMI> 25, taking secondfor diet and exercise education
Compliance not reported. No dropout in treatment
generation antipsychotics
vs. usual care
Intervention: 2.8 kg; usual care: +2.7 kg. Compliance
Jean-Baptiste et al. 2007 (N =18) Weekly group sessions for
BMI> 30, taking any
16 weeks with psychoeducation, not reported. 14/18 completed intervention.
goal setting, self-monitoring;
$25/week for healthy foods
Khazaal et al. 2007 (N= 61)
>2 kg weight gain over
6 months on any

Weekly 2-hour group cognitivebehavioral therapy plus

psychoeducation for 12 weeks
vs. a single 2-hour nutrition
education session

Intervention: 2.9 kg; usual care: 0.8 kg. At end of

treatment, 16.1% of experimental group vs. 13.3% of
control group had lost 5% or more of initial BMI. This
increased to 22.6% and 16.7% at 12-week follow-up.


Follow-up (12 weeks postintervention):

Intervention : 3.5 kg; usual care: +1.7 kg. Compliance
not reported. Dropouts: intervention 8/31, control 7/30.

Behavioral Treatments for Weight Management

Authors and participants

Outcome at final assessment (weight change),

compliance, and attrition


Wu et al. 2007 (N = 56)

BMI 27, clozapine
(300 mg/day) for at least
1 year

Diet (inpatients) reduced by

Intervention: 4.2 kg; usual care: +1 kg. Compliance not
200300 kcal/day; walking
reported. No dropouts in treatment group.
(level and stairs) for 60 minutes
3 days a week for 6 months

Wu et al. 2008 (N = 128)

First-episode schizophrenia,
gained more than 10% of
body weight

Psychoeducation, diet, and

exercise (lifestyle intervention)
over 12 weeks vs. usual care
(placebo), metformin, and
lifestyle plus metformin

BMI=body mass index; NS=not signficant.

Lifestyle intervention: 1.4 kg; usual care: +3.1 kg;

metformin: +3.2 kg; lifestyle plus metformin: 4.7 kg.
Compliance: Diet 61%84%; exercise 50%60%.
Dropouts: Lifestyle plus metformin 2/32, lifestyle only

Medical Illness and Schizophrenia

Authors and participants



TABLE 82. Randomized, controlled trials of behavioral interventions for weight gain in schizophrenia (continued)

Behavioral Treatments for Weight Management


the control group. Thus, the benefit of the intervention might have been
to prevent weight gain rather than to produce weight loss. Because
olanzapine carries a very high risk of clinically significant weight gain
(Newcomer 2005), and weight gain in adulthood is a powerful predictor of cardiovascular disease (Nanas et al. 1987; Pan et al. 1986), prevention of weight gain is a worthwhile benefit of treatment.
Prevention of weight gain was the focus of a study reported by Alvarez-Jimnez et al. (2006). In an early behavioral intervention group,
1014 individual sessions were completed with a clinical psychologist
within the first 3 months of antipsychotic treatment. The sessions consisted of a weight check, agenda setting, review of self-monitoring
records, and assigning new homework. Modules were available on engagement and assessment, psychoeducation, dietary counseling, exercise, and behavioral therapy. Selection of the intervention strategies
was based on a collaborative formulation after initial assessment between the therapist and the patient. At the end of treatment, significantly fewer patients in the intervention group increased their baseline
weight by more than 7% (39.3% vs. 78.8%). All participants randomized
completed the trial.
Brar et al. (2005) developed a manualized 16-week intervention,
adapted from the study conducted by Rotatori et al. (1980). This study
also controlled for confounding effects of medication by first switching
all subjects to the same antipsychotic (risperidone). This is also one of
the few studies to use blinded raters. Participants enrolled because they
desired to lose weight. Also of note, regular mental health clinicians, as
opposed to specialists in behavioral therapy or nutrition, delivered the
intervention, following the manual. This approach was an attempt to
make the results more likely transferable to routine clinical care. Mean
weight loss in this study was larger in those randomly assigned to the
intervention, but both groups lost weight, and the difference was not
statistically significant. However, the proportion of subjects who lost
5% or more of their baseline body weight was three times larger in subjects randomized to the intervention than in controls (32.1% vs. 10.8%),
and the difference was statistically significant.
Jean-Baptiste et al. (2007) published data from an outpatient study
that used standard behavioral techniques from a widely accepted program (Brownell 2004); however, to the standard nutrition and exercise
program, they added a novel indirect method of food provision. Subjects were given lists of healthy food choices and then, at weekly
group sessions, were reimbursed for the cost of these foods, provided
they had receipts showing that they had purchased these food items in
the previous week. Mean weight loss was statistically significantly


Medical Illness and Schizophrenia

greater in the intervention group than in controls. Jean-Baptiste et al.

(2007) also conducted a 6-month follow-up and reported continuing
weight loss of participants, but the number of subjects was only 12,
leaving some uncertainty about the robustness of the results. Nevertheless, this is a promising rigorous evaluation of a multimodal approach
and needs to be followed up.
Focusing on patients with both schizophrenia and diabetes, McKibbin et al. (2006) recruited patients from board-and-care and community
clubhouse settings. Optimal diabetes management requires active selfmanagement, including weight control and weight loss. Patients with
schizophrenia often have difficulty accessing and participating in comprehensive diabetes management programs. This study illustrates the
effectiveness of a well-constructed intervention geared for patients
with schizophrenia and diabetes. The 24-week training program was
developed in collaboration with a community advisory board that comprised consumers, family members, and community clinicians and consisted of weekly, 90-minute group sessions addressing diabetes
education, nutrition, and exercise. Educational material was adapted
by limiting text, introducing one or two topics per session, providing an
overview and summary of material, and using a teach-and-query training method and mnemonic aids. Concrete behavior change strategies
included weekly weigh-ins, pedometers, healthy food sampling, and
reinforcements (raffle tickets for small health gains). Patients in the intervention group lost a mean of 2.3 kg, compared with a mean weight
gain of 2.7 kg in those receiving usual care (medical follow-up and written diabetes information). The intervention group also showed significant improvements in diabetes knowledge and self-efficacy, as well as
self-reported physical activity, but not in fasting plasma glucose or glycosylated hemoglobin.
In the largest study to date (Wu et al. 2008), 128 first-episode schizophrenia patients were randomly assigned to one of the following:
12 weeks of placebo, 750 mg/day of metformin alone, 750 mg/day of
metformin and lifestyle intervention, or lifestyle intervention only. The
lifestyle intervention included psychoeducational, dietary, and exercise
programs. Psychoeducation focused on the role of eating and activity in
weight management. Dietary intervention followed the American
Heart Association Step II diet, which recommends less than 30% of total
calories from fat (<7% saturated fat and <200 mg of cholesterol), 55%
from carbohydrates, more than 15% from protein daily, and fiber intake
of at least 15 g per 100 kcal. Participants maintained a 3-day food diary
at baseline, and a dietitian reviewed their diets and provided feedback
at follow-up sessions. In the first week, exercise sessions were directed

Behavioral Treatments for Weight Management


by an exercise physiologist, and participants performed exercise (walking or jogging) on a treadmill seven times a week for 30 minutes at each
session. After the first week, exercise was home-based, with recommendations to exercise 30 minutes per day. At the end of the trial, the lifestyle intervention plus metformin group (4.7 kg, 95% CI 5.7 to 3.4)
had results superior to those of either the lifestyle intervention alone
(1.4 kg, 95% CI 2.0 to 0.7) or the metformin alone (3.2 kg, 95% CI
3.9 to 2.5) groups. Metformin alone was more effective in weight loss
and improving insulin sensitivity than lifestyle interventions alone.
Notably, although considerable rigor is required to induce weight
loss, in terms of maintaining a negative energy balance, in all of the
identified randomized clinical trials, dropouts have generally not been
a concern in the treatment arm. This might suggest that patients can be
motivated to initiate and then adhere to a lifestyle intervention for
weight management, at least in the short term. Dropout rates have not
been as high as reported in a recent review (Loh et al. 2006), although
attention must be given to the development of retention strategies to
minimize dropouts (Faulkner and Cohn 2006). Furthermore, no adverse
effects explicitly linked to participating in a lifestyle intervention program have been reported.

The results achieved from weight loss interventions in persons with
schizophrenia (WMD =2.56 kg; Alvarez-Jimnez et al. 2008) are within
the range of those reportedly obtained by commercial weight loss programs in the general population (Heshka et al. 2003), although not as
great as suggested by a meta-analysis of randomized clinical trials of
CBT combined with a diet and exercise intervention (WMD =4.9 kg, CI
7.3 to 2.4; Shaw et al. 2005). However, a reasonable question is
whether such modest improvements actually translate into measurable
improvements in health status or risk of cardiovascular disease and diabetes. The general agreement is that in obese individuals, even a 5%
weight loss can produce measurable health benefits. For example, in the
Finnish Diabetes Prevention Study, a modest weight loss of 4.8% of initial body weight was associated with a 58% reduction in the risk of developing diabetes over the following 3 years (Tuomilehto et al. 2001).
As little as 34 kg of weight loss over 3 years results in clinically significant reductions in systolic and diastolic blood pressure (Mertens and
Van Gaal 2000). Also, accumulating evidence from prospective observational studies indicates that increasing physical activity is effective in
improving the health profile of individuals who are overweight and


Medical Illness and Schizophrenia

obese (e.g., Lee et al. 1999). Weight loss might be considered a primary
goal, but clinicians should keep in mind that small, sustained positive
changes in physical activity and dietary intake may be associated with
significant health benefits irrespective of weight loss per se. With very
few exceptions, even randomized controlled interventions have rarely
followed patients for longer than 6 months. Weight loss is challenging,
but even more challenging is weight loss maintenance (Wing et al.
2006). Obesity is unquestionably a chronic condition, and it is likely that
long-term success may require some form of maintenance treatment.
Hence, the real health benefits of weight loss interventions, even for
those who respond to interventions, will be known only when data
from longer studies become available. Fortunately, several ongoing
clinical trials have up to 2 years of follow-up in their designs. Overall,
interventions will probably need to set realistic goals, be highly structured, provide early and intensive support initially, and offer reduced
but continued support over time if not indefinitely (Faulkner and Cohn
Most currently published research has evaluated pharmacological
and nonpharmacological treatments for weight loss in separate studies
(pharmacological treatments are reviewed in Chapter 4, Obesity and
Schizophrenia). However, evidence is accumulating that combining
behavioral and pharmacological weight loss interventions can be more
effective than either approach alone (Wadden et al. 2005). One study has
been published that demonstrates the greater effectiveness of combined
weight loss interventions specifically in patients with schizophrenia
(Wu et al. 2008). However, given concerns about potential polypharmacy, the demands of adding further medication to an existing medical
regimen, and the cost of medication, we suggest that adjunctive pharmacotherapy for weight loss be reserved for patients who do not respond adequately to lifestyle interventions alone (Faulkner et al. 2007).
Further studies evaluating the combination of behavioral and pharmacological weight loss therapies are required before routinely recommending such a dual approach in clinical practice. Switching a patient
to an antipsychotic medication with low liability for weight gain has
also emerged as an effective strategy for weight loss and metabolic benefit, particularly when the increase in weight was clearly associated
with prior antipsychotic treatment (Weiden 2007; Weiden and Buckley
2007) (see Chapter 4). No studies to date have investigated the combination of antipsychotic switching and behavioral strategies for weight loss.

Behavioral Treatments for Weight Management


Given that the current system of care for persons with severe mental
illness is routinely described as underfunded and overburdened (Frank
and Glied 2006), economic considerations may well determine which
interventions for weight loss, if any, will make it to the front lines of
community mental health. Thus, cost-benefit analyses should be included in the evaluation of proposed interventions. At this point, such
analyses are almost entirely missing from the evidence base on this subject. Even without sophisticated economic analyses, clinicians should
be able to evaluate the potential benefit of investing time and resources
in particular interventions if the published results systematically reported the proportions of subjects who benefited and, preferably, the
number needed to treat for each threshold of response. Unfortunately,
most studies already published limit the results to reporting mean
changes in body weight.

On the basis of existing studies, we can conclude that persons with
schizophrenia want to and will participate in behavioral weight loss
interventions. For individuals who are unmotivated or difficult to engage, consideration could be given to broader environmental interventions that aim to shape the environment in ways that are conducive to
encouraging greater physical activity while restricting energy intake
(Gorczynski et al. 2008). Taken together, the evidence from controlled
trials indicates that patients who do participate in weight loss interventions increase their chances of losing weight. The results of simple and
practical interventions are modest but clinically meaningful. The data
on long-term maintenance of weight loss is essentially lacking, but
some ongoing studies will provide data in the next few years. The data
on preventing weight gain in persons with schizophrenia is developing
and looking positive. With these observations in mind, standard behavioral weight loss interventions should be widely and routinely offered
to patients with schizophrenia who are overweight or obese. In addition, discussion about the risk of weight gain and monitoring of weight
should be routinely offered to all patients with schizophrenia. Given
that the current trend is for the rates of obesity to continue to increase,
research into enhancing the effectiveness of current interventions and
the development of new approaches to weight loss need to be urgently


Medical Illness and Schizophrenia

Key Clinical Points

Education about the health hazards of being overweight or becoming

overweight should be included in the psychoeducational interventions
offered to persons with schizophrenia, along with simple advice about
health nutrition and exercise measures.

Regular measurement of body weight should be part of routine care in

mental health settings, and patients should be given feedback on their
own weight regularly. Patients should also be encouraged to weigh

Patients who ask for active interventions to help them lose weight
should be offered group or individual interventions, preferably within
the mental health treatment setting.

Referral to specialized programs, including the full range of options for

severe and/or treatment-resistant obesity, should be pursued for patients who are in need of these services.

Alvarez-Jimnez M, Gonzlez-Blanch C, Vzquez-Barquero JL, et al: Attenuation of antipsychotic-induced weight gain with early behavioral intervention in drug-naive first-episode psychosis patients: a randomized
controlled trial. J Clin Psychiatry 67:12531260, 2006
Alvarez-Jimnez M, Hetrick SE, Gonzlez-Blanch C, et al: Non-pharmacological management of antipsychotic-induced weight gain: systematic review
and meta-analysis of randomised controlled trials. Br J Psychiatry 193:101
107, 2008
Bandura A: A Social Learning Theory. New York, General Learning Press, 1977
Brar JS, Ganguli R, Pandina G, et al: Effects of behavioral therapy on weight loss
in overweight and obese patients with schizophrenia or schizoaffective disorder. J Clin Psychiatry 66:205212, 2005
Brownell KD: The Learn Program for Weight Management, 10th Edition. Dallas, TX, American Health, 2004
Brownell KD, Marlatt GA, Lichtenstein E, et al: Understanding and preventing
relapse. Am Psychol 41:765782, 1986
Buis W: Patients opinions concerning side effects of depot neuroleptics. Am J
Psychiatry 149:844845, 1992
Casey DE, Haupt DW, Newcomer JW, et al: Antipsychotic-induced weight gain
and metabolic abnormalities: implications for increased mortality in patients with schizophrenia. J Clin Psychiatry 65 (suppl 7):418, 2004

Behavioral Treatments for Weight Management


Cooper Z, Fairburn CG: Cognitive-behavioral treatment of obesity, in Handbook of Obesity Treatment. Edited by Wadden TA, Stunkard AJ. New
York, Guilford, 2002
DZurilla TJ, Goldfried MR: Problem solving and behavior modification. J Abnorm Psychol 78:107126, 1971
Evans S, Newton R, Higgins S: Nutritional intervention to prevent weight gain
in patients commenced on olanzapine: a randomized controlled trial: Aust
N Z J Psychiatry 39:479486, 2005
Faulkner G, Cohn TA: Pharmacologic and nonpharmacologic strategies for
weight gain and metabolic disturbance in patients treated with antipsychotic medications. Can J Psychiatry 51:502511, 2006
Faulkner G, Cohn T, Remington G: Interventions to reduce weight gain in
schizophrenia. Cochrane Database Syst Issue 1, Art. No.: CD005148, DOI:
10.1002/14651858.CD005148.pub2, 2007
Ferster CB, Nurnberger JI, Levitt EE: The control of eating. Journal of Mathetics
1:87109, 1962
Frank RG, Glied SA: Better But Not Well: Mental Health Policy in the United
States Since 1950. Baltimore, Johns Hopkins Community Press, 2006
Ganguli R: Weight gain associated with antipsychotic drugs. J Clin Psychiatry
60 (suppl 21):2024, 1999
Ganguli R: Behavioral therapy for weight loss in patients with schizophrenia.
J Clin Psychiatry 68 (suppl 4):1925, 2007
Gorczynski P, Faulkner G, Zeglen L, et al: Dissecting the obesogenic environment of a psychiatric hospital. J Sport Exerc Psychol 30:S172, 2008
Harmatz MG, Lapuc P: Behavioral modification of overeating in a psychiatric
population. J Consult Clin Psychol 32:583587, 1968
Hennekens CH, Hennekens AR, Hollar D, et al: Schizophrenia and increased
risks of cardiovascular disease. Am Heart J 150:11151121, 2005
Heshka S, Anderson JW, Atkinson RL, et al: Weight loss with self-help compared with a structured commercial program: a randomized trial. JAMA
289:17921798, 2003
Jakicic JM: The role of physical activity in prevention and treatment of body
weight gain in adults. J Nutr 132:3826S3829S, 2002
Jakicic JM, Gallagher KI: Exercise considerations for the sedentary, overweight
adult. Exerc Sport Sci Rev 31:9195, 2003
Jakicic JM, Clark K, Coleman E, et al: American College of Sports Medicine position
stand: appropriate intervention strategies for weight loss and prevention of
weight regain for adults. Med Sci Sports Exerc 33:21452156, 2001
Jakicic JM, Marcus BH, Gallagher KI, et al: Effect of exercise duration and intensity on weight loss in overweight, sedentary women: a randomized trial.
JAMA 290:13231330, 2004
Jaspers K: General Psychopathology. Chicago, IL, University of Chicago Press, 1923
Jean-Baptiste M, Tek C, Liskov E, et al: A pilot study of a weight management
program with food provision in schizophrenia. Schizophr Res 96:198205,


Medical Illness and Schizophrenia

Jeffery R, French WSA: Preventing weight gain in adults: design, methods and
one year results from the Pound of Prevention study. Int J Obes Relat
Metab Disord 21:457464, 1997
Kazdin AE: Self-monitoring and behavior change. Monterey, CA, Brooks/Cole,
Khazaal Y, Fresard E, Rabia S, et al: Cognitive behavioural therapy for weight
gain associated with antipsychotic drugs. Schizophr Res 91:169177, 2007
Klem ML, Viteri JL, Wing RR: Primary prevention of weight gain for women
aged 2534: the acceptability of treatment formats. Int J Obes Relat Metab
Disord 24:219225, 2000
Knowler WC, Barrett-Connor E, Fowler SE, et al; Diabetes Prevention Program
Research Group: Reduction in the incidence of type 2 diabetes with lifestyle
intervention or metformin. N Engl J Med 346:393403, 2002
Kraepelin E: Dementia Praecox and Paraphrenia. Edinburgh, Scotland: E and S
Livingstone, 1919
Kwon J, Choi JS, Bahk WM, et al: Weight management program for treatmentemergent weight gain in olanzapine-treated patients with schizophrenia or
schizoaffective disorder: a 12-week randomized controlled clinical trial.
J Clin Psychiatry 67:547553, 2006
Lee CD, Blair SN, Jackson AS: Cardiorespiratory fitness, body composition, and
all-cause mortality in men. Am J Clin Nutr 69:373380, 1999
Littrell KH, Hilligoss NM, Kirshner CD, et al: The effects of an educational intervention on antipsychotic-induced weight gain. J Nurs Scholarsh 35:237
241, 2003
Loh C, Meyer JM, Leckband SG: A comprehensive review of behavioral interventions for weight management in schizophrenia. Ann Clin Psychiatry
18:2331, 2006
Look AHEAD Research Group, Pi-Sunyer X, Blackburn G, et al: Reduction in
weight and cardiovascular disease risk factors in individuals with type 2
diabetes: one-year results of the Look AHEAD trial. Diabetes Care 30:1374
1383, 2007
McGrath J, Saha S, Chant D, et al: Schizophrenia: a concise overview of incidence, prevalence, and mortality. Epidemiol Rev 30:6776, 2008
McKibbin CL, Patterson TL, Norman G, et al: A lifestyle intervention for older
schizophrenia patients with diabetes mellitus: a randomized controlled
trial. Schizophr Res 86:3644, 2006
Mertens IL, Van Gaal LF: Overweight, obesity, and blood pressure: the effects
of modest weight reduction. Obes Res 8:270278, 2000
Nanas S, Pan WH, Stamler J, et al: The role of relative weight in the positive association between age and serum cholesterol in men and women. J Chronic
Dis 40:887892, 1987
National Heart, Lung, and Blood Institute: Clinical guidelines on the identification,
evaluation, and treatment of overweight and obesity in adults. 1998. Available
at: Accessed
October 22, 2008.

Behavioral Treatments for Weight Management


Newcomer JW: Second-generation (atypical) antipsychotics and metabolic effects:

a comprehensive literature review. CNS Drugs 19 (suppl 1):193, 2005
Osby U, Correia N, Brandt L: Time trends in schizophrenia mortality in Stockholm County, Sweden: cohort study. BMJ 321:483484, 2000
Pan WH, Nanas S, Dyer A, et al: The role of weight in the positive association
between age and blood pressure. Am J Epidemiol 124:612623, 1986
Perri MG, Nezu AM, McKelvey WF, et al: Relapse prevention training and
problem-solving therapy in the long-term management of obesity. J Consult Clin Psychol 69:722726, 2001
Rotatori AF, Fox R, Wicks A: Weight loss with psychiatric residents in a behavioral self control program. Psychol Rep 46:483486, 1980
Ryan DH, Espeland MA, Foster GD, et al: Look AHEAD (Action for Health in
Diabetes): design and methods for a clinical trial of weight loss for the prevention of cardiovascular disease in type 2 diabetes. Control Clin Trials
24:610628, 2003
Saha S, Chant D, McGrath J: A systematic review of mortality in schizophrenia:
is the differential mortality gap worsening over time? Arch Gen Psychiatry
64:11231131, 2007
Scocco P, Longo R, Caon F: Weight change in treatment with olanzapine and a
psychoeducational approach. Eat Behav 7:115124, 2006
Shaw K, ORourke P, Del Mar C, et al: Psychological interventions for overweight or obesity. Cochrane Database Syst Rev Issue 2, Art. No.: CD003818.
DOI: 10.1002/14651858.CD003818.pub2, 2005
Strassnig M, Ganguli R: Weight loss interventions for patients with schizophrenia. Clinical Schizophrenia and Related Psychoses 1:4253, 2007
Stuart RB: Behavioral control of overeating. Behav Res Ther 5:357365, 1967
Tuomilehto J, Lindstrom J: Prevention of type 2 diabetes mellitus by changes in
lifestyle among subjects with impaired glucose tolerance. N Engl J Med
344:13431350, 2001
Wadden TA, Foster GD: Behavioral treatments of obesity. Med Clin North Am
84:441461, 2000
Wadden TA, Butryn ML, Byrne KJ: Efficacy of lifestyle modification for longterm weight control. Obes Res 12(suppl):151S162S, 2004
Wadden TA, Berkowitz RI, Womble LG, et al: Randomized trial of lifestyle
modification and pharmacotherapy for obesity. N Engl J Med 353:2111
2120, 2005
Weber M, Wyne K: A cognitive/behavioral group intervention for weight loss
in patients treated with atypical antipsychotics. Schizophr Res 83:95101,
Weiden PJ: Switching antipsychotics as a treatment strategy for antipsychoticinduced weight gain and dyslipidemia. J Clin Psychiatry 68 (suppl 4):34
39, 2007
Weiden PJ, Buckley PF: Reducing the burden of side effects during long-term
antipsychotic therapy: the role of switching medications. J Clin Psychiatry 68 (suppl 6):1423, 2007


Medical Illness and Schizophrenia

Werneke U, Taylor D, Sanders TA, et al: Behavioural management of antipsychotic-induced weight gain: a review. Acta Psychiatr Scand 108:252259,
Wing RR: Behavioral strategies for weight reduction in obese type II diabetic
patients. Diabetes Care 12:139144, 1989
Wing RR: Behavioral approaches to the treatment of obesity, in Handbook of
Obesity. Edited by Bray GA, Bouchard C. New York, Marcel Dekker, 2004,
pp 227248
Wing RR, Tate DF, Gorin AA, et al: A self-regulation program for maintenance
of weight loss. N Engl J Med 355:15631571, 2006
Wu M-K, Wang C-K, Bai YM, et al: Outcomes of obese, clozapine-treated inpatients with schizophrenia placed on a six-month diet and physical activity
program. Psychiatr Serv 58:544550, 2007
Wu R-R, Zhao J-P, Jin H, et al: Lifestyle intervention and metformin for treatment of antipsychotic-induced weight gain. JAMA 299:185193, 2008

Nicotine and Tobacco
Use in Patients With
Andrea H. Weinberger, Ph.D.
Tony P. George, M.D., F.R.C.P.C.

Epidemiology and Significance

of the Problem
Patients with schizophrenia have a high prevalence of cigarette smoking based on clinical (58%88%) and population-based (45%) samples
(Kalman et al. 2005) compared to the general population rate of approximately 20.5% (Giovino 2007). Also, smoking cessation rates are lower
in patients with schizophrenia than in nonpsychiatric control smokers

We thank Erin L. Reutenauer, B.A., for assistance with this work. This work was
supported by National Institute on Drug Abuse grants K02-DA16611, R01DA13672, R01-DA14039 (to Dr. George), and K12-DA000167 (to Dr. Weinberger); National Alliance for Research in Schizophrenia and Depression
(NARSAD) Young Investigator (to Dr. Weinberger) and Independent Investigator (to Dr. George) Awards; and the University Chair in Addiction Psychiatry
at the University of Toronto (to Dr. George).



Medical Illness and Schizophrenia

(Kalman et al. 2005). Despite the fact that patients with schizophrenia
constitute only about 1% of the general population, the medical and
economic burden of cigarette smoking on mentally ill persons is enormous (Kalman et al. 2005). For example, the rates of smoking-related illnesses such as cardiovascular disease and certain cancers appear to be
higher in patients with schizophrenia than in the general population
(Capasso et al. 2008). Hence, an urgent need exists to address tobacco
addiction in this population through the development of effective treatments, both pharmacological and behavioral. A better understanding
of the biology of tobacco addiction in patients with schizophrenia may
assist in the development of better treatments for tobacco use in this
vulnerable population.

Pathophysiology of Schizophrenia:
Relationship to Nicotinic, Dopaminergic,
and Glutamatergic Effects
Over the past 20 years, scientific understanding of both the neurobiology of schizophrenia and nicotine addiction has been increasing. For
the purposes of this discussion, nicotine is assumed to be the active ingredient in tobacco and cigarette smoking that exerts psychopharmacological effects, although other components of tobacco smoke may be
active in this respect (George and OMalley 2004). There are three possible reasons for the high comorbid rates of nicotine addiction and
schizophrenia: 1) self-medication of clinical and cognitive deficits associated with schizophrenia by tobacco use; 2) abnormalities in brain reward pathways in patients with schizophrenia that make these patients
vulnerable to tobacco (and other drug) use; 3) common genetic and environmental factors that are independently associated with both smoking and schizophrenia. We briefly describe next the pharmacological
effects of nicotine, including its effects on neurotransmitter systems
critical to the neurobiology of schizophrenia, and how such effects may
link nicotine addiction with schizophrenia.
Nicotine alters the function of neurotransmitter systems implicated
in the pathogenesis of major psychiatric disorders, including dopamine, norepinephrine, serotonin, glutamate, gamma-aminobutyric acid
(GABA), and endogenous opioid peptides (Dani and Bertrand 2007).
Nicotines receptor in the brain is the nicotinic acetylcholine receptor
(nAChR), where stimulation of presynaptic nAChRs on neurons increases transmitter release and metabolism (Dani and Bertrand 2007;
Picciotto 2003). The two general families of central nAChRs are high-

Nicotine and Tobacco Use


affinity receptors (2 subunitcontaining nAChRs, which exist in a heteropentameric configuration of subunits combined with subunits)
and low-affinity receptors (7 subunitcontaining nAChR homopentameric complexes) (see Dani and Bertrand 2007 for review). Both highand low-affinity nAChRs appear to be present on mesocorticolimbic
dopamine neurons (Wooltorton et al. 2003), and 7 nAChRs are enriched in the hippocampus and cortex and appear to facilitate information processing and sensory integration (Leonard and Bertrand 2001).
Chronic administration of nicotine, unlike most agonists, leads to desensitization and inactivation of nAChRs, with subsequent upregulation of nAChR sites, a process that might explain why many smokers
report that the most satisfying cigarette of the day is the first one in the
morning, at which time upregulated nAChRs are resensitized. Mesolimbic dopamine (reward pathway) neurons possess presynaptic
nAChRs and may be of particular importance in mediating the rewarding effects of nicotine through projections from the ventral tegmental
area in the midbrain to anterior forebrain structures such as the nucleus
accumbens and cingulate cortex (Dani and Bertrand 2007).
Domino et al. (2004) proposed that individuals with schizophrenia
have a relative hyperfunction of subcortical dopamine systems and a
relative hypofunction of prefrontal cortex dopamine systems. It is a
well-known fact that chemical or physical lesioning of the prefrontal
cortex produces hyperfunctional mesolimbic dopamine neurons. Preclinical studies in nonhuman primates suggest that chronic treatment
with the N-methyl-D-aspartate receptor antagonist MK-801 produces a
depletion of prefrontal cortex dopamine levels, which is accompanied
by an increase in D 1 receptor binding and an impairment in working
memory performance (Tsukada et al. 2005b). Interestingly, acute intravenous administration of nicotine normalizes prefrontal cortex dopamine levels, and D1 receptor binding, and improves working memory
performance in these MK-801 exposed monkeys (Tsukada et al. 2005a).
This animal model of schizophrenia may provide an outstanding opportunity to study the interactive effects of nicotine on the putative
pathophysiological processes associated with schizophrenic disorders
and their frequent comorbidity with tobacco dependence (Domino et al.
Converging lines of evidence suggest that dysregulation of both
high- and low-affinity nAChRs occur in the brains of individuals with
schizophrenia. Evidence indicates that high-affinity nAChR expression
is reduced in postmortem brains of patients with schizophrenia
(Leonard et al. 2000) and that the expected upregulation of nAChRs
with tobacco exposure is blunted in patients with schizophrenia com-


Medical Illness and Schizophrenia

pared with control smokers and nonsmokers; however, most of these

data come from cross-sectional comparisons, and within-subject
changes in nAChR levels have not been studied either with respect to
smoking initiation or smoking cessation.
Schizophrenia is associated with a number of neurocognitive deficits, including attention, verbal learning and memory, spatial working
memory, and executive function (Green et al. 2004), and nicotine is
thought to have procognitive effects (Sacco et al. 2005). Acetylcholine is
associated with arousal, learning, and memory, and cholinergic inhibitors are used clinically to enhance memory (Levin et al. 2006; Newhouse et al. 2004; Sacco et al. 2004). The use of selective nAChR agonists
and partial agonists to treat cognitive deficits in schizophrenia is
thought to be the most appealing area of treatment possibilities at this
time. The Measurement and Treatment Research to Improve Cognition
in Schizophrenia consortium has identified several classes of compounds that may be effective for this purpose; topping this list are 7
nicotinic receptor partial agonists (Marder 2006; Tamminga 2006). One
promising medication is the 7 agent dimethoxybenzylidene anabaseine (DXMB-A), which is likely to alter deficits of P50 auditory
evoked potential inhibition and smooth pursuit eye movements in this
population (Olincy et al. 2006). In addition, varenicline (Chantix), as an
42 nAChR-selective partial agonist, might be considered for its potential effects on visuospatial working memory, attention, and prepulse inhibition. Other novel medications used off-label, such as galantamine
hydrobromide (Razadyne), an nAChR allosteric modulator (Coyle and
Kershaw 2001), are also currently being used to study nAChR mechanisms relevant to neurocognition in patients with schizophrenia and
other neuropsychiatric disorders. The development of other novel selective nAChR agents will allow more careful exploration of the therapeutic potential of nicotinic agents for the treatment of neurocognitive
dysfunction associated with schizophrenia.

Pharmacokinetic Implications of Smoking

for Psychotropic Drug Use in Schizophrenia
Strong evidence indicates that tobacco smoking induces the liver cytochrome P450 1A2 (CYP1A2) enzyme system, a major route for the metabolism of antipsychotic drugs such as olanzapine and clozapine
(Kroon 2007). Accordingly, smoking cessation would be expected to
lead to increases in plasma concentrations of antipsychotic drugs
metabolized by the CYP1A2 system, a finding demonstrated repeatedly

Nicotine and Tobacco Use


in both prospective and retrospective studies (Kroon 2007). Such an increase in circulating levels would be expected to increase the likelihood
of extrapyramidal reactions and other antipsychotic drug side effects.
Although no smoking cessation study of patients with schizophrenia to
date has prospectively measured antipsychotic plasma levels before
and after smoking cessation, several case reports and case series suggest increases in medication levels with smoking cessation. In one of the
larger case series, Meyer (2001) reported on serial clozapine levels measured in 11 patients with schizophrenia or schizoaffective disorder
treated with stable dosages in a state hospital before and after a hospital-wide ban on smoking. A mean increase of 57.4% was noted in these
clozapine-treated patients who quit smoking; one patient in particular
had an increase in his serum clozapine level to over 3,000 ng/ml that
was associated with aspiration. In the few published controlled smoking cessation trials in this population (Addington et al. 1998; George et
al. 2000a), no significant increases in medication side effects have been
noted in patients who quit smoking, including those treated with medications known to be metabolized by CYP1A2, but clearly further study
is warranted.

Effects of Smoking on Clinical and

Cognitive Deficits Associated With
Several cross-sectional studies have examined the effects of cigarette
smoking on psychotic symptoms in patients with schizophrenia, with
mixed results (e.g., Barnes et al. 2006; Goff et al. 1992; Patkar et al. 2002;
Tang et al. 2007; Ziedonis et al. 1994). Goff et al. (1992) found that compared with nonsmokers with schizophrenia, smokers with schizophrenia had higher Brief Psychiatric Rating Scale (BPRS) total scores and
higher subscale scores for both positive and negative symptoms. Tang
et al. (2007) found no differences in overall BPRS scores between smoking and nonsmoking patients with schizophrenia but did find that
smokers had a lower score on the BPRS Depressive and Anxious Symptoms subscale. Ziedonis et al. (1994) found that patients with schizophrenia who smoked had increased positive symptom scores and
reduced negative symptom scores compared with nonsmoking patients
with schizophrenia, with heavy smokers having the highest positive
and lowest negative symptom scores. Patkar et al. (2002) found that
higher levels of nicotine dependence were associated with higher


Medical Illness and Schizophrenia

negative symptoms but not with positive symptoms. A study by our

group found that patients with schizophrenia who were former smokers had higher negative symptoms than current smokers, after adjusting for differences in age, depressive symptoms, and education (George
et al. 2002a). Two recent reports found no differences in positive or negative symptoms of schizophrenia when comparing patients who were
smokers with those who did not smoke (Barnes et al. 2006; Tang et al.
2007). Clearly, the interpretation of data from cross-sectional studies is
confounded by methodological differences.
Few direct studies have been done of the effects of smoking or nicotine
administration on clinical symptoms in patients with schizophrenia. In
contrast to results from cross-sectional studies, controlled laboratory
studies of smoking abstinence (George et al. 2002a) and nicotine patch
administration (Dalack et al. 1999) have not shown significant effects on
the clinical symptoms of schizophrenia. One study (Smith et al. 2002)
found that although high-nicotine cigarettes led to a greater decrease in
negative symptoms than denicotinized cigarettes, the type of cigarette
did not affect positive symptoms, anxiety, or depression. In this study,
nicotine nasal spray did not alter clinical symptoms. Furthermore, data
from smoking cessation trials (Addington et al. 1998; Baker et al. 2006;
Evins et al. 2005, 2007; Gallagher et al. 2007; George et al. 2000a, 2008), all
of which used the nicotine patch, found no evidence for significant
changes in psychotic symptoms with smoking abstinence in patients
with schizophrenia. Thus, the effects of cigarette smoking and smoking
abstinence on schizophrenia symptoms are not clear. Perhaps some trait
differences in psychotic symptoms in smokers versus nonsmokers with
schizophrenia (e.g., more refractory symptoms in nonsmokers) might explain these findings, independent of smoking status (George et al. 2002a).
Several human laboratory studies have suggested that patients with
schizophrenia possess deficits in auditory information processing (P50
event-related potentials), which can be transiently normalized by cigarette smoking or administration of nicotine gum (Adler et al. 1998).
Similarly, Griffith et al. (1998) reported a smaller but still significant improvement in P50 auditory gating after providing a nicotine patch to
smokers with schizophrenia following a 2-hour period of abstinence
from smoking. Leonard et al. (2002) found that these P50 response deficits may be linked to a locus on chromosome 15 (q14) near the coding
region for the 7 nicotinic acetylcholine receptor. This subtype of
nAChR has been strongly implicated in P50 responses, leading to speculation that if some schizophrenia patients possess defective 7
nAChR-mediated neurotransmission, they may smoke heavily to overcome the related neurophysiological deficits.

Nicotine and Tobacco Use


In addition, patients with schizophrenia have abnormalities in another auditory information processing response known as prepulse inhibition of the acoustic startle reflex. The neural substrates mediating
this response appear to be distinct from those mediating P50 responses
(Swerdlow et al. 1992). Nicotine seems to improve performance on
prepulse inhibition for smokers with schizophrenia. Kumari et al.
(2001) found that when male patients with schizophrenia were acutely
smoking, their prepulse inhibition was higher than that in smokingdeprived or nonsmoking patients with schizophrenia, whereas Postma
et al. (2006) found that subcutaneous nicotine improves prepulse inhibition in smokers with schizophrenia. Moreover, George et al. (2006)
showed that deprived smokers with schizophrenia demonstrated reduced prepulse inhibition but that baseline performance was reinstated
upon resumption of smoking. However, the nicotinic antagonist
mecamylamine dose-dependently blocked reinstatement in the schizophrenia group, whereas no effect was found in nonpsychiatric control
subjects, suggesting that stimulation of central nAChRs mediates cigarette smokinginduced improvement of prepulse inhibition in patients
with schizophrenia. Thus, patients with schizophrenia may be using
cigarette smoking to ameliorate defects in cognitive function, further
supporting the self-medication hypothesis of cigarette smoking in
A growing body of research has examined cognitive function in
patients with schizophrenia and nonpsychiatric control subjects as
a function of smoking status. In addition to P50 gating and prepulse
inhibition, nicotine appears to enhance areas of cognition including
working memory and sustained attention. Smokers with schizophrenia
demonstrate a deficit in visual-spatial working memory function after
overnight abstinence, with an immediate improvement in performance
with resumption of cigarette smoking (Sacco et al. 2005) or after administration of nicotine nasal spray (Smith et al. 2006). Furthermore, such
persistent effects on visual-spatial working memory are observed with
longer-term abstinence. Smokers who remained abstinent from cigarettes over a 10-week period demonstrated enduring decrements in
visual-spatial working memory (George et al. 2002a). Similar cognitive
enhancement is achieved with nicotine gum administration, as seen
using an auditory N-back task with a high verbal working memory
demand (Jacobsen et al. 2004), and with nicotine patch, nasal spray, or
cigarettes on measures of sustained attention using the Conners Continuous Performance Test (Sacco et al. 2005). In summary, nicotine
administration appears to enhance several areas of cognitive function
in patients with schizophrenia, including working memory, specifically


Medical Illness and Schizophrenia

spatial working memory; sustained attention; P50 gating; smooth pursuit eye movement; and prepulse inhibition.

Effects of Atypical Versus Typical

Antipsychotic Drugs on Smoking
in Schizophrenia
Cross-sectional research has shown lower levels of smoking in patients
taking atypical antipsychotic medications than in those taking typical
antipsychotics (e.g., Barnes et al. 2006). Our group has found that
switching smokers with schizophrenia from typical antipsychotic
agents to clozapine leads to reductions in self-reported cigarette smoking, especially in heavier smokers (George et al. 1995). Similar findings
were reported by McEvoy and colleagues, who found that the degree of
reduction may be dependent on clozapine plasma levels (McEvoy et al.
1995b). A related study found that compared to a baseline medicationfree condition, taking the typical antipsychotic drug haloperidol leads
to increased smoking in patients with schizophrenia (McEvoy et al.
1995a). Thus, a role for atypical antipsychotic drugs in reducing smoking behavior is suggested.
Use of the transdermal nicotine patch (TNP) is known to facilitate
smoking reduction and cessation (Addington et al. 1998; George et al.
2000b) in smokers with schizophrenia, albeit at lower rates (36%42% at
trial endpoint) than in healthy control smokers (50%70%) (George and
OMalley 2004). Nonetheless, nicotine patch use at the dosage of
21 mg/day appears to effectively reduce cigarette smoking and nicotine withdrawal symptoms in smokers with schizophrenia (Addington
et al. 1998; George et al. 2000b). Other studies indicate that the combination of TNP with the atypical antipsychotic risperidone or olanzapine
may enhance smoking cessation rates compared with the combination
of typical antipsychotic drugs and TNP among smokers with schizophrenia with high motivation to quit smoking (George et al. 2000b).
Data from a preliminary placebo-controlled trial comparing bupropion
versus placebo in smokers with schizophrenia suggest that atypical
antipsychotic treatment significantly enhances smoking cessation responses to bupropion (George et al. 2002b).
One can speculate that in patients with schizophrenia, atypical antipsychotic drugs may be helpful for smoking cessation compared with
typical neuroleptic agents for the following reasons: 1) atypical agents
have fewer extrapyramidal side effects and improve negative symp-

Nicotine and Tobacco Use


toms, both of which may be further improved by cigarette smoking;

2) treatment with atypical agents is associated with improvement in
certain neuropsychological functional deficits, which also appear to be
alleviated by smoking; 3) sensory gating deficits (e.g., P50 responses,
prepulse inhibition) that are transiently normalized by nicotine administration or cigarette smoking are also ameliorated by atypical antipsychotic drugs; and 4) atypical agents are associated with augmentation
of dopamine release in the prefrontal cortex in rodent studies, and these
agents may normalize those presumed deficits in cortical dopamine
function in schizophrenia that are remediated by nicotine administration via cigarette smoking.

Optimizing Cessation Treatments to

Prevent Smoking-Related Medical Illness
Despite the enhanced quality of life possibly afforded to individuals
with schizophrenia through the use of atypical antipsychotics, smokers
with schizophrenia are more vulnerable than the general population of
smokers to developing smoking-related morbidity and mortality, including an increased risk of cardiovascular disease (Capasso et al.
2008). Previous epidemiological studies suggested that smokers with
schizophrenia were protected against the development of malignancies
(Tsuang et al. 1983), and this was thought to relate to neuroleptic drug
exposure (Mortensen 1987). In addition, evidence suggests that urinary
levels of the peptide bombesin, a possible marker of precancerous cigarette smokinginduced lung damage, are lower in patients with schizophrenia than in controls (Olincy et al. 1999). This reduction in urinary
bombesin levels is independent of smoking status in patients with
schizophrenia, supporting the notion that these patients may be less
vulnerable to the development of cancer. However, several subsequent
epidemiological studies have found no evidence for a decreased risk of
lung cancer in patients with schizophrenia or other patients with serious mental disorders (Lichtermann et al. 2001). Previous studies may
have been confounded by selection bias, because rates of these medical
illnesses in older patients with schizophrenia are lower, likely related to
the fact that much of this cohort had died from other causes related to
their psychiatric illness (e.g., suicide) by the time they reached the age
when cancer risk is substantially increased (age 50 years or older). Thus,
disease prevention through smoking cessation or reduction in this population is an important public health issue, especially because patients
with schizophrenia constitute 1% of the U.S. population.


Medical Illness and Schizophrenia

Although the interest in smoking cessation by patients with schizophrenia is generally thought to be low, approximately 20%40% have
substantial desire to quit smoking (Ziedonis and Trudeau 1997), based
on ratings of motivational level using the Stages of Change scale (e.g.,
preparation or action stages) (Prochaska and DiClemente 1983). In
many cases where smoking cessation is not possible, a reduction in
smoking consumption (e.g., a harm reduction approach) might accrue some health benefits for smokers with schizophrenia (McChargue
et al. 2002); however, no studies have been published to suggest that reducing smoking can reduce the risk of developing smoking-related illness in nonpsychiatric individuals or smokers with schizophrenia. In
fact, one study suggests that a 50% reduction in smoking does not reduce the risk of developing cardiac and pulmonary disease, compared
with the risk in persistent heavy smokers (Tverdal and Bjartveit 2006).
Thus, the approach to treatment of tobacco dependence in these patients should focus on the ultimate goal of smoking cessation, although
given the difficulties of cessation in patients with schizophrenia, more
flexible approaches (McChargue et al. 2002), including reduction as a
transition to abstinence, should be encouraged. Thus, an understanding
of biological and psychosocial factors that render schizophrenia patients at high risk for developing nicotine addiction and that contribute
to their low intrinsic motivation to change smoking behaviors is critical
to guiding efforts directed toward improving smoking cessation treatment in this population. One study using a single motivational interview versus an educational program in smokers with schizophrenia
(Steinberg et al. 2004) found that motivational interventions were associated with a significant increase in the chance that these smokers initiated smoking cessation treatment.
Experience in controlled treatment research studies for smoking cessation in schizophrenia has suggested the need to optimize both
pharmacological and psychosocial interventions (see Table 91). Although use of atypical antipsychotic drugs may be one patient factor
that predicts better smoking cessation or reduction outcomes, the reality is that patients with schizophrenia need persistent encouragement
to cease smoking through the use of motivational enhancement therapies and, once smoking abstinence has been achieved, require ongoing
teaching in methods to prevent smoking relapse (Addington et al. 1998;
Ziedonis and Trudeau 1997). Steinberg et al. (2004) found that a larger
percentage of smokers with schizophrenia attended a session of tobacco
dependence treatment after receiving a motivational interviewing session than after receiving standard psychoeducational treatment or brief
advice. In addition, educating patients about the dangers of smoking

Nicotine and Tobacco Use


for their health is helpful, because they often know surprisingly little
about the adverse effects of tobacco. Finally, drug refusal techniques are
important, because peer pressure is typically very high on these individuals to resume smoking after successful abstinence. Accrued experience in the Yale Program for Research in Smokers with Mental Illness
indicates that teaching assertiveness skills (in the context of social skills
training) has been quite effective for helping patients who maintain
high motivation to remain tobacco-free.
Standard smoking cessation pharmacotherapies approved by the
U.S. Food and Drug Administration (FDA), such as the transdermal nicotine patch (TNP) (Addington et al. 1998; Baker et al. 2006; Chou et al.
2004; George et al. 2000b), nicotine nasal spray (Williams et al. 2004),
and sustained-release bupropion (Evins et al. 2001, 2005; Fatemi et al.
2005; George et al. 2002b; Weiner et al. 2001), appear to be safe and efficacious treatments for smoking cessation in patients with schizophrenia during the course of controlled studies. A summary of treatment
and nontreatment studies directed at addressing tobacco addiction in
patients with schizophrenia is presented in Table 91.
The smoking cessation rates of patients with schizophrenia at the end
of drug treatment are in the range of 30%42% with the nicotine patch
(Addington et al. 1998; Baker et al. 2006; Chou et al. 2004; George et al.
2000b), and 11%50% with bupropion (Evins et al. 2001, 2005; George et
al. 2002b; Weiner et al. 2001), which are modest compared to rates
achieved in nonschizophrenic control smokers (50%75%) (Hughes et al.
1999) but may be improved when patients are prescribed atypical antipsychotic agents (George et al. 2000b, 2002b). Recently, the combination
of TNP and bupropion has been found to be safe and to increase shortterm abstinence rates in comparison to TNP alone (Evins et al. 2007;
George et al. 2008). Differences in study design, patient variables (e.g.,
level of motivation to quit smoking), medication dosage (in the studies
with bupropion, used at 150300 mg/day), and criteria used to determine
smoking abstinence may explain the variability in cessation rates across
these studies. In studies that have used TNP, patients are expected to stop
all smoking on the designated quit date (when they begin TNP).
When using the TNP, patients should be cautioned not to smoke
while wearing the patch due to concerns about nicotine toxicity, symptoms of which can include tremor, nausea and vomiting, dizziness, and
in rare cases seizures, arrhythmias, and death. In our research clinic, we
have not encountered nicotine toxicity to be a significant problem, but
we tell patients who feel they must smoke to remove the patch and wait
12 hours before resuming smoking. The craving to smoke and continuing withdrawal symptoms typically indicate incomplete nicotine


Smoking cessation and reduction approaches using pharmacotherapies in schizophrenia: summary of controlled

Evins et al.
Weiner et al.

Short-term abstinence rates Long-term abstinence rates

Open-label trial of
TNP +group counseling
TNP +group counseling

42% at study endpoint


3 months: 15%;
6 months: 12%

TNP vs. no-patch control

Nicotine nasal spray

26.9% vs. 0%

3 months: 26.9% vs. 0%

3 months: 42%


18 outpatients


8 outpatients

Open-label BUP
+supportive group

3 months: 30.0% vs. 6.0%;

6 months: 18.6% vs. 4.0%;
12 months: 18.6% vs. 6.6%
66% vs. 11% (50%
reduction in smoking)
0% quit smoking;
reduction in CO levels

6 months: 11% vs. 0%

(sustained abstinence)

Medical Illness and Schizophrenia

Nicotine replacement therapy

50 outpatients
et al. 1998
George et al. 45 outpatients
Chou et al.
68 inpatients
Williams et al. 12 outpatients
Baker et al.
298 outpatients with
nonacute psychotic

Study design




Smoking cessation and reduction approaches using pharmacotherapies in schizophrenia: summary of controlled
studies (continued)

Study design

Short-term abstinence rates Long-term abstinence rates

George et al.
Evins et al.
Fatemi et al.

32 outpatients

BUP+ counseling vs.

PLA + counseling

50% vs. 12.5%

18.8% vs. 6.3%

16% vs. 0%

3 months: 4% vs. 3.6%

Crossover design BUP and


Trend for reduction in CO,

cotinine, and nicotine levels


BUP+ TNP+ counseling vs.
PLA + TNP+ counseling

36% vs. 19%

53 outpatients
10 outpatients

Combination treatments
Evins et al.
51 outpatients
George et al. 58 outpatients
et al. 2007

71.7% vs. 55.2%


Open-label varenicline vs.

111 patients with
mental health
disorders (31 with
psychosis or
psychosis +

34.5% vs. 10.3%

6 months: 20% vs. 8%;

12 months: 12% vs. 8%
6 months: 13.8% vs. 0%

Nicotine and Tobacco Use


Smoking cessation and reduction approaches using pharmacotherapies in schizophrenia: summary of controlled
studies (continued)

Study design

Short-term abstinence rates Long-term abstinence rates

Evins and
Goff 2008

19 outpatients

Open-label varenicline


29 outpatients

Retrospective study of
Within-subjects study of
Within-subjects study of
TNP +counseling;
compared ATP vs. TYP

Reduction in smoking
with clozapine treatment
Reduction in smoking
with clozapine treatment
Reduction in smoking
with clozapine treatment
55.6% (ATP) vs. 22.2%
6 months: 16.7% vs. 7.4%

George et al.
McEvoy et al.
McEvoy et al.
George et al.

12 inpatients
70 inpatients
45 outpatients

6 months: 68% (4 patients

reported slips but
regained abstinence
within a week)

Note. All abstinence percentage rates are for 7-day point prevalence unless otherwise noted. ATP=atypical; BUP=bupropion;
CBT=cognitive-behavioral therapy; CO=carbon monoxide; MI=motivational interviewing; NRT=nicotine replacement therapy; PLA=placebo; TNP=transdermal nicotine patch; TYP=typical.

Medical Illness and Schizophrenia




Nicotine and Tobacco Use


replacement; if needed, another patch of 721 mg/day can be added to

therapy with the 21-mg/day TNP. For bupropion, controlled studies
have started dosing at 150 mg/day orally, with an increase to 150 mg
twice daily by the fourth day of treatment. The quit date is typically
set once levels reach steady-state, usually 34 days after beginning the
full 300-mg/day dosage. A history of seizures of any etiology is a contraindication to the use of bupropion, as indicated by the product labeling, and we recommend not exceeding 300 mg/day, because some
antipsychotic drugs may reduce seizure threshold. At the same time,
bupropion at 150300 mg/day does not appear to worsen positive
symptoms of schizophrenia and may reduce negative symptoms (Evins
et al. 2001, 2005; Fatemi et al. 2005; George et al. 2002b; Weiner et al.
2001). The typical duration of therapy studied in patients with schizophrenia with these agents is 812 weeks; studies with longer durations
of treatment in this population have not yet been conducted.
Varenicline tartrate (Chantix in the United States, Champix in Europe and Canada), an 42 nAChR partial agonist, is the newest medication approved by the FDA as a first-line smoking cessation agent.
Although clinical trials of varenicline with psychiatric smokers have
not yet been published, case reports suggested temporary increases in
psychiatric symptoms in a patient with schizophrenia (Freedman 2007)
and a patient with bipolar disorder (Kohen and Kremen 2007). An open
case series of 19 smokers with schizophrenia reported that 13 patients
were able to maintain abstinence (verified by measuring carbon monoxide level) for 12 weeks (Evins and Goff 2008). Over 6 months, four
patients reported slips and were able to regain abstinence within a
week. None of the 19 patients showed significant worsening of psychiatric symptoms, psychotic relapse, or hospitalization. Finally, a recent
study (Stapleton et al. 2007) compared the effectiveness of varenicline
and nicotine replacement in 412 smokers with and without psychiatric
disorders. Of the 111 patients with mental health disorders, 7 were
diagnosed with psychosis and 24 were diagnosed with combined psychosis and depression. Of those 111 patients with psychiatric disorders,
53 patients received varenicline and 58 patients received TNP. The results of this study suggested that varenicline was equally effective in
patients with and without mental illness, that varenicline was more effective than TNP, and that there was no evidence that taking varenicline
exacerbated psychiatric symptoms (Stapleton et al. 2007). Further controlled research is needed to determine the safety and efficacy of varenicline for patients with schizophrenia and other psychiatric disorders.


Medical Illness and Schizophrenia

Key Clinical Points

Smokers with schizophrenia smoke at higher rates and have lower

smoking cessation rates than the general population. This group of
smokers also shows higher rates of smoking-related illnesses, making
the development of effective treatments for this population an important priority.

Motivation to quit smoking is often low in patients with schizophrenia,

and efforts need to be undertaken to increase the awareness of both
patients and their clinicians of the dangers of habitual tobacco smoking. Motivational interviewing and relapse-prevention methods are the
mainstays of behavioral treatment for tobacco dependence in these patients.

The high rates of comorbid tobacco dependence in patients with schizophrenia may relate to abnormal biology of nicotinic receptor systems
and central dopaminergic, glutamatergic, and GABAergic pathways associated with this disorder.

Nicotine administration enhances several areas of cognitive function in

patients with schizophrenia, including working memory (specifically
spatial working memory), sustained attention, P50 gating, smooth pursuit eye movement, and prepulse inhibition, which may contribute to
the initiation and maintenance of tobacco dependence in these patients.

Controlled studies have suggested that treatments for tobacco dependence (e.g., nicotine replacement therapy, bupropion) are safe and improve smoking cessation outcomes for patients with schizophrenia.
Recent studies have examined the efficacy of combining medications
and newer pharmacological agents (e.g., varenicline) to further improve short- and long-term smoking cessation rates.

Although there is little evidence from controlled clinical studies that

smoking cessation produces a deterioration of clinical function (e.g.,
positive and negative symptoms) in stabilized patients, further research
is needed, especially for newer agents such as varenicline that have not
been tested in controlled trials. Clinicians should not encourage patients with schizophrenia to quit smoking when they are clinically unstable, because the likelihood of success is low and the risk of symptom
exacerbation may be elevated.

Nicotine and Tobacco Use


Addington J, el-Guebaly N, Campbell W, et al: Smoking cessation treatment for
patients with schizophrenia. Am J Psychiatry 155:974976, 1998
Adler LE, Hoffer LD, Waldo M, et al: Schizophrenia, sensory gating and nicotinic receptors. Schizophr Bull 24:189202, 1998
Baker A, Richmond R, Haile M, et al: A randomized controlled trial of a smoking cessation intervention among people with a psychotic disorder. Am J
Psychiatry 163:19341942, 2006
Barnes M, Lawford BR, Burton SC, et al: Smoking and schizophrenia: is symptom profile related to smoking and which antipsychotic medication is of
benefit in reducing cigarette use? Aust N Z J Psychiatry 40:575580, 2006
Capasso RM, Lineberry TW, Bostwick JM, et al: Mortality in schizophrenia and
schizoaffective disorder: an Olmsted county, Minnesota cohort: 19502005.
Schizophr Res 98:287294, 2008
Chou K-R, Chen R, Lee J-F, et al: The effectiveness of nicotine-patch therapy for
smoking cessation in patients with schizophrenia. Int J Nurs Stud 41:321
330, 2004
Coyle J, Kershaw P: Galantamine, a cholinesterase inhibitor that allosterically
modulates nicotinic receptors: effects on the course of Alzheimers disease.
Biol Psychiatry 49:289299, 2001
Dalack GW, Becks L, Hill E, et al: Nicotine withdrawal and psychiatric symptoms in cigarette smokers with schizophrenia. Neuropsychopharmacology
21:195202, 1999
Dani JA, Bertrand D: Nicotinic acetylcholine receptors and nicotinic cholinergic
mechanisms of the central nervous system. Annu Rev Pharmacol Toxicol
47:699729, 2007
Domino EF, Mirzoyan D, Tsukada H: N-methyl-D-aspartate antagonists as drug
models of schizophrenia: a surprising link to tobacco smoking. Prog Neuropsychopharmacol Biol Psychiatry 28:801811, 2004
Evins AE, Goff DC: Varenicline treatment for smokers with schizophrenia:
a case series. J Clin Psychiatry 69:1016, 2008
Evins AE, Mays VK, Rigotti NA, et al: A pilot trial of bupropion added to cognitive behavioral therapy for smoking cessation in schizophrenia. Nicotine
Tob Res 3:397403, 2001
Evins AE, Cather C, Deckerbach T, et al: A double-blind placebo-controlled trial
of bupropion sustained-release for smoking cessation in schizophrenia.
J Clin Psychopharmacol 25:218225, 2005
Evins AE, Cather C, Culhane MA, et al: A 12-week double-blind, placebocontrolled study of bupropion SR added to high-dose dual nicotine replacement therapy for smoking cessation or reduction in schizophrenia.
J Clin Psychopharmacol 27:380386, 2007
Fatemi SH, Stary JM, Hatsukami DK, et al: A double-blind placebo-controlled
cross over trial of bupropion in smoking reduction in schizophrenia.
Schizophr Res 76:353356, 2005


Medical Illness and Schizophrenia

Freedman R: Exacerbation of schizophrenia by varenicline. Am J Psychiatry

164:1269, 2007
Gallagher SM, Penn PE, Schindler E, et al: A comparison of smoking cessation
treatments for persons with schizophrenia and other serious mental illnesses. J Psychoactive Drugs 39:487497, 2007
George TP, OMalley SS: Current pharmacological treatments for nicotine dependence. Trends Pharmacol Sci 25:4248, 2004
George TP, Sernyak MJ, Ziedonis DM, et al: Effects of clozapine on smoking in
chronic schizophrenic outpatients. J Clin Psychiatry 56:344346, 1995
George TP, Verrico CD, Picciotto MR, et al: Nicotinic modulation of mesoprefrontal dopamine systems: pharmacologic and neuroanatomic characterization. J Pharmacol Exp Ther 295:5866, 2000a
George TP, Zeidonis DM, Feingold A, et al: Nicotine transdermal patch and
atypical antipsychotic medications for smoking cessation in schizophrenia.
Am J Psychiatry 157:18351842, 2000b
George TP, Vessicchio JC, Termine A, et al: Effects of smoking abstinence on
visuospatial working memory function in schizophrenia. Neuropsychopharmacology 26:7585, 2002a
George TP, Vessicchio JC, Termine A, et al: A placebo-controlled trial of bupropion for smoking cessation in schizophrenia. Biol Psychiatry 52:5361,
George TP, Termine A, Sacco KA, et al: Effects of cigarette smoking on prepulse
inhibition in schizophrenia: involvement of nicotinic receptor mechanisms.
Schizophr Res 87:307315, 2006
George TP, Vessicchio JC, Sacco KA, et al: A placebo-controlled trial of bupropion combined with nicotine patch for smoking cessation in schizophrenia.
Biol Psychiatry 63:10921096, 2008
Giovino GA: The tobacco epidemic in the United States. Am J Prev Med
33:S318S326, 2007
Goff DC, Henderson DC, Amico E: Cigarette smoking in schizophrenia: relationship to psychopathology and medication side effects. Am J Psychiatry
149:11891194, 1992
Green MF, Neucherterlein KH, Gold JM, et al: Approaching a consensus cognitive battery for clinical trials in schizophrenia: the NIMH-MATRICS conference to select cognitive domains and test criteria. Biol Psychiatry 56:301
307, 2004
Griffith JM, ONeill JE, Petty F, et al: Nicotinic receptor desensitization and sensory gating deficits in schizophrenia. Biol Psychiatry 44:98106, 1998
Hughes JR, Goldstein MG, Hurt RD, et al: Recent advances in the pharmacotherapy of smoking. JAMA 281:7276, 1999
Jacobsen LK, DSouza DC, Mencl WE, et al: Nicotine effects on brain function
and functional connectivity in schizophrenia. Biol Psychiatry 55:850858,
Kalman D, Morrisette SB, George TP: Co-morbidity of smoking with psychiatric
and substance use disorders. Am J Addict 14:106123, 2005

Nicotine and Tobacco Use


Kohen I, Kremen N: Varenicline-induced manic episode in a patient with bipolar disorder. Am J Psychiatry 164:12691270, 2007
Kroon LA: Drug interactions with smoking. Am J Health Syst Pharm 64:1917
1921, 2007
Kumari V, Soni W, Sharma T: Influence of cigarette smoking on prepulse inhibition of the acoustic startle response in schizophrenia. Hum Psychopharmacol 16:321326, 2001
Leonard S, Bertrand D: Neuronal nicotinic receptors: from structure to function.
Nicotine Tob Res 3:203223, 2001
Leonard S, Breese C, Adams C, et al: Smoking and schizophrenia: abnormal nicotinic receptor expression. Eur J Pharmacol 393:237242, 2000
Leonard S, Gault J, Hopkins J, et al: Promoter variants in the alpha-7 nicotinic
acetylcholine receptor subunit gene are associated with an inhibitory deficit found in schizophrenia. Arch Gen Psychiatry 59:10851096, 2002
Levin ED, McClernon FJ, Rezvani AH: Nicotinic effects on cognitive function:
behavioral characterization, pharmacological specification, and anatomic
localization. Psychopharmacology 184:523539, 2006
Lichtermann D, Ekelund E, Pukkala E, et al: Incidence of cancer among persons
with schizophrenia and their relatives. Arch Gen Psychiatry 58:573578,
Marder SR: Drug initiatives to improve cognitive function. J Clin Psychiatry
67:3135, 2006
McChargue DE, Gulliver SB, Hitsman B: Would smokers with schizophrenia
benefit from a more flexible approach to smoking treatment? Addiction
97:785793, 2002
McEvoy J, Freudenreich O, Levin E, et al: Haloperidol increases smoking in patients with schizophrenia. Psychopharmacology 119:124126, 1995a
McEvoy J, Freudenreich O, McGee M, et al: Clozapine decreases smoking in patients with chronic schizophrenia. Biol Psychiatry 37:550552, 1995b
McEvoy JP, Freudenreich O, Wilson WH: Smoking and therapeutic response to
clozapine in patients with schizophrenia. Biol Psychiatry 46:125129, 1999
Meyer JM: Individual changes in clozapine levels after smoking cessation: results and a predictive model. J Clin Psychopharmacol 21:569574, 2001
Mortensen PB: Neuroleptic medication and other factors modifying cancer risk
in schizophrenic patients. Acta Psychiatr Scand 75:585590, 1987
Newhouse PA, Potter A, Singh A: Effects of nicotinic stimulation on cognitive
performance. Curr Opin Pharmacol 4:3646, 2004
Olincy A, Leonard S, Young DA, et al: Decreased bombesin peptide response to
cigarette smoking in schizophrenia. Neuropsychopharmacology 20:5259,
Olincy A, Harris JG, Johnson LL, et al: Proof-of-concept trial of an alpha-7 nicotinic agonist in schizophrenia. Arch Gen Psychiatry 63:630638, 2006
Patkar AA, Gopalakrishnan R, Lundy A, et al: Relationship between tobacco
smoking and positive and negative symptoms in schizophrenia. J Nerv
Ment Dis 190:604610, 2002


Medical Illness and Schizophrenia

Picciotto MR: Nicotine as a modulator of behavior: beyond the inverted U.

Trends Pharmacol Sci 23:494499, 2003
Postma P, Gray JA, Sharma T, et al: A behavioural and functional neuroimaging
investigation into the effects of nicotine on sensorimotor gating in healthy
subjects and persons with schizophrenia. Psychopharmacology 184:589
599, 2006
Prochaska JO, DiClemente CC: Stages and processes of self-change of smoking:
toward an integrative model of change. J Consulting Clin Psychol 51:390
395, 1983
Sacco KA, Bannon KL, George TP: Nicotinic receptor mechanisms and cognition in normal states and neuropsychiatric disorders. J Psychopharmacol
18:457474, 2004
Sacco KA, Termine A, Seyal AA, et al: Effects of cigarette smoking function on
spatial working memory and attentional function in schizophrenia: involvement of nicotinic receptor mechanisms. Arch Gen Psychiatry 62:649
659, 2005
Smith RC, Singh A, Infante M, et al: Effects of cigarette smoking and nicotine
nasal spray on psychiatric symptoms and cognition in schizophrenia. Neuropsychopharmacology 27:479497, 2002
Smith RC, Warner-Cohen J, Matute M, et al: Effects of nicotine nasal spray on
cognitive function in schizophrenia. Neuropsychopharmacology 31:637
643, 2006
Stapleton JA, Watson L, Spirling LI, et al: Varenicline in the routine treatment
of tobacco dependence: a pre-post comparison with nicotine replacement
therapy and an evaluation in those with mental illness. Addiction 103:146
154, 2007
Steinberg ML, Ziedonis DM, Krejci JA, et al: Motivational interviewing with
personalized feedback: a brief intervention for motivating smokers with
schizophrenia to seek treatment for tobacco dependence. J Consult Clin
Psychol 72:723728, 2004
Swerdlow NR, Caine SB, Braff DL, et al: The neural substrates of sensorimotor
gating of the startle reflex: a review of recent findings and their implications. J Psychopharmacol 6:176190, 1992
Tamminga CA: The neurobiology of cognition in schizophrenia. J Clin Psychiatry 67:e11, 2006
Tang YL, George TP, Mao PX, et al: Cigarette smoking in Chinese male inpatients with schizophrenia: a cross-sectional analysis. J Psychiatr Res 41:43
48, 2007
Tsuang MT, Perkins K, Simpson JC: Physical diseases in schizophrenia and affective disorder. J Clin Psychiatry 44:4246, 1983
Tsukada H, Miyasato K, Nishiyama S, et al: Nicotine normalizes increased prefrontal cortical dopamine D1 receptor binding and decreased working
memory performance produced by repeated pretreatment with MK-801: a
PET study in conscious monkeys. Neuropsychopharmacology 30:2144
2153, 2005a

Nicotine and Tobacco Use


Tsukada H, Nishiyama S, Fukumoto D, et al: Chronic NMDA antagonism impairs working memory, decreases extracellular dopamine, and increases
D 1 receptor binding in the prefrontal cortex of conscious monkeys. Neuropsychopharmacology 30:18611869, 2005b
Tverdal A, Bjartveit K: Health consumption of reduced daily cigarette consumption. Tob Control 15:472480, 2006
Weiner E, Ball MP, Summerfelt A, et al: Effects of sustained-release bupropion
and supportive group therapy on cigarette consumption in patients with
schizophrenia. Am J Psychiatry 158:635637, 2001
Williams JM, Ziedonis DM, Foulds J: A case series of nicotine nasal spray in the
treatment of tobacco dependence among patients with schizophrenia. Psychiatr Serv 55:10641066, 2004
Wooltorton J, Pidoplichko VI, Broide RS, et al: Differential desensitization and
distribution of nicotinic acetylcholine receptor subtypes in midbrain
dopamine areas. J Neurosci 23:31763185, 2003
Ziedonis DM, Trudeau K: Motivation to quit using substances among individuals with schizophrenia: implications for a motivation-based treatment
model. Schizophr Bull 23:229238, 1997
Ziedonis DM, Kosten TR, Glazer WM, et al: Nicotine dependence and schizophrenia. Hosp Community Psychiatry 45:204206, 1994

This page intentionally left blank

Special Topics and

This page intentionally left blank

HIV and Hepatitis C
in Patients With
Milton L. Wainberg, M.D.
Francine Cournos, M.D.
Karen McKinnon, M.A.
Alan Berkman, M.D.
Mark Drew Crosland Guimares,
M.D., D.Sc., M.P.H.

Among the chronic health conditions experienced by people with

schizophrenia, infection with human immunodeficiency virus (HIV) and
hepatitis C virus (HCV) occur with concerning frequency. People who
contract HIV or HCV face chronic illness, the possibility of premature
death, complicated medication regimens, barriers to medical care, and the
neuropsychiatric sequelae of treatment or infection itself. Estimates are
that approximately one-quarter of those infected with HIV and the majority of those with HCV in the United States do not know they are infected
(Agency for Healthcare Research and Quality 2002; Marks et al. 2006).
HIV and HCV are acquired by similar routes of transmission, although HIV is much more likely to be sexually transmitted than HCV.
In the United States, rates of both diseases are higher among African
Americans and Latinos than among Caucasians. Both infections are


Medical Illness and Schizophrenia

overrepresented among people with severe mental illness, including

those with schizophrenia, especially when comorbid substance use is
present. Published rates of HIV infection among psychiatric patients in
the United States are 3.1%23.9% (Cournos and McKinnon 1997; Rosenberg et al. 2001), at least eight times higher than the general population
estimate of 0.6% (UNAIDS 2007). HIV infection rates among patients
with schizophrenia rarely have been differentiated from rates among
others with psychotic disorders, but when they have, no significant differences have emerged. In the United States, estimates of HCV infection
in samples of patients with mental illness, including schizophrenia, have
ranged from 8.2% to 38.0% (Al Jurdi and Burruss 2003; Butterfield et al.
2003, 2004; Dinwiddie et al. 2003; Freudenreich et al. 2007; Huckans et al.
2006; Klinkenberg et al. 2003; Meyer 2003; Rosenberg et al. 2001; Tabibian et al. 2008). These rates are higher than the general population rate of
approximately 1.6% (Centers for Disease Control and Prevention 2008).
Psychiatric symptoms and disabilities may increase risk of HIV or
HCV among people with severe mental illness, either by directly affecting behavior or by interfering with the opportunity or ability to acquire
and/or use information about these illnesses to practice safer behaviors. Psychiatrists and other mental health care providers often are in
the best position to enhance patients skills in modifying risk behaviors
and to connect patients to the medical services they need.

Sexual Risk Behaviors and

Psychiatric or Situational Factors
Studies of sexual risk behavior among people with psychiatric disorders
have thus far not linked behaviors to biological outcomes, and few studies have differentiated risk behaviors by psychiatric diagnosis. However,
risk behaviors among people in psychiatric treatment in the United
States are common, with interview studies revealing that a majority of
patients were sexually active with a partner in the previous year and that
the sexual activity of people with severe mental illness is characterized
by having multiple sex partners and by a lack of condom use in a majority of sexual occasions for both men and women (Meade and Sikkema
2005). Being sexually active has been found to be associated with a diagnosis of schizophrenia but not with bipolar disorder, and trading sex for
money or drugs was more than three times as likely among patients with
schizophrenia than among those with other diagnoses, and more than
five times as likely among those with certain positive symptoms such as
delusions (McKinnon et al. 1996; Meade and Sikkema 2007). Diagnosis

HIV and Hepatitis C


has not been shown to be associated with number of sex partners or condom use (Meade and Sikkema 2007), although greater psychiatric symptoms were associated with increased HIV risk (Rosenberg et al. 2001).
Extended periods of institutionalization on same-sex units in hospitals, shelters, or prisons may foster high-risk same-sex activity, often
among those who do not self-identify as gay or lesbian. One study that
directly compared psychiatric patients with nonpsychiatric groups
found that psychiatric patients are more likely to engage in same-sex activity (McDermott et al. 1994). This behavior is particularly risky for men.
Recurrent institutionalization, homelessness, transient living circumstances, alienation from supportive social relationships, and lack of
privacy can all interrupt long-term relationships, reinforcing the tendency to have unfamiliar partners. Studies report that 10%16% of psychiatric patients had sex in the past year with someone they had known
less than 24 hours (Kalichman et al. 1994; Kelly et al. 1992). These conditions also may make changing risk behaviors difficult for people with
severe mental illness or may limit sexual opportunities to those that
confer greater risk. Unemployment, which is common among adults
with severe mental illness, also may contribute to greater sexual risk
taking (Meade and Sikkema 2007), possibly through pressures toward
survival sex or commercial sex work. Sexual victimization, which
increases the likelihood of unprotected intercourse, has been widely
reported by psychiatric inpatients (Malow et al. 2006). Among outpatients, one in eight reported having been pressured, coerced, or forced
into unwanted sex in the past year (Carey et al. 1997).
Having a sexually transmitted infection, which renders a person biologically more susceptible to acquiring subsequent infections when exposed, is common among people with severe mental illness. Between 9%
and 36% of psychiatric inpatients and outpatients are diagnosed with one
or more sexually transmitted infections at some time in their lives (Carey
et al. 1997; Rosenberg et al. 2001). Moreover, some patients may have sexually transmitted infections without being aware of them and will likely
remain untreated unless screened for sexually transmitted infections.

Substance Use Risk Behaviors and

Psychiatric or Situational Factors
In the United States, psychiatric patients with identified comorbid alcohol
or other drug use disorders have a significantly higher rate of HIV infection than those without (McKinnon and Cournos 1998), even if they have
never injected drugs. In this population, any lifetime drug injection or


Medical Illness and Schizophrenia

needle sharing appears to increase the risk of being infected with HIV
more than sixfold (Rosenberg et al. 2001). As in the general population,
HCV infection in psychiatric patients is most strongly associated with injection drug use, and to a lesser extent with other drug use (including sniffing and smoking forms of cocaine), unsafe sexual practices, and poverty.
A history of drug injection has been noted among 1%26% of psychiatric patients, and 5%12% report having ever shared needles or other
injection paraphernalia (Carey et al. 1997; Rosenberg et al. 2001). Moreover, psychiatric patients who inject tend to do so intermittently rather
than regularly, so drug injection histories often are overlooked in mental health settings, and no conclusive evidence has been found to link
injection drug use with psychiatric diagnosis, chronicity, level of functioning, or psychiatric symptoms. As in the general population, alcohol
or other drugs often are part of psychiatric patients sexual experience
and may decrease the motivation or ability to have protected sex. Of
psychiatric inpatients and outpatients, 15%45% reported using substances during sex in the past year (Carey et al. 1997; Menon and Pomerantz 1997), with comparable rates seen in men and women.

Knowledge About HIV/HCV Among

People With Severe Mental Illness and
Their Providers
Knowledge about HIV and acquired immune deficiency syndrome
(AIDS) among U.S. psychiatric patients appears to be relatively good.
Correct responses to AIDS knowledge questionnaires in a variety of
psychiatric patient groups ranged from 63% to 80%, a comparable accuracy rate to that found in the general U.S. population (Meade and
Sikkema 2005). Still, many psychiatric patients held critical misinterpretations about HIV and related risks. In one study, for example, 48% of
outpatients believed that careful cleansing after sex would provide protection from the virus (Otto-Salaj et al. 1998). Comparable studies about
HCV knowledge among psychiatric patients have not been completed.

HIV/AIDS: Overview of Course of

Illness and Treatment
As of 2007, approximately 1.3 million persons in North America and
33 million persons worldwide were living with HIV (UNAIDS 2007),
which was first identified in 1983. HIV infection is a chronic condition

HIV and Hepatitis C


that normally runs its course over many years, with AIDS being a latestage manifestation of HIV infection, characterized by increasing viral
load of HIV in the blood and declining immune function (as measured
by CD4 T-cell counts), leading to a broad spectrum of diseases. According to the Centers for Disease Control and Prevention ([CDC] 2008), the
diagnosis of AIDS is made in an HIV-positive individual who has at
least one of 25 AIDS-defining conditions (e.g., Pneumocystis carinii
pneumonia, HIV-associated dementia, or a CD4 T-cell count lower than
200 cells per cubic millimeter). HIV has been identified as the cause of
AIDS (Barre-Sinoussi et al. 1983). The virus invades the nervous system,
causes persistent viremia, and weakens the humoral and cellular immune response. HIV has a remarkably complex viral genome, which
probably underlies its profound pathogenicity.
HIV mutates rapidly, making drug resistance a major problem and
contributing to difficulties in producing an effective vaccine. HIV-1 is
the most common type of HIV virus in the United States and much of
the rest of the world. HIV-2 occurs in West Africa, with scattered cases
appearing in Asia. HIV-2 is associated with considerably slower disease
progression than HIV-1. Having either virus does not alter the susceptibility to infection with the other strain. All vaccine trials thus far have
failed to show protection against acquiring HIV, and one trial even suggested increased susceptibility among a subgroup of those vaccinated,
leading to a call for more basic science research to further clarify vaccine
development strategies (Walker and Burton 2008).

HIV Transmissibility
Although HIV has been isolated from a variety of body fluids, including blood, semen and preseminal fluid, vaginal secretions, breast milk,
urine, saliva, and tears, the risk of transmission is a consequence of the
viral load of the fluid. HIV is found in such small quantities in tears, saliva, and urine that casual contact with these fluids is a very unlikely
mode of transmission. Epidemiological studies indicate that semen and
preseminal fluid, cervical and vaginal secretions, breast milk, and blood
and blood products are the predominant, if not exclusive, vehicles for
viral transmission (Staprans and Feinberg 1997).
HIV is typically spread by sexual contact, exposure to infected blood
(transfusions, blood products, percutaneous and intravenous injections
with contaminated syringes or needles, etc.), and through perinatal
transmission from mother to child. Although uncommon, infection is
possible through the exposure of cuts in skin or mucous membranes to
HIV-infected blood.


Medical Illness and Schizophrenia

Penile-vaginal and penile-anal intercourse are considered the highest-risk sexual behaviors, in addition to activities that cause a rupture of
tissue and the presence of blood. Although infection risk is somewhat
higher for the recipient of semen than for the insertive partner, transmission has been documented in both directions. Male circumcision
significantly reduces the risk of HIV acquisition and transmission by
men during penile-vaginal sex; however, condoms are still needed.
Oral sex has also been documented as a mode of transmission for the recipient of fluids, as has the sharing of sexual toys, although to a lesser
extent. Certain cofactors enhance the risk of sexual transmission of HIV,
including the presence of sexually transmitted infections, genital lesions, or genital or mucous membrane bruising during sexual activity.
Sharing needles or other equipment during injection is a very efficient means of transmitting HIV and amounts to a direct inoculation of
viral particles from the infected to the noninfected person. As with
other routes of transmission, the likelihood increases with the size of
the viral inoculum. Even noninjection substance use may increase the
risk for HIV by increasing the chance that an individual will engage in
high-risk behaviors due to lowered sexual inhibitions, impaired judgment, increased impulsiveness, or the exchange of sex for drugs or
money to buy drugs. Transfusion with infected blood and the use of infected blood products almost always results in acquisition of HIV, although testing of donated blood and blood products has almost
eliminated the chances of this occurrence in the industrialized countries.

HIV Testing
The most commonly used HIV test is the enzyme-linked immunosorbent assay (ELISA), which, if positive, is followed by Western blot testing. The Western blot test is needed to rule out false positives on the
ELISA test. Antibody-based assays are available as rapid (20-minute)
tests of either blood or saliva, but the blood test is more accurate; positive results on either test must be confirmed by Western blot. These
tests detect antibodies produced by the host as an immune response to
certain genetic components of the virus but do not detect the virus itself.
Although sensitive and specific, even the Western blot can give falsenegative or indeterminate results, especially during the first few weeks
of infection. However, direct measurement of viral presence using the
polymerase chain reaction assay will usually show extremely high levels of virus during this window period. False-positive results may occur
as well. Despite the latter concern, the CDC now recommends routine

HIV and Hepatitis C


opt-out HIV testing (testing for HIV as part of routine medical care unless the patient refuses) in all medical settings (Branson et al. 2006).

Viral Load and Resistance Assays

Viral load, a quantifiable measurement of how many viral particles are
present in a milliliter of blood, can be determined by several different
assays, some of which detect as few as 50 viral particles per milliliter of
blood. Below that threshold of measurement, the result is reported as
nondetectable. This does not indicate that there is no virus present in
blood at all, nor does it measure the amount of virus in lymphoid tissue
or in the central nervous system (CNS). Viral load is a strong predictor
of disease progression in untreated patients. For those on antiretroviral
therapy, CD4 T-cell counts are the stronger predictor of clinical outcome. Both CD4 T-cell counts and viral load are used to monitor clinical
response to therapy (Hulgan et al. 2007). Predicting responses to antiviral medications can be accomplished by testing for drug resistance, a
problem increasingly seen in clinical practice. Resistance assays include
genotyping to examine the patients virus for mutations known to confer drug resistance; phenotyping to test the susceptibility of the patients virus to specific medications; and virtual phenotyping, which
uses a large database that associates genotypes with phenotypes to predict drug susceptibility.

Natural History of HIV Disease

HIV contains ribonucleic acid (RNA) as its genetic material. HIV targets
host CD4 T-lymphocytes and other susceptible cells by identifying surface molecules and attaching to and entering the cells (Staprans and
Feinberg 1997). This begins the process of using the viral reverse transcriptase enzyme to transcribe viral RNA to deoxyribonucleic acid
(DNA), which then allows the virus to use the host cells machinery to
replicate itself. The replicated virus particles must be split apart into
virions by the protease enzyme before they are extruded from the host
cell so as to become functional and capable of infecting other CD4+
cells, leading to ongoing damage to the immune system. As the host
produces circulating antibodies against HIV, the host is said to seroconvert. Genetic information to reproduce the virus is also integrated into
the host cells genome. The immune system may initially contain but
does not clear the infection; the course is set for chronic persistent viral
replication. Without treatment, near complete destruction of the CD4 Tlymphocyte population eventually occurs in the vast majority of infected people. A persons initial viral setpoint, the capacity of his or


Medical Illness and Schizophrenia

her immune system to limit viral replication, is the strongest predictor

of untreated disease progression (Staprans and Feinberg 1997).
The range and severity of symptoms in acute HIV infection vary considerably, with approximately 40%90% of patients developing a retroviral syndrome characterized by a variety of symptoms that may
include fever, lymphadenopathy, pharyngitis, malaise, myalgia, arthralgia, rash, headache, stiff neck, and other meningeal signs and
symptoms. Acute HIV infection is highly transmissible due to transient
intense viremia, and is accompanied by an acute fall in CD4 T-cell
counts in the peripheral blood from its normal range of 8001,200 cells/
mm3 (Soogoor and Daar 2005). The more severe this syndrome is, the
more likely the untreated patient will progress rapidly to AIDS (Fidler
et al. 2008). Although there are reasons to believe that favorable treatment during acute HIV infection can alter its long-term course, this is
not yet established beyond question, and patients are encouraged to enroll in clinical trials studying the course of acute HIV infection (http://
Once the symptoms of primary infection subside and an antiviral
immune response appears, patients usually enter a chronic, clinically
asymptomatic or minimally symptomatic state despite continuous active viral replication. This period may last only a few years in some infected individuals, but the majority of HIV-positive patients develop
overt immunodeficiency in approximately 810 years, and a small cohort demonstrates sustained long-term (>10 years), symptom-free HIV
infection (Staprans and Feinberg 1997). During chronic infection, the
development of symptoms, a low CD4 cell count, and a high viral load
should initiate a discussion between clinician and patient about antiretroviral treatment. The patients ability to adhere to the regimen is, of
course, pivotal to the decision.
The spectrum of HIV-associated illnesses that eventually develop includes constitutional symptoms (e.g., weight loss, fatigue, fever, night
sweats) and involvement of multiple organ systems. Opportunistic infections (OIs) are multiple and can occur throughout the body. These
include fungal infections (e.g., oral or esophageal candidiasis [thrush],
cryptococcal infection, Pneumocystis carinii pneumonia); protozoan infections (e.g., toxoplasmosis); mycobacterial infections (e.g., tuberculosis and Mycobacterium avium complex); and bacterial and viral
infections (e.g., recurrent pulmonary and gynecological infections).
Cancers, such as Kaposis sarcoma and lymphoma, are other manifestations of severe immunosuppression. Prophylactic regimens can reduce
the occurrence of many OIs in immunocompromised patients.

HIV and Hepatitis C


Neuropsychiatric Manifestations of HIV

HIV infection presents a spectrum of neuropsychiatric sequelae that
can pose diagnostic and treatment quandaries for clinicians. In patients
with serious and persistent psychiatric illness, some of the early, subtle
neuropsychiatric symptoms may be difficult to differentiate from preexisting symptoms of psychiatric illness. HIV is neurotropic and enters
the CNS soon after infection. Long-term clinical sequelae of CNS infection range from subtle neurocognitive impairment to frank dementia,
and their incidence increases with HIV illness progression. Opportunistic infections and neoplasms that follow immunosuppression can also
affect the CNS, resulting in mood disorders, psychosis, cognitive disorders, delirium, and other neuropsychiatric abnormalities. In addition,
prescribed and recreational psychoactive substance use may create
neuropsychiatric complications and must be considered in the differential diagnosis of patients who present with new mental status changes
(American Psychiatric Association 2000; Wainberg et al. 2000).
HIV-associated neurocognitive disorders (HAND) are diagnoses of
exclusion made after other etiologies have been ruled out through a
comprehensive evaluation. Table 101 describes the diagnostic criteria
for the three categories of HAND: asymptomatic cognitive disorder,
mild neurocognitive disorder, and dementia. (Antinori et al. 2007).
HAND are subcortical disorders that affect primarily and initially
the CNS white matter and, as time progresses, other CNS tissues. Although the exact pathophysiology of HAND remains unclear, diagnosis of HAND is relatively common, particularly in more advanced
stages of HIV infection. Effective antiretroviral treatment has dramatically reduced the incidence of HAND. Antiretroviral treatment should
be initiated or the regimen evaluated following diagnosis of symptomatic HAND, and other medications can be prescribed to treat associated
symptoms (e.g., antidepressants, stimulants, testosterone).
Psychiatric symptoms due to HIV-related medical conditions are
most common in advanced stages of illness (American Psychiatric Association 2000). Therefore, among patients with advanced HIV who
have a preexisting severe mental illness, psychiatric changes should not
be attributed to a relapse until a complete medical workup has ruled
out other causes. Mental status changes that can have a medical etiology include shifts in level of consciousness characteristic of delirium,
cognitive impairment, mood changes, and psychotic symptoms. The
differential diagnosis includes not only the neuropsychiatric manifestations of HIV itself but also opportunistic infections (e.g., toxoplasmosis,


Medical Illness and Schizophrenia

TABLE 101. Criteria for HIV-associated neurocognitive disorders

HIV-associated asymptomatic neurocognitive
impairment (ANI)a
1. Acquired impairment in cognitive functioning, involving at least
two ability domains, documented by performance of at least
1 standard deviation (SD) below the mean for age- and educationappropriate norms on standardized neuropsychological tests. The
neuropsychological assessment must survey at least the following
abilities: verbal/language; attention/working memory;
abstraction/executive; memory (learning, recall); speed of
information processing; sensory-perceptual and motor skills.
2. The cognitive impairment does not interfere with everyday
3. The cognitive impairment does not meet criteria for delirium or
4. There is no evidence of another preexisting cause for the ANI.b
HIV-1associated mild neurocognitive
disorder (MND)c
1. Acquired impairment in cognitive functioning, involving at least
two ability domains, documented by performance of at least 1 SD
below the mean for age- and education-appropriate norms on
standardized neuropsychological tests. The neuropsychological
assessment must survey at least the following abilities: verbal/
language; attention/working memory; abstraction/executive;
memory (learning, recall); speed of information processing;
sensory-perceptual and motor skills.
2. The cognitive impairment produces at least mild interference in
daily functioning (at least one of the following):
a) Self-report of reduced mental acuity or of inefficiency in work,
homemaking, or social functioning.
b) Observation by knowledgeable others that the individual has
undergone at least mild decline in mental acuity with
resultant inefficiency in work, homemaking, or social
3. The cognitive impairment does not meet criteria for delirium or
4. There is no evidence of another preexisting cause for the MND.d

HIV and Hepatitis C


TABLE 101. Criteria for HIV-associated neurocognitive disorders

(HAND) (continued)
HIV-1associated dementia (HAD)e
1. Marked acquired impairment in cognitive functioning, involving
at least two ability domains; typically the impairment is in
multiple domains, especially in learning of new information,
slowed information processing, and defective attention/
concentration. The cognitive impairment must be ascertained by
neuropsychological testing with at least two domains reduced by
2 SD or greater than demographically corrected means. (Note that
where neuropsychological testing is not available, standard
neurological evaluation and simple bedside testing may be used,
but this should be done using well-defined criteria.)
2. The cognitive impairment produces marked interference with
day-to-day functioning (work, home life, social activities).
3. The pattern of cognitive impairment does not meet criteria for
delirium (e.g., clouding of consciousness is not a prominent
feature), or if delirium is present, criteria for dementia need to have
been met on a prior examination when delirium was not present.
4. There is no evidence of another, preexisting cause for the dementia
(e.g., other central nervous system [CNS] infection, CNS neoplasm,
cerebrovascular disease, preexisting neurological disease, or
severe substance abuse compatible with CNS disorder).f

If the patient had a prior diagnosis of ANI but currently does not meet
criteria, the diagnosis of ANI in remission can be made.

If the individual with suspected ANI also satisfies criteria for a major depressive episode or substance dependence, the diagnosis of ANI should
be deferred to a subsequent examination conducted at a time when the
major depression has remitted or at least 1 month after cessation of substance use.

If the patient had a prior diagnosis of MND but currently does not meet
criteria, the diagnosis of MND in remission can be made.

If the individual with suspected MND also satisfies criteria for 1) a severe
episode of major depression with significant functional limitations or psychotic features or 2) substance dependence, the diagnosis of MND should
be deferred to a subsequent examination conducted at a time when the
major depression has remitted or at least 1 month after cessation of substance use.

If the patient had a prior diagnosis of HAD but currently does not meet
criteria, the diagnosis of HAD in remission can be made.


Medical Illness and Schizophrenia

TABLE 101. Criteria for HIV-associated neurocognitive disorders

(HAND) (continued)

If the individual with suspected HAD also satisfies criteria for 1) a severe
episode of major depression with significant functional limitations or psychotic features or 2) substance dependence, the diagnosis of HAD should
be deferred to a subsequent examination conducted at a time when the
major depression has remitted or at least 1 month after cessation of substance use. Note that the consensus was that even when major depression
and HAD occurred together, there is little evidence that pseudodementia
exists and the cognitive deficits do not generally improve with treatment
of depression.


Adapted from Antinori et al. 2007.

cryptococcus, tuberculous meningitis), lymphoma, or delirium from

metabolic derangement, substance use, or drug toxicity (Wainberg et al.

HIV/AIDS Treatment
Clinical management of HIV/AIDS is intended to maximally and durably suppress viral load, restore and preserve immune function, provide
prophylaxis against opportunistic infections as appropriate, treat OIs
when present, and decrease morbidity and mortality. Guidelines for the
prophylaxis of OIs are based on the prevalence of each OI at various
stages of immunodeficiency. In addition to the diagnosis of AIDS, other
indications for treatment in an HIV-positive person include pregnancy,
HIV-associated nephropathy, symptomatic HAND, and a CD4 cell
count between 200 and 350. The need to treat other infections, such as
hepatitis B and tuberculosis, also affects the timing and choice of antiretroviral agents for HIV. Guidelines for treatment of HIV infection are
continually updated as new medications and more data from clinical
trials become available. U.S. recommendations can be found at http:// This site, maintained by the U.S. Department of
Health and Human Services, also includes guidelines for postexposure
prophylaxis following either an injury that carries risk in the health care
setting or a sexual exposure.
Part of instituting antiretroviral therapy involves ensuring that patients recognize their HIV infection, have access to health care, develop
ongoing provider relationships, and are motivated to adhere to treatment, even during an asymptomatic phase of infection. When patients
are not ready to adhere to antiretroviral therapy (e.g., because of chaotic

HIV and Hepatitis C


life due to psychiatric instability, substance use, and/or homelessness;

lack of motivation; insufficient social supports), barriers to adherence
should be addressed first. High levels of adherence are required to suppress fully the replication of virus that is sensitive to the regimen being
administered and to prevent the emergence of drug resistance. Mental
health professionals often have important roles in assessing and promoting adherence.
As of June 2008, the U.S. Food and Drug Administration had approved 22 individual antiretroviral medications, which are classified
into six categories according to the specific step inhibited in the HIV life
cycle (see Table 102). Individually, these agents are not very potent,
but by combining three or more agents, effective viral suppression can
be achieved even among many patients with resistant strains. Moreover, many regimens are simpler and contain fewer pills than in the

HCV: Overview of Course of

Illness and Treatment
Approximately 4 million persons in the United States and 170 million
persons worldwide are infected with hepatitis C. The virus causes persistent infection in a majority of infected people, although most have
relatively mild disease with slow progression. However, chronic and
progressive HCV carries significant morbidity and mortality and is a
major cause of cirrhosis, end-stage liver disease, and liver cancer. Six
different genotypes have been identified. Most people in the United
States who have HCV have genotype 1, which is difficult to treat. Development of an effective HCV vaccine is not imminent but continues
to be pursued, because current therapy for HCV is poorly tolerated and
not effective for a substantial number of patients.

HCV Transmission
HCV is spread primarily through infected blood and is therefore a common complication of injection drug use and sharing of razors or tattooing equipment. HCV may also be spread by maternal fetal transmission
and noninjection drug use activities. Although HCV is sexually transmissible, its efficiency of transmission via this mode is far less than
other blood-borne viruses, including HIV (Clarke and Kulasegaram


Medical Illness and Schizophrenia

TABLE 102. Medications for HIV infection approved by the U.S. Food
and Drug Administration as of June 2008
Generic name

Trade name

Nucleoside/nucleotide reverse transcriptase inhibitors

Videx EC
Non-nucleoside reverse transcriptase inhibitors
Protease inhibitors


Fusion inhibitor


CCR5 inhibitor


Integrase inhibitor


Combination reverse transcriptase inhibitors

Abacavir and lamivudine
Abacavir, zidovudine, and lamivudine
Efavirenz, tenofovir, emtricitabine
Tenofovir and emtricitabine
Zidovudine and lamivudine


HIV and Hepatitis C


HCV Testing
Testing for HCV does not involve the counseling, confidentiality, or
stigma issues usually associated with HIV testing. Providers vary
greatly in how they proceed after identifying HCV-positive patients.
Given that about 20%33% of patients with severe mental illness in the
United States are infected with HCV, a strong argument can be made
that all such persons should be screened for the virus (Goldberg and
Seth 2008). Certainly, all patients known to be infected with HIV should
be screened for anti-HCV antibodies as part of their initial evaluation,
as should those with high-risk behavior. HCV infection should be confirmed with qualitative polymerase chain reaction assay if the patient is
at low risk and the diagnosis seems in doubt. Moreover, there is a small
but measurable false-negative rate for antibody testing in patients who
are severely immunosuppressed.

Natural History of HCV Disease

Acute HCV disease is usually asymptomatic, but 25%35% of patients
develop some constitutional symptoms or jaundice. Serum alanine aminotransferase levels frequently rise, fluctuate, and fall again, suggesting
recovery from the acute phase. However, following acute infection,
HCV is not easily cleared by the immune system, and 75%80% of acute
HCV infections become chronic, as evidenced by persistent or intermittent HCV viremia.
Chronic HCV infection can cause inflammatory infiltration, particularly of the portal tracts, as well as focal liver cell necrosis and fibrosis
that bridges between portal tracts. Hepatitis C is the leading cause of
liver transplantation in the United States. About 10%20% of chronically infected patients progress to cirrhosis within 2030 years of infection, and among those with cirrhosis, 1%4% per year will develop
hepatocellular carcinoma. Patients infected with HCV should be advised to minimize or preferably discontinue intake of alcohol.
Immunosuppression associated with HIV significantly alters the natural history and clinical course of HCV, with HCV-associated cirrhosis
occurring more frequently in patients with HCV/HIV co-infection (33%)
than in HCV alone (11%). People with concurrent HCV do not tolerate
highly active antiretroviral therapy as well as people with HIV alone, and
this can interfere with effective HIV treatment (Soriano et al. 2008). Liver
failure due to HCV is the leading non-AIDS cause of death in HIVinfected individuals (de Ldinghen et al. 2008). Because of increased risk
of severe liver damage, all patients with HCV and/or HIV should be
screened for immunity to hepatitis A and B and immunized accordingly.


Medical Illness and Schizophrenia

Neuropsychiatric Manifestations of HCV

HCV-related cognitive impairment was previously assumed to be a consequence of cirrhosis-associated hepatic encephalopathy. Recent evidence suggests that one-third of people with chronic HCV experience
cognitive impairment even in the absence of cirrhosis and that its occurrence is unrelated to laboratory values, viral load, and genotype (Perry
et al. 2008). Attention, concentration, and psychomotor speed are the
most impaired cognitive functions and may influence adherence to
medical care and medications. HCV injury to the brain can occur via infection of resident (e.g., astrocytes) and migrating (e.g., macrophages)
cells of the central nervous system; adaptation to neural cells; inflammation of basal ganglia and white matter; or neuronal loss (Letendre 2008).

HCV Treatment
Diagnostic testing to determine the presence of HCV viremia and the
extent of liver pathology should be completed as early as possible in the
care of a patient infected with HCV. Patients with active HCV infection
or evidence of chronic liver disease should be referred for care to a specialist with experience in treating hepatitis C. Not all patients with HCV
require or benefit from treatment. Table 103 describes the treatment
guidelines according to patient characteristics. In patients who are also
HIV positive, antiretroviral therapy may need to be modified, delayed,
or interrupted to complete an adequate course of therapy for HCV. The
effectiveness of treatment is assessed by following HCV viral load.
The goal of treatment is cure of HCV infection, manifested by reduction in hepatitis C viral load to undetectable, normalization of transaminases (alanine aminotransferase and aspartate transaminase), and
cessation of liver disease progression. Combination therapy with daily
oral ribavirin and once-weekly subcutaneous peginterferon alpha-2b is
the standard therapy for HCV (Deutsch and Hadziyannis 2008).
Exacerbation of psychotic symptoms during HCV treatment with interferon has been reported in sporadic cases, especially in patients with
a history of abuse of various drugs, including hallucinogens. Interferon
and ribavirin have potential neuropsychiatric side effects, including
apathy, cognitive changes, irritability, depression, psychosis, and suicidal thoughts. Prospective studies with psychiatric patients, including
people with schizophrenia (n=38), found that the response rates and adherence during treatment with standard interferon and ribavirin were
similar to those of patients in a nonpsychiatric control group (Schaefer et
al. 2007). Psychiatric patients were also not at increased risk of worsening
depression or psychosis during antiviral treatment compared with con-

HIV and Hepatitis C


TABLE 103. Hepatitis C virus (HCV) treatment guidelines

Patient characteristics

Treatment recommendations

HCV genotype 2 or 3 with

persistent HCV viremia

Treat with pegylated interferon

plus ribavirin for 24 weeks.
Liver biopsy is optional, but
treat regardless.

HCV genotype 1 with persistent

viremia and

liver biopsy showing

portal or bridging fibrosis,
compensated cirrhosis, or
moderate inflammation or

Treat with pegylated interferon

plus ribavirin if no contraindications (assess risks and benefits on a case-by-case basis).
If HCV viral load decreases
within the initial 612 weeks of
therapy, continue treatment
for 48 weeks; if no decrease,
discontinue treatment.


no or minimal fibrosis
on liver biopsy

Follow with observation, serial

alanine aminotransferase
measures, and liver biopsy
35 years.

Decompensated cirrhosis, any


Consider evaluation for liver


trols. Of note is that clinical trials using interferon as an add-on treatment

for patients with schizophrenia have shown improvement in psychotic
symptoms and reduction in the daily dosage of antipsychotic medication
needed (Freudenreich et al. 2007). Nonetheless, patients with depression,
psychotic disorders, or drug addiction should be carefully monitored for
the development of psychiatric symptoms during interferon treatment,
with interdisciplinary involvement to optimize adherence and response
rates and to manage potential side effects. The neurocognitive complications of HCV itself, together with the symptoms of liver failure, such as
fatigue, loss of appetite, loss of sexual drive, and impotence, can overlap
with the complications of interferon treatment, the symptoms of psychiatric disorders, and the side effects of psychotropic drugs.
Unfortunately, the majority of people with schizophrenia are not tested
for HCV, and among those who are tested and are diagnosed with chronic


Medical Illness and Schizophrenia

HCV infection, few receive treatment. Among a sample of 108 outpatients

with severe mental illness, only 32% reported having been previously
tested for HCV (Goldberg and Seth 2008). Of those newly identified as
HCV infected (n =18) within the study, half received an HCV-specific
medical follow-up, and only one person was started on treatment.

Clinical Considerations
Risk-Reduction Interventions and Strategies for
The most successful HIV prevention interventions are those addressing
perinatal transmission, which has been reduced to approximately 1% in
the United States with the effective use of antiretroviral treatment, and
may be further reduced by intrapartum delivery strategies (Jamieson et
al. 2007). However, a variety of interventions and strategies have been
recommended to lower the risk of transmitting HIV and HCV.
Harm reduction is the ultimate goal for both drug-related and sexual
risk behaviors. The use of prevention strategies for injection drug users
(e.g., syringe exchange and syringe availability at pharmacies) has reduced HIV and HCV transmission among injection drug users in the
United States and in a number of other countries (Des Jarlais and Semaan 2008). No studies have been done of interventions to address the
intermittent injection drug use that has been documented in the population with severe mental illness.
Engaging patients in efficacious individual and group interventions
to practice assertive behaviors and negotiation skills to increase selfprotective behaviors, including condom use, is necessary to stem new
infections in this highly affected population. The best predictor of using
a condom is having a condom. Institutional obstacles to condom acquisition are likely to impede patients initiative and ability to practice
safer sex. Making condoms available anonymously to all patients, including inpatients with off-ward privileges or who engage in consensual sex on the ward, has been shown to be a cost-effective primary
prevention intervention (Carmen and Brady 1990).
HIV prevention programs that primarily dispense AIDS information
have not been shown to influence risk behavior. As evidence indicates,
knowledge is necessary but not sufficient to produce behavioral changes.
Intensive, small-group programs that simultaneously target knowledge,
attitudes, motivations, and cognitive and behavioral skills have been
tried and found to produce reductions in high-risk sexual behaviors, including some that are substance related, among people with severe men-

HIV and Hepatitis C


tal illness. Effective elements from randomized controlled trials of these

HIV risk-reduction interventions (Wainberg et al. 2007) include 1) providing risk information; 2) enhancing awareness of attitudes, intentions,
and readiness for change; 3) acquiring and rehearsing sexual risk-reduction behavioral skills; 4) problem solving for handling triggers for sexual
risk taking; and 5) reinforcing behavior changes between intervention
sessions. Mental health programs need to provide an environment in
which healthy sexuality can be discussed. Patients can benefit from participating in mixedHIV serostatus prevention groups; participants need
not reveal their HIV status unless they wish to do so. Whatever the
groups composition, group leaders should leave time at the end of each
session to discuss patients personal issues privately and to address their
needs by making appropriate referrals, including for HIV testing.
Because HCV shares modes of transmission with HIV, groups focused on HIV risk reduction may also wish to impart information about
HCV infection through high-risk behaviors. Prevention efforts aimed at
HCV are essential for patients who are currently (or at risk for) injecting
drugs, because injection drug use is the risk factor responsible for a majority of HCV cases in the United States.
Once staff have received training to do prevention interventions,
they typically become motivated to start intervention groups for patients. Groups work best in longer-term day programs and outpatient
programs. On inpatient units, short lengths of stay may limit the number of sessions patients can attend, but that issue should not discourage
staff from setting up such programs.
Individual counseling can reinforce patients motivations to protect
themselves and others. Clinicians can encourage patients with HIV or
HCV to disclose their illness to drug-use partners or sex partners and to
use condoms. Fully informed decisions about risk and protection of
others are the goals.

Accessing HIV- and HCV-Related Services That

Psychiatric Patients Need
Mental health service settings vary in their ability to offer HIV- and
HCV-related screening, testing, and prevention interventions, and the
range of services available may not yet be meeting the needs of people
with severe mental illness (Goldberg and Seth 2008; Satriano et al. 2007).
Risk assessment for HIV or HCV is intended to elicit specific information about patients sexual behaviors and drug use. Most patients, when
asked in a direct, nonjudgmental way, are cooperative and forthcoming.
Rather than asking Have you ever..., asking How often have you...


Medical Illness and Schizophrenia

is more likely to elicit useful risk information without raising a patients defenses by implying that the clinician will judge such behaviors negatively
or consider them unusual. The ease with which the clinician is able to discuss sex and drug use will set the anxiety level for the patient, and normalizing any patient discomfort can create a more relaxed tone.
HIV testing is guided by regulations that vary from state to state, but
the CDC-recommended opt-out testing does not require counseling. To
better prepare patients for potential positive results, mental health programs may want to provide pretest and posttest counseling that is coordinated with medical care or conduct the testing themselves.
The initial diagnosis of HIV infection may occur when a patient first
becomes infected or has advanced AIDS, or any time in between. Shock
and disbelief may be followed by depression, anxiety, and fear in adjusting to having contracted a serious and potentially deadly illness.
Like severe mental illness, HIV and AIDS can be highly stigmatizing,
possibly resulting in rejection, abandonment, and further social isolation. If a worsening of psychiatric symptoms follows the initial HIV diagnosis, the most effective intervention is individual counseling and
supportive therapy geared to the patients current mental status and his
or her knowledge and understanding of HIV infection.
Legal regulations or guidelines for confidentiality and the disclosure
of HIV- or AIDS-related information set the stage for upholding humane and responsible individual and public health standards. Contact
notification by physicians may not be required, but legal statutes may
allow doctors or public health officers to inform a contact who is at significant risk if they determine that a patient will not do so. Laws that are
applicable to the locality should be consulted when making decisions
that involve confidentiality and contact notification.
For HIV-infected psychiatric patients who are asymptomatic, supportive groups may encourage behavioral change and promote ways to
preserve physical health in the community and within the psychiatric
setting. This kind of group intervention can prevent worsening of psychiatric symptoms and provide a sense of community that can decrease
social isolation, reinforce safer peer norms, and encourage altruism,
which appeals to ego strengths and gives patients a sense of worth and
Services may be directed at patients on the basis of their suspected or
presumed risks rather than thorough individual risk assessments (Brunette et al. 2000; McKinnon et al. 1999). This model of risk evaluation is
typically relied upon when there are unmet needs for staff training.
Training is one important way to improve service delivery systems.

HIV and Hepatitis C


Testing for HCV can occur at any point and does not require special
consent. Routine screening for HCV, followed by referral for evaluation
of every HCV-positive patient, including those with schizophrenia
whose psychiatric condition is stable, is warranted. Clinicians need to
explain the test result to patients found to be HCV positive and to refer
such patients for further evaluation and, as appropriate, hepatitis A and
B immunization.
Clinicians may be in the best position to help their patients access
HIV and HCV testing, treatment, and prevention opportunities. They
can employ many strategies to help their patients determine their risk
of acquiring or transmitting HIV or HCV, prevent new infections, promote healthier behaviors, and reduce the impact of HIV- and HCVrelated illness on this vulnerable population. Clinicians also are
uniquely qualified to help their patients manage the many medications
required to maintain their psychiatric and physical health.

Psychopharmacology for People With HIV or

HCV and Schizophrenia
Important points to keep in mind when prescribing psychotropic medication to people with schizophrenia who have comorbid HIV or HCV
infection are listed in Table 104. Notably, as HIV progresses, people
with preexisting psychiatric conditions may develop new or increased
side effects when taking medications they were previously able to tolerate.
In the course of the medical management of HIV infection, particularly in later stages of HIV disease, a large number of medications may
be prescribed, including antibacterial, antifungal, antineoplastic, antiretroviral, and other antiviral agents (Wainberg et al. 2000). Any attempt to diagnose drug-induced neuropsychiatric syndromes requires
an appreciation of both the therapeutic use and potential side effects of
these medications. Some of these are described in the Practice Guidelines of the American Psychiatric Association (American Psychiatric
Association 2000), but new antiretroviral agents are being developed at
a rapid pace and existing medications to treat HIV-related infections
and neoplasms are too numerous to describe fully. Neuropsychiatric
side effects of antiretrovirals have been reported most commonly with
efavirenz and occasionally with nevirapine. Despite predictions of
drug-drug interactions (which can be found in handheld databases,
through hospital pharmacies, and at a variety of Web sites), experienced prescribers generally find they can effectively use the full range


Medical Illness and Schizophrenia

of psychotropic medications. Overlapping toxicities between different

medications must also be taken into consideration. Because advanced
HIV infection is associated with greater sensitivity to both the therapeutic effects and side effects of psychotropic medications, the best approach is to start a patient taking the lowest possible dosage and
increase the dosage slowly. Drug levels, if available, should be monitored closely, especially when patients are on complex medication regimens. Methadone, recreational drugs, and herbal preparations also can
affect medication levels.
Most psychiatric medications are metabolized by the liver and therefore may require more careful monitoring in patients chronically infected with HCV. In particular, for patients who manifest clinical or
laboratory signs of liver failure, medication metabolism can be dangerously reduced such that the patients may accumulate toxic levels of
drugs at dosages they were previously able to tolerate.
In general, the use of most psychiatric medications is relatively safe
with both HIV and HCV treatments. For patients with HCV illness, periodic liver function tests are the standard of care. Because some psychotropic medications (e.g., antidepressants, lithium, valproic acid,
carbamazepine) may elevate liver enzymes in HCV-infected patients, it
is important to check these at baseline, early after initiation of therapy
(after 24 weeks), and every 23 months thereafter. However, alanine
aminotransferase elevations may be due to fluctuations in the HCV itself or other factors (e.g., unacknowledged alcohol use) (Felker et al.
2003). Combined toxicities are important to consider, particularly during treatment with medications that can have an impact on bone marrow activity (e.g., clozapine, carbamazepine, zidovudine, interferon,
ribavirin) or cause other additive side effects.
Psychiatric patients, like medical patients, are often nonadherent to
medication, which they may perceive as one of the few aspects of their
lives they can control. It is important to time the beginning of an antiviral regimen with a commitment to the treatment and to help patients
see that following an HIV or HCV medication regimen can be part of
gaining control. Working with a patient to promote adherence to psychotropic medications will be the best predictor of whether the patient
can follow an HIV or HCV medication regimen. The clinician and infectious disease specialist must communicate about the patients need for
and readiness to begin antivirals; select, in concert with the patient, regimens that minimize drug-drug interactions; keep the number and dosages of pills the patient has to take to the minimum necessary; and
coordinate their treatment plans.

HIV and Hepatitis C


TABLE 104. Psychotropic medication in patients with schizophrenia

with comorbid HIV and/or HCV infection
Medically asymptomatic and not receiving antiviral treatment
Most psychotropics can be used as usual.
Monitor use of typical antipsychotics for increased incidence of
extrapyramidal side effects.
Evaluate for testosterone deficiency if depressed or fatigued.
Whenever possible, avoid drugs that cause liver toxicity.
Follow liver enzymes.
Medically ill and/or receiving antiviral treatment
For most psychotropics, start with low doses and increase slowly.
Check for interactions and overlapping toxicities between
psychotropics and antiretrovirals. Note that protease inhibitors are
potent inhibitors of one or more cytochrome P450 enzymes, which
may have implications for interactions with psychiatric and other
Monitor use of typical antipsychotics for severe and rapid onset of
extrapyramidal side effects.
Monitor use of atypical antipsychotics with antiretroviral
medications, which overlap in causing metabolic abnormalities.
Avoid carbamazepine, which lowers antiretrovirals.
Avoid lithium with HIV-associated nephropathy.
Consider that antianxiety drugs metabolized by glucuronidation
have fewer drug interactions with antiretrovirals (i.e., clorazepate,
lorazepam, oxazepam, temazepam).
Evaluate for testosterone deficiency if depressed or fatigued.
Be aware that antiretrovirals alter levels of methadone and
buprenorphine, usually by decreasing them. Check effects of
specific antiretroviral regimens and modify doses as needed.
Caution against St. Johns wort, which lowers antiretrovirals.
Avoid drugs that cause liver toxicity.
Follow liver enzymes.
Monitor for depression and worsening of psychotic symptoms
while on antiviral therapy; treat as necessary.
Source: American Psychiatric Association 2000 and


Medical Illness and Schizophrenia

Understanding the HIV and HCV epidemics and their underpinnings
among people with severe mental illness can help clinicians to recognize when their patients with schizophrenia are at risk of acquiring or
transmitting these viruses and to intervene appropriately and effectively with preventive, counseling and testing, medical, and supportive
services. It is critical for clinicians to receive training to enhance their
knowledge and skills in these areas, to ask comfortably and nonjudgmentally about their patients risks rather than wait until patients are ill
to learn about risk behaviors, and to continually access up-to-date information and technical assistance about how to prevent and treat these
infections among their patients.

Key Clinical Points

People with schizophrenia are at heightened risk for acquiring HIV and
HCV infection, and both of these infections can result in increased
morbidity and mortality.

Both HIV and HCV enter the central nervous system and can cause
neurocognitive symptoms that complicate the psychiatric presentations
and treatment of people with schizophrenia.

Improvements in the treatments for HIV and HCV infections warrant

assertive efforts to identify patients who are infected.

Testing for HIV and HCV infection and offering prevention interventions should be part of routine care for people with schizophrenia.

Patients who test positive for either HIV or HCV should routinely be
referred to medical providers who can assess and treat these conditions.
Mental health care providers need to stay involved to ensure patient
adherence to medical care and help manage mental health complications of treatment.

Mental health providers often need to advocate to ensure that people

with schizophrenia are not rejected from HIV and HCV treatment
solely because they have a mental illness.

HIV and Hepatitis C


Agency for Healthcare Research and Quality: Management of chronic hepatitis
C. June 2002. Available at:
Accessed October 25, 2008.
Al Jurdi RK, Burruss JW: Prevalence of hepatitis C in psychiatric institutions.
Psychosomatics 44:439440, 2003
American Psychiatric Association: APA practice guidelines: treatment of patients
with HIV/AIDS. November 2000. Available at: http://www.psychiatryonline
.com/pracGuide/pracGuideChapToc_4.aspx. Accessed October 25, 2008.
Antinori A, Arendt G, Becker JT, et al: Updated research nosology for HIVassociated neurocognitive disorders. Neurology 69:17891799, 2007
Barre-Sinoussi F, Chermann JC, Rey F, et al: Isolation of a T-lymphotropic retrovirus from a patient at risk for acquired immunodeficiency syndrome.
Science 220:868871, 1983
Branson BM, Handsfield HH, Lampe MA, et al; Centers for Disease Control and
Prevention (CDC). Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR Recomm
Rep 55:117, 2006
Brunette MF, Mercer CC, Carlson CL, et al: HIV-related services for persons
with severe mental illness: policy and practice in New Hampshire community mental health. J Behav Health Serv Res 27:347353, 2000
Butterfield MI, Bosworth HB, Meador KG, et al: Gender differences in hepatitis
C infection and risks among persons with severe mental illness. Psychiatr
Serv 54:848853, 2003
Butterfield MI, Bosworth HB, Stechuchak KM, et al: Racial differences in hepatitis B and hepatitis C and associated risk behaviors in veterans with severe
mental illness. J Natl Med Assoc 96:4352, 2004
Carey MP, Carey KB, Kalichman SC: Risk for human immunodeficiency virus
(HIV) infection among persons with severe mental illnesses. Clin Psychol
Rev 17:271291, 1997
Carmen E, Brady SM: AIDS risk and prevention for the chronic mentally ill.
Hosp Community Psychiatry 41:652657, 1990
Centers for Disease Control and Prevention, Division of HIV/AIDS Prevention,
National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention:
Revised surveillance case definitions for HIV infection among adults, adolescents, and children aged <18 months and for HIV infection and AIDS
among children aged 18 months to <13 yearsUnited States, 2008. MMWR
Morb Mortal Wkly Rep 57(RR10):18, 2008
Clarke A, Kulasegaram R: Hepatitis C transmission: where are we now? Int J
STD AIDS 17:7480, 2006
Cournos F, McKinnon K: HIV seroprevalence among people with severe mental
illness in the United States: a critical review. Clin Psychol Rev 17:259269,


Medical Illness and Schizophrenia

de Ldinghen V, Barreiro P, Foucher J, et al: Liver fibrosis on account of chronic

hepatitis C is more severe in HIV-positive than HIV-negative patients despite antiretroviral therapy. J Viral Hepat 15:427433, 2008
Des Jarlais DC, Semaan S: HIV prevention for injecting drug users: the first 25
years and counting. Psychosom Med 70:606611, 2008
Deutsch M, Hadziyannis SJ: Old and emerging therapies in chronic hepatitis C:
an update. J Viral Hepat 15:211, 2008
Dinwiddie SH, Shicker L, Newman T: Prevalence of hepatitis C among psychiatric patients in the public sector. Am J Psychiatry 160:172174, 2003
Felker BL, Sloan KL, Dominitz JA, et al: The safety of valproic acid use for patients with hepatitis C infection. Am J Psychiatry 160:174178, 2003
Fidler S, Fox J, Porter K, et al: Primary HIV infection: to treat or not to treat?
Curr Opin Infect Dis 21:410, 2008
Freudenreich O, Gandhi RT, Walsh JP, et al: Hepatitis C in schizophrenia:
screening experience in a community-dwelling clozapine cohort. Psychosomatics 48:405411, 2007
Goldberg RW, Seth P: Hepatitis C services and individuals with serious mental
illness. Community Ment Health J 44:381384, 2008
Huckans MS, Blackwell AD, Harms TA, et al: Management of hepatitis C disease among VA patients with schizophrenia and substance use disorders.
Psychiatr Serv 57:403406, 2006
Hulgan T, Shepherd BE, Raffanti SP, et al: Absolute count and percentage of
CD4+ lymphocytes are independent predictors of disease progression in
HIV-infected persons initiating highly active antiretroviral therapy. J Infect
Dis 195:425431, 2007
Jamieson DJ, Read JS, Kourtis AP, et al: Cesarean delivery for HIV-infected
women: recommendations and controversies. Am J Obstet Gynecol 197
(suppl 3): S96S100, 2007
Kalichman SC, Kelly JA, Johnson JR, et al: Factors associated with risk for HIV infection among chronic mentally ill adults. Am J Psychiatry 151:221227, 1994
Kelly JA, Murphy DA, Bahr GR, et al: AIDS/HIV risk behavior among the
chronic mentally ill. Am J Psychiatry 149:886889, 1992
Klinkenberg WD, Caslyn RJ, Morse GA, et al: Prevalence of human immunodeficiency virus, hepatitis B, and hepatitis C among homeless persons with
co-occurring severe mental illness and substance use disorders. Compr
Psychiatry 44:293302, 2003
Letendre SL: Neurocognitive effects of HCV, methamphetamine abuse, and HIV:
multiple risks and mechanisms. Ann Gen Psychiatry 7 (suppl 1):S39, 2008
Malow RM, Devieux JG, Martinez L, et al: History of traumatic abuse and HIV
risk behaviors in severely mentally ill substance abusing adults. J Fam Violence 21:127135, 2006
Marks G, Crepaz N, Janssen RS: Estimating sexual transmission of HIV from
persons aware and unaware that they are infected with the virus in the
USA. AIDS 20:14471450, 2006

HIV and Hepatitis C


McDermott BE, Sautter FJ, Winstead DK, et al: Diagnosis, health beliefs, and
risk of HIV infection in psychiatric patients. Hosp Community Psychiatry
45:580585, 1994
McKinnon K, Cournos F: HIV infection linked to substance use among hospitalized patients with severe mental illness. Psychiatr Serv 49:1269, 1998
McKinnon K, Cournos F, Sugden R, et al: The relative contributions of psychiatric symptoms and AIDS knowledge to HIV risk behaviors among people
with severe mental illness. J Clin Psychiatry 57:506513, 1996
McKinnon K, Cournos F, Herman R, et al: AIDS-related services and training in
outpatient mental health care agencies in New York. Psychiatr Serv
50:12251228, 1999
Meade CS, Sikkema KJ: HIV risk behavior among adults with severe mental illness: a systematic review. Clin Psychol Rev 25:433457, 2005
Meade CS, Sikkema KJ: Psychiatric and psychosocial correlates of sexual risk
behavior among adults with severe mental illness. Community Ment
Health J 43:153169, 2007
Menon AS, Pomerantz S: Substance use during sex and unsafe sexual behaviors
among acute psychiatric inpatients. Psychiatr Serv 48:10701072, 1997
Meyer JM: Prevalence of hepatitis A, hepatitis B and HIV among hepatitis C seropositive state hospital patients: results from Oregon State Hospital. J Clin
Psychiatry 64:540545, 2003
Otto-Salaj LL, Heckman TG, Stevenson LY, et al: Patterns, predictors and gender differences in HIV risk among severely mentally ill men and women.
Community Ment Health J 34:175190, 1998
Perry W, Hilsabeck RC, Hassanein TI: Cognitive dysfunction in chronic hepatitis C: a review. Dig Dis Sci 53:307321, 2008
Rosenberg SD, Goodman LA, Osher FC, et al: Prevalence of HIV, hepatitis B,
and hepatitis C in people with severe mental illness. Am J Public Health
91:3137, 2001
Satriano J, McKinnon K, Adoff S: HIV service provision for people with severe
mental illness in outpatient mental health care settings in New York. J Prev
Interv Community 33:95108, 2007
Schaefer M, Hinzpeter A, Mohmand A, et al: Hepatitis C treatment in difficultto-treat psychiatric patients with pegylated interferon-alpha and ribavirin: response and psychiatric side effects. Hepatology 46:991998, 2007
Soogoor M, Daar ES: Primary human immunodeficiency virus type 1 infection.
Curr HIV/AIDS Rep 2:5560, 2005
Soriano V, Puoti M, Garcia-Gasc P, et al: Antiretroviral drugs and liver injury.
AIDS 22:113, 2008
Staprans SI, Feinberg MB: Natural history and immunopathogenesis of HIV-1
disease, in The Medical Management of AIDS, 5th Edition. Edited by Sande
MA, Volberding PA. Philadelphia, PA, WB Saunders, 1997, pp 2956
Tabibian JH, Wirshing DA, Pierre JM, et al: Hepatitis B and C among veterans
on a psychiatric ward. Dig Dis Sci 53:16931698, 2008


Medical Illness and Schizophrenia

UNAIDS: 2007 AIDS epidemic update. 2007. Available at: http://www.unaids

Accessed October 25, 2008.
Wainberg ML, Forstein M, Berkman A, et al: Essential medical facts for mental
health practitioners, in What Mental Health Practitioners Need to Know
About HIV and AIDS. Edited by Cournos F, Forstein M . San Francisco, CA,
Jossey-Bass, 2000
Wainberg ML, McKinnon K, Mattos P, et al: Is it Brazilian? A model for adapting evidence-based behavioral interventions to a new culture: HIV prevention for psychiatric patients in Rio de Janeiro, Brazil. AIDS Behav 11:872
883, 2007
Walker BD, Burton DR: Towards an AIDS vaccine. Science 320:760764, 2008

Substance Abuse and
Peter F. Buckley, M.D.
Jonathan M. Meyer, M.D.

The co-occurrence of schizophrenia and substance use or abuse is

common and presents complex issues for mental health practitioners.
Approximately 50% of individuals with schizophrenia will develop substance use disorders at some point during their lives, and about half of
this group will exhibit current substance abuse or dependence (Buckley
2006). The abuse of drugs and alcohol by persons with severe mental illnesses has a wide range of adverse effects on the course of mental illness
and psychosocial functioning, including compliance, prognosis, and
rates of acute service utilization. Overall, substance abuse comorbidity
substantially complicates the course and management of schizophrenia.
In this chapter, we provide a brief overview of the scope of the problem,
with a detailed focus on the demographics of substance use, the adverse
consequences of the major drugs of abuse in patients with schizophrenia,
and the clinical approach to patients with dual diagnoses.

Demographics of Substance Use in

Patients With Schizophrenia
The proportion of patients with schizophrenia who have a comorbid
drug or alcohol use disorder varies tremendously in published studies,


Medical Illness and Schizophrenia

from as low as 10% to as high as 70% (Mueser et al. 1990). This wide
range is partially due to variability in the diagnostic criteria employed
for schizophrenia, sample demographic characteristics (e.g., male vs.
female, urban vs. rural), types of patient populations studied (e.g., inpatient vs. outpatient), and criteria for defining drug and alcohol disorders (e.g., Diagnostic and Statistical Manual of Mental Disorders [DSM]
diagnosis, positive urine toxicology screens, rating scales) (Mueser et al.
1990). Although structured clinical interviews have been found to produce the most reliable diagnoses, these are time consuming and expensive due to the need for trained personnel and, therefore, are not often
used in clinical studies (Mueser et al. 1995). Surveys conducted exclusively in inpatient settings tend to produce higher rates of substance use
disorders, in part because persons with dual disorders (i.e., substance
use and another major Axis I disorder) are more likely to enter into
treatment because exacerbation of either problem may become a focus
of clinical attention (Mueser et al. 1995).
The Epidemiologic Catchment Area (ECA) study revealed that 47%
of all individuals in the United States with a lifetime diagnosis of
schizophrenia or schizophreniform disorder met criteria for some form
of substance abuse or dependence (33.7% for alcohol disorder and
27.5% for another drug abuse disorder) (Regier et al. 1990). In addition
to the lifetime prevalence data, the ECA study found that the odds of
having a substance abuse diagnosis were 4.6 times greater for persons
with schizophrenia compared with the rest of the population, with the
odds of having an alcohol disorder over 3 times greater and of having
another drug disorder 6 times greater (Regier et al. 1990). This finding
was replicated in the National Comorbidity Survey (Kessler et al. 1997)
and again in the most recent National Comorbidity Survey Replication
(Merikangas et al. 2007). The high rate of comorbidity is comparable
with the range of 40%60% lifetime prevalence gleaned from an analysis of 47 published studies with sample sizes of at least 30 patients with
schizophrenia and in which the criteria for abuse or dependence were
clearly delineated (Cantor-Graae et al. 2001). Cantor-Graae et al. (2001)
also found that studies in which more than one method of diagnosis
was employed (e.g., chart review plus interview) yielded higher prevalence rates compared with those that relied on a single method. Regardless of method, in community samples of patients with schizophrenia
that are not comprised solely of inpatient groups, a figure of approximately 50% lifetime prevalence is found repeatedly. Additionally, studies from all over the world confirm a strong association between
schizophrenia and substance abuse (ODely et al. 2005). For example, a
Swedish study combining both structured interview and chart review

Substance Abuse


of 87 patients with schizophrenia noted a lifetime prevalence of any

substance use disorder of 48.3% (47.1% for alcohol alone or in combination with other drugs, 26.4% for alcohol alone) (Cantor-Graae et al.
2001). A British study of 168 patients prospectively evaluated at the
time of their first episode of psychosis found that 37% of patients were
abusing drugs or alcohol (Cantwell et al. 1999).
Nicotine is clearly the most frequently abused agent by patients with
schizophrenia, with prevalences ranging from 70% to 90%, over three
times greater than use by the general population (De Leon et al. 2002,
2007; see also Chapter 9, Nicotine and Tobacco Use in Patients With
Schizophrenia). Excluding nicotine use, the ECA study results are consistent with other prevalence studies of schizophrenia demonstrating
that alcohol tends to be the most frequently abused agent (20%60%),
followed by cannabis (12%42%) and cocaine (15%50%) (Chambers et
al. 2001). Use of amphetamines (2%25%), hallucinogens, opiates, and
sedatives/hypnotics is less common. Findings from a British firstepisode study (Barnett et al. 2007) indicated that 43% of patients abused
alcohol, 51% abused cannabis, and 38% had a pattern of polysubstance
abuse. Archie et al. (2007) reported a small shift in these abuse patterns
when patients are followed over time. Harrison et al. (2008) reported an
almost 50% decline in overall substance use, except smoking (which
rose from 60% to 64%) when patients are reassessed after 1 year. Aside
from nicotine, alcohol and marijuana were the most frequently abused
substances, and this also remained constant after 14 months of mean
follow-up. In the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study, marijuana was the most commonly used drug by
patients with schizophrenia, followed by cocaine and opiates (Swartz et
al. 2008). Polysubstance abuse was also common in the CATIE sample.
The most consistent findings with regard to demographic characteristics are that those who are younger, were younger at age of schizophrenia onset, or are male are more likely than those who are older and
female to abuse drugs or alcohol (Barnett et al. 2007; Hambrecht and
Hafner 2000). Males were more frequent among the abuser group (81%)
than among the nonabuser group (69%) in the CATIE study (Swartz et
al. 2008). It is important to note, however, that evidence is accumulating
to document that substance use difficulties among women are not sufficiently recognized, and that women with schizophrenia and comorbid
substance disorders are less likely to receive substance abuse treatment
(Alexander 1996; Comtois and Ries 1995). The undertreatment of
women may lead to unrealistically low prevalence estimates in retrospective chart review studies, yet even in interview studies male gender
appears as an independent risk factor after the disparities in gender


Medical Illness and Schizophrenia

frequency are controlling for among study participants (Cantor-Graae

et al. 2001). Also, evidence from a Malmo, Sweden, cohort (CantorGraae et al. 2001) and other studies indicates that the onset of substance
use in schizophrenia occurs at a younger age in males than in females
(ODely et al. 2005; Salyers and Mueser 2001).
Conflicting data have been reported regarding the functional capabilities of substance-using patients compared with non-substanceusing patients with severe mental illness. In one of the first studies to
examine this issue, Mueser et al. (1990) reported findings on 149 recently hospitalized patients who met DSM-III-R criteria for schizophrenia, schizoaffective disorder, or schizophreniform disorder, and found
lower educational levels among those with substance abuse disorders
than among those without substance use disorders. Arndt et al. (1992),
in a study of 131 schizophrenia patients with and without comorbid
substance abuse disorder (n = 64 and 67, respectively) matched for
symptomatology and clinical history, noted better premorbid adjustment among the so-called pathological substance users (Arndt et al.
1992). In another study, Zisook et al. (1992) compared 34 patients with
schizophrenia who had histories of substance abuse with 17 patients
with schizophrenia who were abstinent and found that the substance
abusers were more likely to have been married or gainfully employed.
In the CATIE study (Swartz et al. 2008), abusers were overall similar in
terms of marital status and years of education. They were also similar
on overall functioning, based on their Clinical Global Impression Scale
Salyers and Mueser (2001) examined 404 patients with a history of
recent hospitalization (i.e., prior 3 months) recruited from within a
large multicenter psychosocial treatment strategies study. Individuals
were ages 1855, agreed to receive fluphenazine decanoate but not
other major psychotropic agents, and received at least three of the
monthly assessments of psychiatric symptoms, social functioning, side
effects, and substance use over the 3- to 6-month follow-up period. Notably, those who were homeless or transient and patients with active
ongoing physical dependence (or suspected drug-induced psychosis
but not schizophrenia) were excluded from the study. In this study,
those patients who consistently reported low or no use of substances
scored significantly higher than regular drug or alcohol users on assessments of negative symptoms, particularly the social amotivation and
diminished expression scores, although the groups were comparable
on ratings of psychotic symptoms (Brief Psychiatric Rating Scale) and
distress (Symptom Checklist90) (Salyers and Mueser 2001). Also, users
and nonusers had no significant differences in ratings of tardive dyski-

Substance Abuse


nesia, parkinsonism, or akathisia, although those who reported the

greatest frequency of social problems had higher akathisia ratings irrespective of degree of substance use. As one might expect from those
who manifest greater negative symptomatology, the low- or no-use
group also demonstrated more severe impairment in leisure activities
and less frequent social contacts. Although the substance users enjoyed
a higher level of social functioning, the drug users in particular reported significantly greater interpersonal problems compared to the
low- or no-use group. Interestingly, despite the greater social functional
status of substance users, this group had an earlier age of onset for their
mental illness and more hospitalizations than the low- or no-use group.
In the CATIE study (Swartz et al. 2008), symptom measures were similar between abusers and nonabusers, although abusers had more prior
illness relapses and an earlier age at onset.
Demographic differences exist between urban and rural groups with
schizophrenia, especially with regard to housing, and data also suggest
that patterns of substance use differ between the two groups. The use
of alcohol alone or in combination with cannabis is common among
schizophrenia patients living in a rural setting, whereas the use of multiple substances, particularly cocaine, is common among those living in
urban settings (Mueser et al. 2001).
Also, patients with a dual diagnosis of schizophrenia and substance
use or abuse tend to engage in a variety of risky behaviors that result in
a higher prevalence of human immunodeficiency virus (HIV), hepatitis
B, and hepatitis C (Meyer 2003; Wirshing et al. 2005; also see Chapter
10, HIV and Hepatitis C in Patients With Schizophrenia, in this volume). Whereas the former may be transmitted sexually, hepatitis C in
the United States is transmitted primarily by use of shared needles
among intravenous drug users. Although patients with schizophrenia
are more likely to engage in high-risk behavior resulting in HIV or hepatitis C infection, limited data are available about the extent of intravenous drug abuse by individuals with this disorder. In one large study
of persons with mental illness (65% of whom had schizophrenia or
schizoaffective disorder), 62.1% of those infected with hepatitis C
(n =145) reported a lifetime history of intravenous drug abuse (IVDA),
whereas only 5.1% of the hepatitis Cnegative group (n =604) reported
a history of IVDA (Rosenberg et al. 2001). Among the HIV-positive persons with severe mental illness, the lifetime prevalence of IVDA was
8.6% compared with 1.4% for the HIV-negative group (Rosenberg et al.
2001). A smaller study of 91 patients with schizophrenia employing two
self-reporting measures found a 22.4% lifetime prevalence of injected
drug use; moreover, despite IVDA and high-risk sexual behaviors, 65%


Medical Illness and Schizophrenia

reported no concern with HIV infection, and AIDS knowledge was significantly lower among the patients with schizophrenia than the control
group, particularly among those with long-standing illness and multiple
psychiatric admissions (Grassi et al. 1999). Meyer (2003) showed a high
rate of hepatitis exposure in long-stay patients at a state hospital. Wirshing et al. (2005) showed similar findings in the Veterans Administration
dual-diagnosis population (schizophrenia plus substance abuse).

Patterns of Substance Abuse Among

Persons With Schizophrenia
Although substance use disorders occur significantly more frequently
in persons diagnosed with schizophrenia than in the general population, little evidence is available to suggest that persons with schizophrenia abuse substances for different reasons than the general population.
Specifically, the literature demonstrates that patients with schizophrenia do not differentially choose to abuse specific drugs to ameliorate specific psychic states. Multiple studies indicate that patients with
schizophrenia do not preferentially abuse certain agents and that choice
of agent does not correlate with extent of psychopathology; rather,
these individuals use those substances that are most available and affordable (Cantor-Graae et al. 2001; Chambers et al. 2001; Mueser et al.
1992). For the most part, research has also failed to find different patterns of drug choice between persons with schizophrenia and groups of
patients with other major mental disorders. Thus, although the causes
of dysphoria in schizophrenia may be linked to the illness (e.g., neuroleptic side effects, negative symptoms, demoralization), data simply do
not support the notion that patients with schizophrenia uniquely prefer
certain drugs (Degenhardt et al. 2007).
The notion that schizophrenia patients abuse substances to alleviate
certain mental states or medication side effects (the self-medication
hypothesis) derives from research focusing on patients reported
reasons for substance use. Dixon et al. (1991) found that persons with
schizophrenia abuse drugs for reasons that are similar to those for
persons without schizophrenia (i.e., to get high, to feel better, to escape,
and to be less depressed), rather than for symptom control. Other observational data lend credence to the hypothesis that addictive behavior occurs independently as an inherent aspect of the neural
abnormalities that contribute to schizophrenia, as opposed to being motivated or driven secondarily by symptoms of the disorder or negative
psychic states induced by medication. In a review of the literature on

Substance Abuse


the biological basis for substance abuse in schizophrenia, Chambers et

al. (2001) noted that some individuals do report symptomatic improvement during substance use, whereas others note symptom exacerbation
but nevertheless persist in the use of substances. Moreover, self-reports
of improvement are often at variance with the clinical observation of
symptomatic worsening. In addition, substances of abuse have widely
divergent effects on neurotransmitters, thereby weakening the assertion that the concurrent use of agents with opposing effects is intended
to ameliorate a specific psychic state.
Dervaux et al. (2001) used diagnostic interview to assess the reasons
for and extent of substance use in a study of 100 patients with schizophrenia. Other assessment tools in this study included the Positive and
Negative Syndrome Scale (PANSS), self-report measures of impulsivity
and sensation seeking, and a self-report measure of anhedonia. Fortyone percent of this cohort met lifetime criteria for substance abuse or
dependence, but this group did not differ from the nonabusing patients
with schizophrenia on the basis of ratings of anhedonia, or in symptom
severity (as rated by PANSS). Interestingly, this study did not find differences between users and nonusers on the basis of age of first psychiatric contact and number of hospitalizations. Nonetheless, the
substance abusers rated significantly higher on the measures of impulsivity and sensation seeking, in a manner consistent with substance users without schizophrenia; moreover, these measures were elevated
even among the subgroup reporting only past abuse or dependence,
leading the authors to suggest that impulsivity and drug-seeking behavior may not be induced by the use of substances, but rather are inherent aspects of schizophrenia or correspond to the traits of novelty
seeking and impulsivity that are seen in the general population who
abuse drugs. Indeed, Swann et al. (2004) invoked a model of heightened
impulsivity as well as behavioral sensitization to explain the high rates
of comorbid substance abuse in patients with bipolar disorder. Notably,
some studies have found higher reported rates of depression among
patients with comorbid substance abuse (Duke et al. 1994; Hambrecht
and Hafner 2000), although others have not replicated this finding
(Drake et al. 1989; Mueser et al. 1990). Recently, Degenhardt et al. (2007)
found no difference in depression between schizophrenia patients with
and those without cannabis abuse. Additionally, cannabis abuse was
not associated with worsening of depressive symptoms.
The research on antipsychotic side effects and substance abuse has
generally produced mixed results when investigators examined tardive
dyskinesia or other measures of extrapyramidal side effects. Akathisia
and related dysphoria have been found in some studies to be associated


Medical Illness and Schizophrenia

with current alcohol use and increased risk of future alcohol use (Duke
et al. 1994; Voruganti et al. 1997). However, these reasons may be less
relevant in the current era of treatment with second-generation antipsychotic medications, which are generally considered to have low rates of
extrapyramidal side effects and to have mood stabilization and enhancing properties (Green et al. 2008).
Other observations suggesting that schizophrenia and addictive behavior are independent manifestations of common underlying dysregulation of neural circuitry are that use of both drugs and alcohol
commonly precedes psychosis and that 77% of first-episode patients are
smokers prior to neuroleptic treatment (Chambers et al. 2001). The relationship between substance use and symptom onset is thus an area of
intense scrutiny for those who see the predilection toward psychosis
and addictive behavior as independent but parallel processes engendered by common neuropsychiatric deficits.
In the literature review cited earlier, Cantor-Graae et al. (2001) noted
two questions involving schizophrenia and substance use that are not
clearly answerable with the current body of data. The first is the extent
to which substance use (especially psychoactive substance use) contributes to the development of schizophrenia, whether as an independent
risk factor in itself or by precipitating illness onset in vulnerable individuals (p. 72). The second, discussed in the following section, is the
degree to which history of substance abuse is associated with a more
chronic clinical course of schizophrenia (p. 72).
Because comorbidity is not merely a chance co-occurrence of two
disorders and because self-relief of symptoms and side effects also appears to be an inadequate explanation of drug use by people with
schizophrenia, a reasonable question is whether drug use actually
causes schizophrenia. This is a vexing question (and often comes across
as a chicken and egg problem), especially at the first presentation of
psychosis, where substance abuse is common. Moreover, studies have
produced some evidence that patients at high clinical risk for psychosis
who abuse drugs during the vulnerable prodromal stage are at great
risk of conversion to frank psychosis (Cannon et al. 2008). Caton et al.
(2005) examined this relationship among patients who were presenting
with their first psychotic episode, all of whom were actively abusing
drugs. On the basis of longitudinal evaluation, the authors reported
that key predictors can help determine whether a patient has substanceinduced psychosis or primary schizophrenia. In the longitudinal evaluation of patients who presented as acutely psychotic, 44% were later
classified as having a drug-induced psychosis and 56% received a diagnosis of schizophrenia. Patients with a drug-related psychosis had

Substance Abuse


fewer positive and negative symptoms but more visual hallucinations.

They were also more likely to have a family history of substance abuse.
Mathias et al. (2008) pointed out that little is known about substanceinduced psychosis. This nosology may get an extreme makeover in
the workings toward DSM-V (see Mathias et al. 2008).
Because use of substances, particularly psychotomimetic agents
(e.g., hallucinogens) and stimulants, may increase the likelihood of developing psychosis, investigators have attempted to determine the
course of substance use in relation to onset of psychosis. To assess the
temporal sequence of substance abuse and illness onset, Buhler et al.
(2002) performed a structured interview of 232 first-episode patients
with schizophrenia, and then prospectively interviewed and followed
another sample of 115 first-episode patients representing 86% of consecutive admissions in the local area. The investigators found that 62%
of those patients with drug abuse and 51% with alcohol abuse began
their habit before any signs of schizophrenia were manifest, including
prodromal nonpsychotic symptoms. Importantly, no correlation was
found between onset of substance abuse and onset of psychotic symptoms, although the onset of abuse and the psychotic disorder occurred
in the same month in 18.2% who abused alcohol and 34.6% who used
drugs, implying that in a subset of schizophrenia patients, development
of psychotic symptoms was speeded up or precipitated by substance
use, particularly cannabis use. These data from the combined pools of
first-episode patients are similar to those reported for the original
group of 232 first-episode patients analyzed separately (Hambrecht
and Hafner 2000), of whom 27.5% had a cannabis use problem more
than 1 year (often more than 5 years) before onset of prodromal schizophrenia symptoms, 34.6% had the onset of symptoms and cannabis
use in the same month, and 37.9% had symptoms of schizophrenia
before beginning substance use. Archie et al. (2007) illustrated the opportunity to reduce the rates of substance abuse early in the course of

Impact of Substance Use Disorders on

Development and Course of Illness
Some researchers have attempted to determine which drugs, if any,
are most likely to affect the development and course of schizophrenia.
Despite having other negative medical consequences, alcohol does
not appear to increase the risk of developing schizophrenia. The evidence for illicit drugs such as cocaine, heroin, methamphetamines, and


Medical Illness and Schizophrenia

3,4-methylenedioxymethamphetamine (ecstasy) is at best mixed. The

most compelling evidence exists for cannabis causing schizophrenia.
Overall, the prevalence of cannabis abuse disorders in patients with
schizophrenia is about 4.5 times higher than in the general population.
There is also some (albeit less robust) evidence suggestive of a dose dependency in that the more cannabis consumed, the greater the chance
of developing schizophrenia. The public health significance of these
data is considerable, and these data have fueled fierce debate (especially in England and Canada) regarding the legalization of marijuana.
However, most people who abuse cannabis do not develop schizophrenia, and even if cannabis causes schizophrenia, it is probably only in
a small minority of patients (ODely et al. 2005).
In attempting to tease out this vulnerability, Caspi et al. (2005) conducted a genetic analysis of blood samples from participants in a large
epidemiological study of schizophrenia in New Zealand. The authors
reported that adolescents who abused cannabis and who had the most
inefficient allele of the catechol-O-methyltransferase gene had the earliest onset of schizophrenia. This study is important not just because of
this finding, but also because it serves as a yardstick for the application
of neurobiological research probes to the study of dual diagnosis. Kishimoto et al. (2008) described an association between the dysbindin gene
and methamphetamine-related psychosis. This report is of interest because of the already substantial genetic evidence for the dysbindin gene
in schizophrenia, and this study suggests that this vulnerability might
extend to drug-related psychosis. Yucel et al. (2008) reported on magnetic resonance imaging in people with no prior psychiatric history who
were chronic abusers of cannabis. They found bilateral loss of hippocampal and amygdalar tissue. Interestingly, those who were heavy
abusers showed a relationship between hippocampal tissue loss and
mild paranoia. In an earlier study, Scheller-Gilkey et al. (1999) examined
differences in magnetic resonance imaging scans for 176 patients with
schizophrenia. The investigators noted that the rate of gross brain abnormalities among both alcohol and drug abusers was less than half the rate
found among patients with no history of alcohol or substance abuse, although this finding did not reach the 0.05 level of statistical significance
(Scheller-Gilkey et al. 1999). There is now a growing appreciation that
the shared vulnerability for substance abuse and schizophrenia may be
due to dopamine sensitization (Chambers et al. 2007; ODely et al. 2005);
in this model, patients may abuse drugs because of chronic dysregulation of dopamine, particularly in the amygdala.
Although research on the impact of drug and alcohol use on the
course of illness has produced variable results, the overwhelming

Substance Abuse


weight of evidence points toward substance abuse and dependence

having adverse short-term and long-term effects. Overall, persons with
schizophrenia who abuse drugs and/or alcohol have more psychotic
symptoms and psychotic relapses than do persons with schizophrenia
without substance use disorders (Degenhardt et al. 2007; Harrison et al.
2008; Negrete et al. 1986). In one of the first short-term prospective studies of recent-onset schizophrenia patients, Linszen et al. (1994) found
that significantly more and earlier psychotic relapses occurred in those
who abused cannabis. This finding was replicated in a long-term follow-up case-control study of 39 cannabis-abusing schizophrenia patients without other major drug use matched for age, gender, and year
of admission with 39 noncannabis using schizophrenia controls (Caspari 1999). After a mean 68.7 28.3 months of follow-up, the cannabis
abusers had a significantly greater hospitalization rate, higher symptom ratings, and greater unemployment.
A larger study of relapse rates was performed in Australia using a recently hospitalized sample of 99 patients, ages 1865, with schizophrenia or related disorders who were followed prospectively for 4 years
(Hunt et al. 2002). Of the 99 who entered into the study, 66 were still being followed after 4 years. Demographically, compared with nonusers,
the substance users were more likely to be male, to be younger, and to
have a forensic history, although no differences were found in age of
schizophrenia onset or number of prior hospitalizations. The investigators found that the median time until rehospitalization was 10 months
for medication-compliant patients who used substances and 37 months
for medication-compliant nonusers. Among medication-noncompliant
patients, the survival times to rehospitalization were 5 months for
substance users and 10 months for nonusers. Although the medicationnoncompliant substance users composed 28.3% of the total study sample, they accounted for 57% of all psychiatric admissions recorded by
study participants, with an average of 1.5 per year. Overall, the substance users had an average of 3.6 admissions over the 4 years compared with 1.1 for nonusers (P< 0.05) and were significantly more likely
to be medication noncompliant (users 67% vs. nonusers 34%, P<0.05).
In a first-episode study in London, in which patients with dual diagnosis were followed for 1 year, Harrison et al. (2008) found that overall the
extent of substance abuse had fallen by half. Those who continued to
abuse drugs or alcohol were more symptomatic.
The abuse of cocaine and stimulants, due to their direct effects on
dopamine, has obviously been linked to exacerbation of psychotic
symptoms (Dixon et al. 1990), yet increased severity of positive symptoms is seen in users of all substances. Caspari (1999) found higher Brief


Medical Illness and Schizophrenia

Psychiatric Rating Scale ratings in a long-term case-control study of

cannabis abusers, some of whom subsequently also abused alcohol but
not cocaine or stimulants. Similarly, Buhler et al. (2002) noted a greater
extent of positive symptoms during 5 years of follow-up in a group of
29 first-episode schizophrenia patients who were users of various substances compared with matched nonusing control subjects. In particular, hallucinations were present for 1.8 months per year for substance
users compared with 0.6 months per year for nonusers (P<0.05). The
studys findings also indicated a trend toward fewer negative symptoms among users, which reached statistical significance at the 5-year
endpoint (P<0.03).

Substance use disorders in patients with schizophrenia have been associated with violent behavior toward others, suicide (Barry et al. 1996;
Dassori et al. 1990; Drake et al. 1989; Fulwiler et al. 1997), and increased
risk for contact with the legal system (Buckley et al. 2004; Hunt et al.
2002). The association of violence with substance use is not unique to
schizophrenia or severe mental illness, and has been observed among
patients with other mental disorders (Cuffel et al. 1994; Swartz et al.
1998); moreover, this association between substance use and increased
risk for violent behavior among persons with severe mental illness has
also been described in patient populations outside the United States.
Rasanen et al. (1998) completed a prospective study in Finland on an
11,017-person unselected birth cohort followed to age 26, which
showed that men with schizophrenia who abused alcohol were 25.2
times more likely to commit violent crimes than men without mental illness, whereas nonalcoholic men with mental illness were only 3.6 times
more likely to commit violent crimes than males without any psychiatric diagnosis. Buckley et al. (2004) found that approximately 50% of patients with schizophrenia who committed violent acts were abusing
alcohol or drugs at the time of the incident.

Housing Instability and Homelessness

Persons with comorbid substance use disorders and schizophrenia are
at increased risk for homelessness (Drake et al. 1991). In Caton et al.s
(1994) case-control study comparing 100 homeless men with severe
mental illness with 100 men with schizophrenia who were not homeless, homeless subjects had significantly higher rates of drug abuse.
Studies of innovative service models for persons who are homeless and
mentally ill have found that persons with substance use comorbidity do

Substance Abuse


not benefit as much from these programs as do those who do not use
substances, in part due to the fact that substance-using patients with severe mental illness lead a more transient lifestyle. In one study of assertive community treatment for individuals who are homeless and
mentally ill, homeless persons with substance use disorders had more
moves during the treatment year than other patients with severe mental
illness (Holohan et al. 1997). Providing residential support is a key aspect of management for this problem (Drake 2007).

Medical Consequences of the Most

Commonly Abused Substances in
In the general population, alcohol is the most widely used substance of
abuse. Sequelae of ethanol use disorders represent the third leading
cause of death in the United States, with an estimated 111,000 deaths
per year directly attributable to alcohol ingestion. The most common
causes of death in alcohol-related disorders are suicide, cancer, heart
disease, and hepatic disease (Schuckit 1999). Chronic liver disease and
cirrhosis together are rated the twelfth leading cause of death in the
United States (Centers for Disease Control and Prevention 2008).
The acute action of ethanol in the central nervous system (CNS) derives from two main mechanisms: ethanol 1) facilitates the activation of
gamma aminobutyric acid type A (GABAA) receptors, the main inhibitory neurotransmitter system in the CNS, and 2) inhibits the N-methylD-aspartate (NMDA) subtype of glutamate receptors, the main excitatory neurotransmitter receptor system in the CNS. The net effects of
GABA are thereby potentiated, leading to sedation, and inhibition of
NMDA receptors via allosteric modulation appears to be responsible
for the intoxicating effects (Schuckit 1999). Sudden cessation of ethanol
use by a chronic, heavy drinker often results in an uncomplicated withdrawal syndrome characterized by mild confusion associated with diaphoresis, tremulousness, and increased heart rate, blood pressure, and
temperature, all of which can be blocked by administration of GABAacting drugs such as benzodiazepines. In about 5%10% of patients
with alcohol dependence, this syndrome may progress to delirium tremens (DTs), characterized by significant autonomic instability, marked
confusion, disorientation, agitation, tremulousness, and hallucinations
(auditory, visual, and tactile). The mortality rate for untreated DTs is


Medical Illness and Schizophrenia

approximately 5%. Alcohol withdrawal seizures (rum fits) may also

develop independently of DTs, typically within the first 2448 hours after cessation of alcohol use (Schuckit 1998).
Alcohol consumption plays a role in acute or chronic medical illness
through a variety of mechanisms, some or all of which may be operative
in a specific individual (Schuckit 1998). Repeated exposure may lead to
alcoholic hepatitis and eventually cirrhosis, with secondary cognitive
impairment via hepatic encephalopathy in advanced cirrhosis due to
the accumulation of nitrogenous compounds that are inadequately metabolized by the compromised liver. The cirrhotic changes of the liver
are also manifested in impaired synthetic function (e.g., decreased production of clotting factors) and reduced metabolism of exogenous toxins such as medications. In the CNS, prolonged alcohol abuse leads to
cerebellar degeneration, presenting as unsteady gait and mild nystagmus, in about 1% of patients with chronic alcoholism. These symptoms
are irreversible and often accompanied by global cognitive decline from
direct or indirect effects of chronic alcohol consumption (e.g., head injury, nutritional deficiency, direct neurotoxicity). Finally, peripheral
neuropathy is seen in approximately 5%15% of alcoholics due to nutritional deficiency and direct toxic effects of alcohol on neuronal axons
(Schuckit 1998).
Alcohol abuse increases blood pressure and serum triglycerides, and
may lead to cardiomyopathy and arrhythmias via toxic effects on cardiac muscle (Schuckit 1998). During the acute intoxication period, tests
for blood alcohol level, serum electrolytes, glucose, aspartate transaminase, alanine transaminase, gamma-glutamyltransferase, hematocrit,
and amylase can be useful. For the chronic alcoholic, additional laboratory values that should be monitored include albumin, red blood cell
indices, white blood cell count, platelet count, prothrombin time, hepatitis B and C screening, vitamin B12, and folate.

Cocaine and Amphetamines

Cocaine is derived from the coca plant, native to South America, and is
typically insufflated nasally (snorted), smoked (in the form of freebase
or crack), or injected intravenously. Amphetamines, which are analogs
of naturally occurring ephedrine, have been abused in various forms
since their original synthesis in 1887. Currently, d-methamphetamine
(crystal meth or meth) is the most popular form of amphetamine
abused, and, in some areas of the country, such as the West Coast, abuse
of methamphetamine is more prevalent than cocaine. Cocaine acts as a
competitive blocker of dopamine reuptake in the synaptic cleft, which

Substance Abuse


increases the concentration in the cleft, with resultant activation of

dopamine type 1 and 2 receptors. In addition, cocaine increases norepinephrine and serotonin neurotransmission via reuptake inhibition,
but these monoamines do not play the dominant role in its CNS effects
(Jaffe 1999b). By contrast, amphetamine increases the availability of all
synaptic monoamines by stimulating the release of catecholamines,
particularly dopamine, from the presynaptic terminals. This effect is especially potent for dopaminergic neurons projecting from the ventral
tegmental area to the cerebral cortex and limbic areas, known as the reward pathway.
Patients with schizophrenia who abuse stimulants experience significant increases in the extent of positive symptoms of their psychosis. In
comparison to patients with schizophrenia, cocaine-abusing patients
without underlying psychotic disorders tend to seek treatment more often for depression and anxiety than for psychosis (Serper et al. 1999).
Withdrawal following heavy stimulant use may be associated with significant lethargy and depression and an increased risk for suicide
(Gawin and Kleber 1986).
Cocaine and amphetamines have similar health effects based on their
similar pharmacological activities that result in increased synaptic
dopamine and other monoamines. The resultant sympathomimetic and
vasoconstrictive properties of these agents are responsible for many of
the acute (e.g., myocardial infarction, cardiac arrhythmias, cerebrovascular accidents, seizures) and chronic (e.g., hypertension) effects of ingestion (Jaffe 1999a). The nasal route of administration can lead to
general sinus congestion and septal perforation due to chronic ischemia. Not only does the intravenous route of administration increase
the transmission risk of HIV and hepatitis B and C viruses, but intravenous and subcutaneous (skin popping) use may result in significant
cellulitis, bone and joint infections, endocarditis, and renal failure (usually as a result of adulterants to the drug). Smoking cocaine can exacerbate preexisting asthma or chronic obstructive pulmonary disease, and
can produce specific, fibrotic lung changes known as crack lung
(Tashkin 2001). Chronic use of stimulants often results in significant
weight loss due to the anorexic effects of these agents.

Cannabis plants and derived products (e.g., hashish) contain many substances that are believed to have psychoactive properties, although the
most important of these is delta-9-tetrahydrocannabinol (THC). The endogenous cannabis receptor is a member of the G proteinlinked family


Medical Illness and Schizophrenia

of anandamide receptors, found in the highest concentration in the

basal ganglia, hippocampus, and cerebellum. In animal studies, activation of the receptor has an effect on monoamine oxidase and GABA
(Chaperon and Thiebot 1999).
The peak intoxication from smoking cannabis occurs after 1030
minutes. THC and its metabolites accumulate in fat cells and have a
half-life of approximately 50 hours (Franklin and Frances 1999). In an
intoxicated state, behavioral changes include a heightened sensitivity to
external stimuli, derealization, impaired motor skills, increased reaction time, and euphoria. Panic attacks can occur in inexperienced users.
The drug is usually smoked, but users may ingest orally (e.g., hashish
brownies). Psychosis can also occur during the intoxicated period, evidenced by transient paranoid ideation and, rarely, frank hallucinations.
Subjective reported effects of acute cannabis intoxication in individuals
with schizophrenia include a decrease in anxiety and depression and an
increase in suspiciousness (Dixon et al. 1990). In addition, chronic
heavy cannabis use can result in an amotivational syndrome that is described as passivity, decreased drive, diminished goal-directed activity,
decreased memory, fatigue, problem-solving deficits, and apathy that
can last for weeks following abstinence (Degenhardt et al. 2007). This
has been described by various authors in uncontrolled studies, although controversy still exists regarding the exact nature of the syndrome (Franklin and Frances 1999).
The deleterious health effects of cannabis are related to route of administration and direct effects on body systems. Cannabis cigarettes
(joints or blunts) have more tar and respiratory irritants than tobacco and are more carcinogenic to laboratory animals. Long-term use
has been found to lead to large airway obstruction. Smoking cannabis
also has direct inhibitory effects on pulmonary antibacterial mechanisms such as destruction of alveolar macrophages, neutrophils, and
lymphocytes. In addition, cannabis can cause variations in heart rate
and blood pressure, which can lead to an increase in oxygen consumption and increased risk of myocardial infarction in individuals with coronary artery disease (Woody and MacFadden 1996).

Screening for Substance Use

Clearly, case finding remains a priority if patients are to be offered any
form of treatment aimed at reduction in substance use. Simply put, clinicians must take the initiative to identify patients who have substance
use comorbidity. The first step requires screening and assessment. Most

Substance Abuse


patients will not spontaneously volunteer that they are using substances. Thus, clinicians must actively seek out this information, with
best results achieved using nonjudgmental forms of inquiry such as
How often have you used.. . instead of Do you use . .. Given the
prevalence of substance use, a high index of suspicion is essential. Clinicians should routinely ask all patients about use of alcohol or other
drugs and should continue to inquire on a periodic basis, especially
with newer patients who may be reluctant to discuss substance abuse
with a new clinician. Because patients sometimes deny their drug use,
a multimodal strategy is optimal, including urine toxicology screens,
interviews with collateral sources, records from recent hospitalizations,
and consultation with other care providers or family if permitted by the
The extent to which each approach adds to the confirmation of substance abuse has been studied. In a French comparative study of selfreport versus confirmatory, biological measures of substance abuse
among over 400 inpatients, the authors found that patients underreported their abuse of illicit drugs in 52% of cases (De Beaurepaire et al.
2007). Patients underreported abuse of alcohol in 56% of cases. In a
study in a U.S. community mental health center using the timeline follow-back method (an intensive, research-based retrospective reconstruction of abuse patterns), poor correlation was found between
subject self-report and collateral information (Stasiewicz et al. 2008).
Once a patient acknowledges using substances, a first step in treatment is to conduct a specialized assessment. In addition to questioning
the amount and frequency of substance use, the clinician should get an
understanding of each patients personal economy of using substances.
What benefits and costs does the patient perceive to result from using
substances? What are the patients motivations and expectations? A detailed understanding of patients perspectives on these questions is critical to engaging them in treatment and helping them to negotiate the
phases of treatment to recovery.

Treatment of Patients With Dual Diagnoses

The Presidents New Freedom Commission on Mental Health (2003)
noted that the failure to integrate substance abuse services and psychiatric/mental health services leaves people with dual diagnoses in no
mans land. These patients present complex addiction, medical, and
mental health challenges that are beyond the scope of isolated, independent services, thus posing administrative, policy, and service delivery


Medical Illness and Schizophrenia

issues that systems of care have been grappling with. Additionally, advocacy and self-help groups have realized that the needs of patients
with dual diagnoses are complex and best served with specialized supports such as Double Trouble in Recovery peer support (rather than Alcoholics Anonymous or Narcotics Anonymous groups).
Treatment of substance use by people with schizophrenia requires
an integrated approach, given the extent to which these problems interact and are interconnected (Drake 2007; Hellerstein et al. 2001). Although the deleterious effects of comorbid substance use on relapse
rates are clear, the literature on interventions offers little guidance due
to the paucity of controlled studies. Bennett et al. (2001) noted that only
seven studies available in the literature through 1998 employed experimental designs, five of which examined inpatient care or intensive
outpatient case management for serious mentally ill clients (primarily
for homeless populations), and are not directly applicable to the treatment of the broader population of people with schizophrenia living in
the community (p. 164). Of the remaining two studies, both involving
outpatient treatment, one was a controlled study that found decreased
substance abuse and psychiatric severity among patients who remained in treatment over several months, and the other was a semicontrolled study that found that both behavioral skills training and
intensive case management were more effective than a 12-step program
on several outcome measures, but with minimal effects on substance
use. Interestingly, the behavioral treatment was the most effective
even though it was not designed specifically to address substance abuse
problems and sessions were only held once per week (p. 164).
More recent controlled studies of integrated dual-diagnosis treatment have demonstrated the efficacy of this approach (Drake 2007).
McHugo et al. (1999) showed that programs with high fidelity to dualdiagnosis treatment principles had markedly greater rates of sobriety
than other treatment strategies. After 36 months, patients in high-fidelity
programs had a 55% rate of stable remission, whereas only about 15%
of patients in low-fidelity programs were in stable remission. In
Manchester, United Kingdom, Barrowclough et al. (2001) conducted a
controlled 12-month study comparing outcomes in 18 patients with
schizophrenia and substance use disorders employing an integrated
approach of motivational interviewing, family interventions (including
a family support worker), and cognitive-behavioral therapy, with outcomes in 12 patients receiving care as usual. The integrated care group
had a high retention rate (94%) and significant improvements compared with the usual care group on the Global Assessment of Functioning scale, PANSS positive symptom scores, and relapse rates. The

Substance Abuse


benefits of a combined, integrated approach, particularly with respect

to retention, are also seen in the results of a randomized controlled
study by the Combined Psychiatric and Addictive Disorders Program
(COPAD) at Beth Israel Medical Center in New York. The components
of the COPAD treatment approach include supportive group substance
abuse counseling, a multifaceted educational program (mental illness,
psychiatric medications, alcohol and drug abuse, HIV), ongoing assessment of substance use via weekly urine toxicology, encouragement to
attend self-help groups, monthly psychiatric medication visits, regular
communication with other clinicians involved in the patients care, and
as-needed communication with family members (Hellerstein et al.
2001). After 8 months, 11 of 23 patients in the COPAD cohort remained
in treatment versus only 6 of 24 usual-care patients.
Bennett et al. (2001) delineated what they perceived to be the necessary qualities of a dual-diagnosis program geared toward patients with
schizophrenia in their description of the Behavioral Treatment for Substance Abuse in Schizophrenia (BTSAS) model. The special requirements of an effective substance abuse treatment program for the
schizophrenia population derive from the findings of low motivation
for decreasing substance use (41%60% depending on the substance[s]
abused) in schizophrenia patients, and the cognitive and social skills
deficits present with schizophrenia. In creating the BTSAS program, the
investigators relied on social skills training, a behavioral approach for
rehabilitation of schizophrenia patients that has been successfully employed for the past 25 years...that employs instruction, modeling, roleplaying, and social reinforcement (p. 165).
The components of BTSAS include monthly motivational interviews
to discuss treatment goals; urinalysis with rewards for abstinence; social skills training, which teaches patients how to refuse offers of drugs;
education on the effects of drug use in schizophrenia; and problemsolving and relapse prevention training to help patients cope with
urges and high-risk situations. Most skills groups meet twice per week,
a similar frequency to the group counseling in COPAD. Recognizing
that schizophrenia patients have difficulty in changing behavior and
that many are abusing multiple substances, the BTSAS model does not
mandate need to be abstinent or commitment to total abstinence to participate. Any decrement in use is seen as a positive step that will reduce patients overall level of harm (p. 165) and bring the patient
closer to an eventual goal.
Drake et al. (2001, 2007) described several other critical components
for integrated programs, including staged interventions for those at
different stages of the recovery process, assertive outreach to engage


Medical Illness and Schizophrenia

clients (especially the homeless), the need to maintain a long-term perspective on the chronic and relapsing problems of substance use, and
cultural sensitivity. They also noted the barriers that exist in implementing such integrated programs, including administrative issues
(policy, programmatic barriers), clinician barriers due to lack of dualdiagnosis training, and consumer barriers due to denial or low motivation. Although the preceding models have not focused on intensive case
management, the problems with medication noncompliance, housing,
and psychosocial issues associated with substance use often warrant a
case management approach, especially for those who are frequent utilizers of inpatient services. The federal government has also articulated
what the best practices are for treating recurring disorders (Substance
Abuse and Mental Health Services Administration 2008).
For all schizophrenia patients, including those with dual diagnoses,
antipsychotic medication is a fundamental aspect of treatment. Although medication-compliant substance users have psychotic relapses
at higher rates than nonusers, the greatest risk for relapse is with medication-noncompliant substance users (Hunt et al. 2002). In general, there
are no absolute contraindications to the prescription of antipsychotics
for patients with schizophrenia currently using substances, apart from
those who are medically compromised (e.g., hepatic disease, HIV), in
whom dosage adjustment may be necessary. However, clinicians must
use reasonable caution in the use of the more sedating agents in patients
abusing alcohol or other CNS depressants such as opiates.
That antipsychotic medication will reduce the likelihood of psychotic relapse, even in the presence of ongoing use, is substantiated by
data from an Australian 4-year prospective study (Hunt et al. 2002), yet
findings from some recent studies indicate that the use of atypical antipsychotics might be associated with reductions in substance use. The
data are most compelling for clozapine (Green et al. 2008). One 3-year
prospective study of 151 patients with dual diagnoses found that 79.0%
of patients taking clozapine (n= 36) achieved full remission from alcohol use for 6 months or longer, compared with 33.7% of clozapine nonrecipients (Drake et al. 2000).
Some relatively less convincing support has been published for the
use of other antipsychotics in patients with schizophrenia and drug
abuse diagnoses. Several studies have shown that aripiprazole might
reduce abuse in this group (Green et al. 2008; Littrell et al. 2001; Smelson
et al. 2002). Some studies also support the use of risperidone, olanzapine, and quetiapine (Green et al. 2008). Also, an ongoing study
(P. Buckley, principal investigator) is being done of risperidone microspheres in this patient group. No information is available yet on pali-

Substance Abuse


peridone in this group, although the pharmacokinetic profile of this

drug should be more favorable for patients with dual diagnoses and alcohol-related liver damage. There is no information directly on ziprasidone in this group, apart from what can be gleaned from the CATIE
study (Swartz et al. 2008). This examination of CATIE data from phase
showed that the superiority of olanzapine over ziprasidone in prolonging time to discontinuation of treatment was still present but was
attenuated in the subgroup of substance users compared with the nonabusing sample. Curiously, some case reports describe abuse of quetiapine (Paparrigopoulos et al. 2008) and olanzapine (Reeves 2007). In
any regard, the evidence (albeit not compelling) is strongest for clozapine, although of course clozapine is not indicated for substance abuse
and is only indicated for treatment-refractory or antipsychotic-intolerant patients. Although patients with dual diagnoses may be more sensitive to the extrapyramidal side effects and the side-effect profile of
first-generation antipsychotics (Green et al. 2008; Voruganti et al. 1997),
their higher rate of medical comorbidity should lead to careful consideration about the use of second-generation antipsychotic medications,
which have a greater risk of diabetes mellitus and metabolic syndrome
(Green et al. 2008; see also Chapter 3, Medical Outcomes From the
CATIE Schizophrenia Study, in this text). The study of medications in
patients with dual diagnoses is advancing more rapidly now in an era
of more effective, pragmatic treatment trials. The treatment options for
nicotine dependence in schizophrenia are reviewed by Montoya and
Vocci (2007), as well as in Chapter 9, Nicotine and Tobacco Use in Patients With Schizophrenia, in this volume.

Schizophrenia appears to carry with it the propensity for substance
abuse, an independent but parallel process that often precedes by many
years the onset of psychotic symptoms (Buckley 2006). The most commonly abused substances by patients with schizophrenia are alcohol,
cocaine, and cannabis, with the choice of substance based on affordability and availability, and not on inherent aspects of psychopathology.
Because the medical sequelae of abuse are numerous, periodic monitoring of urine toxicology and appropriate laboratory values should be
considered as part of the routine care for schizophrenia patients with
established or suspected substance use disorders. Given the likelihood
that any schizophrenia patient has a past or ongoing substance use
problem, a high index of suspicion must be maintained even with
patients who deny or minimize the extent of their substance abuse. The


Medical Illness and Schizophrenia

greatest chance for optimizing outcomes and reducing the morbidity of

substance abuse comes from identification of patients and enrollment
into an integrated dual-diagnosis treatment modality designed specifically for substance-abusing or substance-dependent persons with
schizophrenia (Drake et al. 2007). Routine screening for and increased
recognition of substance use disorders in those with schizophrenia
must be considered a standard part of psychiatric and medical care for
this population (Substance Abuse and Mental Health Services Administration 2008).

Key Clinical Points

The odds of having any substance use disorder are 4.6 times greater
for persons with schizophrenia than for the rest of the population.

Substance use is an inherent aspect of the neurobiology of schizophrenia, and increased rates of substance abuse can be seen in prodromal

Little evidence is available to suggest that substances are used specifically to ameliorate certain symptom states.

The impact of substance use on poor psychiatric outcomes is mediated

primarily by medication nonadherence.

Screening for substance use disorders includes open-ended nonjudgmental inquiry, routine biological fluid testing, and screening for associated medical disorders (e.g., hepatitis C, HIV).

Effective programs for patients with schizophrenia who use substances

must account for the lower levels of motivation and the cognitive deficits seen in this population, and must strongly emphasize role-playing,
social reinforcement, and problem solving as core aspects of relapse
prevention, typically in ongoing skill-building group sessions.

Alexander MJ: Women with co-occurring addictive and mental disorders: an
emerging profile of vulnerability. Am J Orthopsychiatry 66:6170, 1996
Archie S, Rush BR, Akhtar-Danesh N, et al: Substance use and abuse in firstepisode psychosis: prevalence before and after early intervention.
Schizophr Bull 33:13541363, 2007

Substance Abuse


Arndt S, Tyrrell G, Flaum M, et al: Comorbidity of substance abuse and schizophrenia: the role of pre-morbid adjustment. Psychol Med 22:379388, 1992
Barnett JH, Werners U, Secher SM, et al: Substance use in a population-based
clinic sample of people with first-episode psychosis. Br J Psychiatry
190:515520, 2007
Barrowclough C, Haddock G, Tarrier N, et al: Randomized controlled trial of
motivational interviewing, cognitive behavior therapy, and family intervention for patients with comorbid schizophrenia and substance use disorders. Am J Psychiatry 158:17061713, 2001
Barry KL, Fleming MF, Greenley JR, et al: Characteristics of persons with severe
mental illness and substance abuse in rural areas. Psychiatr Serv 47:8890, 1996
Bennett ME, Bellack AS, Gearon JS: Treating substance abuse in schizophrenia:
an initial report. J Subst Abuse Treat 20:163175, 2001
Buckley PF: Prevalence and consequences of the dual diagnosis of substance
abuse and severe mental illness. J Clin Psychiatry 67:59, 2006
Buckley PF, Hrouda DR, Friedman L: Insight and its relationship to violent behavior in schizophrenia. Am J Psychiatry 161:17121714, 2004
Buhler B, Hambrecht M, Loffler W, et al: Precipitation and determination of the
onset and course of schizophrenia by substance abuse: a retrospective and
prospective study of 232 population-based first illness episodes. Schizophr
Res 54:243251, 2002
Cannon TD, Cadenhead K, Cornblatt B, et al: Prediction of psychosis in youth
at high clinical risk: a multisite longitudinal study in North America. Arch
Gen Psychiatry 65:2837, 2008
Cantor-Graae E, Nordstrom LG, McNeil TF: Substance abuse in schizophrenia:
a review of the literature and a study of correlates in Sweden. Schizophr
Res 48:6982, 2001
Cantwell R, Berwin J, Glazebrook C, et al: Prevalence of substance misuse in
first-episode psychosis. Br J Psychiatry 174:150153, 1999
Caspari D: Cannabis and schizophrenia: results of a follow-up study. Eur Arch
Psychiatry Clin Neurosci 249:4549, 1999
Caspi A, Moffitt TE, Cannon M, et al: Moderation of the effect of adolescentonset cannabis use on adult psychosis by a functional polymorphism in the
catechol-O-methyltransferase gene: longitudinal evidence of a gene environment interaction. Biol Psychiatry 57:11171127, 2005
Caton CL, Shrout PE, Eagle PF, et al: Risk factors for homelessness among schizophrenic men: a case-control study. Am J Public Health 84:265270, 1994
Caton CL, Drake RE, Hasin DS, et al: Differences between early phase primary
psychotic disorders with concurrent substance use and substance-induced
psychoses. Arch Gen Psychiatry 62:137145, 2005
Centers for Disease Control and Prevention: U.S. mortality drops sharply
in 2006, latest data show. June 11, 2008. Available at:
nchs/pressroom/08newsreleases/mortality2006.htm. Accessed October
28, 2008.


Medical Illness and Schizophrenia

Chambers RA, Krystal JH, Self DW: A neurobiological basis for substance abuse
comorbidity in schizophrenia. Biol Psychiatry 50:7183, 2001
Chambers RA, Bickel WK, Potenza MN: A scale-free systems theory of motivation and addiction. Neurosci Biobehav Rev 31:10171045, 2007
Chaperon F, Thiebot MH: Behavioral effects of cannabinoid agents in animals.
Crit Rev Neurobiol 13:243281, 1999
Comtois KA, Ries R: Sex differences in dually diagnosed severely mentally ill
clients in dual diagnosis outpatient treatment. Am J Addict 4:245253, 1995
Cuffel BJ, Shumway M, Chouljioa TL, et al: A longitudinal study of substance
use and community violence in schizophrenia. J Nerv Ment Dis 182:704
708, 1994
Dassori AM, Mezzich JE, Keshavan M: Suicidal indicators in schizophrenia.
Acta Psychiatr Scand 81:409413, 1990
De Beaurepaire R, Lukasiewicz M, Beauverie P, et al: Comparison of self reports
and biological measures for alcohol, tobacco, and illicit drugs consumption
in psychiatric inpatients. Eur Psychiatry 22:540548, 2007
Degenhardt L, Tennant C, Gilmour S, et al: The temporal dynamics of relationships between cannabis, psychosis and depression among young adults
with psychotic disorders: findings from a 10-month prospective study.
Psychol Med 37:927934, 2007
De Leon J, Diaz FJ, Rogers T, et al: Initiation of daily smoking and nicotine dependence in schizophrenia and mood disorders. Schizophr Res 56:4754,
De Leon J, Gurpegui M, Diaz F, et al: Epidemiology of comorbid tobacco use in
schizophrenia: thinking about risks and protective factors. J Dual Diagn 3/
4:926, 2007
Dervaux A, Bayle FJ, Laqueille X, et al: Is substance abuse in schizophrenia related to impulsivity, sensation seeking, or anhedonia? Am J Psychiatry
158:492494, 2001
Dixon L, Haas G, Weiden PJ, et al: Acute effects of drug abuse in schizophrenic
patients: clinical observations and patients self-reports. Schizophr Bull
16:6979, 1990
Dixon L, Haas G, Weiden PJ, et al: Drug abuse in schizophrenic patients: clinical
correlates and reasons for use. Am J Psychiatry 148:224230, 1991
Drake RE: Management of substance use disorder in schizophrenia patients:
treatment guidelines. CNS Spectr 12:2732, 2007
Drake RE, Osher FC, Wallach MA: Alcohol use and abuse in schizophrenia: a
prospective community study. J Nerv Ment Dis 77:408414, 1989
Drake RE, Osher FC, Wallach MA: Homelessness and dual diagnosis. Am Psychol 46:11491158, 1991
Drake RE, Xie H, McHugo GJ, et al: The effects of clozapine on alcohol and drug
use disorders among patients with schizophrenia. Schizophr Bull 26:441
449, 2000
Drake RE, Essock SM, Shaner A, et al: Implementing dual diagnosis services for
clients with severe mental illness. Psychiatr Serv 52:469476, 2001

Substance Abuse


Drake RE, Mueser KT, Burnette MF: Management of persons with co-occurring
severe mental illness and substance use disorder: program implications.
World Psychiatry 6:131136, 2007
Duke PJ, Pantelis C, Barnes TRE: South Westminster schizophrenia survey: alcohol use and its relationship to symptoms, tardive dyskinesia and illness
onset. Br J Psychiatry 164:630636, 1994
Franklin JE, Frances RF: Alcohol and other psychoactive substance use disorders, in Textbook of Psychiatry, 3rd Edition. Edited by Hales RE, Yudofsky
SC, Talbott JA. Washington, DC, American Psychiatric Press, 1999, pp 363
Fulwiler C, Grossman H, Forbes C, et al: Early onset substance abuse and community violence by outpatients with chronic mental illness. Psychiatr Serv
48:11811185, 1997
Gawin F, Kleber H: Abstinence symptomatology and psychiatric diagnosis in
cocaine abusers. Arch Gen Psychiatry 43:107133, 1986
Grassi L, Pavanati M, Cardelli R, et al: HIV-risk behaviour and knowledge
about HIV/AIDS among patients with schizophrenia. Psychol Med
29:171179, 1999
Green AI, Noordsy DL, Burnette MF, et al: Substance abuse and schizophrenia:
pharmacotherapeutic interventions. J Subst Abuse Treat 34:6171, 2008
Hambrecht M, Hafner H: Cannabis, vulnerability, and the onset of schizophrenia: an epidemiological perspective. Aust N Z J Psychiatry 34:468475, 2000
Harrison I, Joyce EM, Mutsatsa SH, et al: Naturalistic follow-up of co-morbid
substance use in schizophrenia: the West London first-episode study. Psychol Med 38:7988, 2008
Hellerstein DJ, Rosenthal RN, Miner CR: Integrating services for schizophrenia
and substance abuse. Psychiatr Q 72:291306, 2001
Holohan N, Dixon L, Drauss N: Outcomes of housing in persons with mental
illness and substance use disorders. Poster presented at the annual meeting
of the American Psychiatric Association, San Diego, CA, May 1997
Hunt GE, Bergen J, Bashir M: Medication compliance and comorbid substance
abuse in schizophrenia: impact on community survival 4 years after a relapse. Schizophr Res 54:253264, 2002
Jaffe JH: Amphetamine (or amphetamine-like)-related disorders, in Comprehensive Textbook of Psychiatry, 7th Edition. Edited by Sadock BJ, Sadock
VA. New York, Lippincott, Williams, and Wilkins, 1999a, pp 971981
Jaffe JH: Cocaine-related disorders. In Comprehensive Textbook of Psychiatry,
7th Edition. Edited by Sadock BJ, Sadock VA. New York, Lippincott Williams & Wilkins, 1999b, pp 9991014
Kessler RC, Crum RM, Warner LA, et al: Lifetime co-occurrence of DSM-III-R
alcohol abuse and dependence with other psychiatric disorders in the National Comorbidity Study. Arch Gen Psychiatry 54:313321, 1997
Kishimoto M, Ujike H, Motohashi Y, et al: The dysbindin gene (DTNBP1) is associated with methamphetamine psychosis. Biol Psychiatry 63:191196,


Medical Illness and Schizophrenia

Linszen DH, Dingemans PM, Lenior ME: Cannabis abuse and the course of recent-onset schizophrenic disorders. Arch Gen Psychiatry 51:273279, 1994
Littrell KH, Petty RG, Hilligoss NM, et al: Olanzapine treatment for patients
with schizophrenia and substance abuse. J Subst Abuse Treat 21:217221,
Mathias S, Lubman DI, Hides L: Substance-induced psychosis: a diagnostic conundrum. J Clin Psychiatry 69:358367, 2008
McHugo GJ, Drake RE, Teague GB, et al: Fidelity to assertive community treatment and client outcomes in the New Hampshire Dual Diagnosis study.
Psychiatr Serv 50:818824, 1999
Merikangas KR, Ames M, Lui L: The impact of comorbidity of mental and physical conditions on sole disability in the U.S. adult household population.
Arch Gen Psychiatry 64:11801188, 2007
Meyer JM: Prevalence of hepatitis A, hepatitis B, and HIV among hepatitis Cseropositive state hospital patients: results from Oregon State Hospital.
J Clin Psychiatry 64:540545, 2003
Montoya ID, Vocci F: Medication development for the treatment of nicotine dependency in individuals with schizophrenia. J Dual Diagn 30:113150, 2007
Mueser KT, Yarnold PR, Levinson DF, et al: Prevalence of substance abuse in
schizophrenia: demographic and clinical correlates. Schizophr Bull 16:31
56, 1990
Mueser KT, Yarnold PR, Bellack AS: Diagnostic and demographic correlates of
substance abuse in schizophrenia and major affective disorder. Acta Psychiatr Scand 85:4855, 1992
Mueser KT, Bennett M, Kushner MG: Epidemiology of substance use disorders
among persons with chronic mental illness, in Double Jeopardy: Chronic
Mental Illness and Substance Use Disorders. Edited by Lehman AF, Dixon
LB. Chur, Switzerland, Harwood Academic, 1995, pp 926
Mueser KT, Essock SM, Drake RE, et al: Rural and urban differences in patients
with a dual diagnosis. Schizophr Res 48:93107, 2001
Negrete JC, Werner PK, Doublas DE, et al: Cannabis affects the severity of
schizophrenic symptoms: results of a clinical survey. Psychol Med 16:515
520, 1986
New Freedom Commission on Mental Health: Achieving the Promise: Transforming Mental Health Care in America. Final Report (DHHS Publ No
SMA-03-3832).. Rockville, MD, Presidents New Freedom Commission on
Mental Health, 2003
ODely OG, Guillin, Tsapakis EM, et al: Schizophrenia and substance abuse comorbidity: a role for dopamine sensitization? J Dual Diagn 1:1140, 2005
Paparrigopoulos T, Karaiskos D, Liappas J: Quetiapine: another drug with potential for misuse? J Clin Psychiatry 69:162163, 2008
Rasanen P, Tiihonen J, Isohanni M, et al: Schizophrenia, alcohol abuse, and violent behavior: a 26-year follow-up study of an unselected birth cohort.
Schizophr Bull 24:437441, 1998

Substance Abuse


Reeves RR: Abuse of olanzapine by substance abusers. J Psychoactive Drugs

39:297299, 2007
Regier DA, Farmer ME, Rae DS, et al: Comorbidity of mental disorders with alcohol and other drug abuse: results from the Epidemiologic Catchment
Area (ECA) study. JAMA 264:25112518, 1990
Rosenberg SD, Goodman LA, Osher FC, et al: Prevalence of HIV, hepatitis B,
and hepatitis C in people with severe mental illness. Am J Public Health
91:3137, 2001
Salyers MP, Mueser KT: Social functioning, psychopathology, and medication
side effects in relation to substance use and abuse in schizophrenia.
Schizophr Res 48:109123, 2001
Scheller-Gilkey G, Lewine RR, Caudle J, et al: Schizophrenia, substance use, and
brain morphology. Schizophr Res 35:113120, 1999
Schuckit MA: Alcohol and alcoholism, in Harrisons Principles of Internal Medicine 14th edition. Edited by Fauci AS, Braunwald E, Isselbacher KJ, et al.
New York, McGraw-Hill, 1998, pp 25032508
Schuckit MA: Alcohol-related disorders, in Comprehensive Textbook of Psychiatry, 7th Edition. Edited by Sadock BJ, Sadock VA. New York, Lippincott
Williams & Wilkins, 1999, pp 953970
Serper MR, Chou JCY, Allen MH, et al: Symptomatic overlap of cocaine intoxication and acute schizophrenia at emergency presentation. Schizophr Bull
25:387394, 1999
Smelson DA, Losonczy MF, Davis CW, et al: Risperidone decreases craving and
relapses in individuals with schizophrenia and cocaine dependence. Can J
Psychiatry 47:671675, 2002
Stasiewicz PR, Vincent PC, Bradizza CM, et al: Factors affecting agreement between severely mentally ill alcohol abusers and collaterals reports of alcohol and other substance abuse. Psychol Addict Behav 22:7887, 2008
Substance Abuse and Mental Health Services Administration: National outcomes measures. Available at http://www.nationaloutcomemeasures Accessed March 17, 2008.
Swann AC, Dougherty DM, Pazzaglia PJ, et al. Impulsivity: a link between bipolar disorder and substance abuse. Bipolar Disord 6:204212, 2004
Swartz MS, Swanson JW, Hiday VA, et al: Violence and severe mental illness:
the effects of substance abuse and nonadherence to medication. Am J Psychiatry 155:226231, 1998
Swartz MS, Wagner HR, Swanson JW, et al: The effectiveness of antipsychotic
medications in patients who use or avoid illicit substances: results from the
CATIE study. Schizophr Res 100:3952, 2008
Tashkin D: Airway effects of marijuana, cocaine, and other inhaled illicit agents.
Curr Opin Pulm Med 7:143161, 2001
Voruganti LNP, Heslegrave RJ, Awad AG: Neuroleptic dysphoria may be the
missing link between schizophrenia and substance abuse. J Nerv Ment Dis
185:463465, 1997


Medical Illness and Schizophrenia

Wirshing D, Tabibian JH, Pierre JM, et al: Hepatitis B and C among veterans on
a psychiatric ward. Paper presented at International Congress on Schizophrenia Research, Savannah, GA, April 2005
Woody G, MacFadden W: Cannabis-related disorders, in Substance Abuse: A
Comprehensive Textbook, 3rd Edition. Edited by Lowinson J, Ruiz P, Millman J, et al. Baltimore, MD, Williams & Wilkins, 1996, pp 810817
Yucel M, Solowij N, Respondek C, et al: Regional brain abnormalities with longterm heavy cannabis use. Arch Gen Psychiatry 65:694701, 2008
Zisook S, Heaton R, Moranville J, et al: Past substance abuse and clinical course
of schizophrenia. Am J Psychiatry 149:552553, 1992

Sexual Dysfunction
and Schizophrenia
Heidi J. Wehring, Pharm.D., B.C.P.P.
Deanna L. Kelly, Pharm.D., B.C.P.P.

The sexuality of people with schizophrenia has received relatively

little attention in the past. Evidence suggests that patients with schizophrenia identify sexual function and medication-related sexual side effects as important issues, yet much remains unknown regarding sexual
dysfunction in this population. Multiple factors have contributed to this
lack of knowledge, including lack of basic information on what constitutes normal sexual functioning of patients with schizophrenia. Earlier studies and theories suggested that sexuality is best not discussed
with patients with schizophrenia, that identifying and treating sexual
issues and side effects might exacerbate illness or lengthen recovery,
and that sexual activity could even contribute to the development of
schizophrenia (Pinderhughes et al. 1972). This perception has changed
dramatically in recent years, and modern data indicate that the sexual
activity, desire, fantasies, and expectations of patients with schizophrenia, although more autoerotic in nature, are not likely to be much different from those of the general population (Kelly et al. 2006).
Although addressing clinical symptoms in patients with schizophrenia
generally takes precedence over discussions of sexual function, clinicians
often underestimate the prevalence of sexual dysfunction and the significance that sexual side effects may have on psychotropic adherence and
outcomes. In one prevalence study, sexual dysfunction occurred in more


Medical Illness and Schizophrenia

than 50% of subjects with schizophrenia, yet the treating nurses and physicians reported significantly lower rates of sexual dysfunction in these
same individuals (Dossenbach et al. 2005). In responding to a survey, patients with schizophrenia rated the area of sexual expression as the third
most significant unmet treatment need (Grinshpoon and Ponizovsky
2008). One contributing element to the underrecognition of sexual dysfunction and its importance for patients with schizophrenia could be that
this patient population may be unwilling or embarrassed to broach the
topic of sexual concerns and not comfortable in discussing their sexuality.
Moreover, because clinicians do not routinely focus on sexual symptoms
or issues, patients with schizophrenia may think that clinicians do not take
these concerns seriously or believe that their concerns are important.
Given the high prevalence of sexual dysfunction in patients with
schizophrenia and the fact that its occurrence may be associated with
lower quality of life and poorer psychiatric outcomes, this chapter is intended to serve as a guide for clinicians on the extent of and contributors to these problems and the means to address unmet needs in this
area. Improving the focus on this domain of treatment could lead to
more integrated and comprehensive psychiatric and medical care and
a better step toward the path to recovery.

Sexual dysfunction is common in the general population, with estimates
of 43% of women and 31% of men in the United States reporting some
type of sexual dysfunction (Laumann et al. 1999). The prevalence of sexual
dysfunction in patients with schizophrenia has been more difficult to ascertain but is considered to be higher than in the general population, with
reported rates averaging 50%75% (Kelly and Conley 2004).
Several reasons may explain the wide range of reported prevalence
rates in patients with schizophrenia, including the nature of the patient
population studied (illness severity and acuity, location of study, men vs.
women, partner vs. no partner), medication status of patients (treated vs.
untreated, type of antipsychotic medication and concomitant medications used), method of data collection (spontaneous report vs. standardized assessment tool), source of data (clinician vs. patient), time frame of
study, and definition of sexual dysfunction measured. Some studies
have suggested that higher psychopathology rating scores have been associated with greater impairment in sexual function (Fan et al. 2007;
Kockott and Pfeiffer 1996); however, these findings have not been widely
replicated, and more work is needed to ascertain this relationship. Antipsychotic treatment is known to cause sexual impairment, yet some

Sexual Dysfunction


authors (Aizenberg et al. 1995; Bitter et al. 2005) have reported sexual
dysfunction in treatment-naive patients with schizophrenia, suggesting
that a treatment-independent factor contributes to this clinical problem.
Evidence indicates that males and females experience different
forms of sexual dysfunction and at different rates, but the bulk of the
schizophrenia literature has focused on males, possibly due to reporting bias. Although a review by Howes et al. (2007) found that the odds
ratios of patients with schizophrenia having sexual dysfunction, compared with the general population, were 15.2 and 3.7 for women and
men, respectively, one recent study of 827 patients on antipsychotic
medications found that 58% of women but only 43% of men reported
sexual dysfunction (ok et al. 2007). Despite higher sexual dysfunction rates, females are less likely to spontaneously complain about sexual dysfunction in treatment settings, leading to an underestimation of
prevalence. A survey study in a diagnostically heterogeneous group of
psychiatric outpatients found that among those with sexual side effects,
80% of women versus 26.7% of men never or infrequently discussed
their sexual lives with clinicians (Rosenberg et al. 2003).
Another methodological issue in interpreting sexual dysfunction prevalence rates relates to the fact that many studies include only those with
sexual partners, thereby excluding a significant portion of the schizophrenia population, because patients with schizophrenia are less likely to be
married or in a relationship than those in the general population (Aizenberg et al. 2001). Medications, other mental and physical comorbidities,
and substance and tobacco use may also affect sexual function. The prevalence rates of environmental exposures to substances and tobacco vary
considerably between schizophrenia patients and the general population.

Types of Sexual Dysfunction

Sexual dysfunction in patients with schizophrenia may occur in any of
the areas of the sexual response cycle. Generally speaking, this cycle includes the following domains:
Interest (libido)
Arousal (vaginal lubrication in women, erectile function in men)
The breadth of what is covered by the term sexual dysfunction and what
has been measured in studies varies greatly, sometimes causing difficulty in providing consensus regarding types of sexual dysfunction that
may occur in the schizophrenia population.


Medical Illness and Schizophrenia

Men with schizophrenia who have sexual dysfunction typically complain of decreased libido and erectile, ejaculatory, and orgasmic dysfunction. The reported erectile and ejaculatory disturbances are
generally related to difficulty achieving and maintaining erections, delayed or inhibited ejaculation, and diminished orgasm quality. The
prevalence of these sexual impairments ranges widely; however, most
reports indicate that problems with desire and arousal as well as problems with ejaculatory disturbances are present in up to 60% of men with
schizophrenia (Bobes et al. 2003; Olfson et al. 2005). Spontaneous and
premature ejaculation also occurs more commonly in men with schizophrenia than in the general population (Aizenberg et al. 1995).
Galactorrhea and gynecomastia are uncommon in men but may occur in males treated with antipsychotics. Gynecomastia has been reported in 1%2% of men taking antipsychotic medications, with some
estimates up to 11%, whereas galactorrhea is thought to occur more
rarely (Byerly et al. 2007). Priapism is an infrequently reported but serious side effect associated with antipsychotic medications. Priapism, a
prolonged, painful erection not usually associated with sexual stimuli,
requires emergency urological evaluation to minimize potentially serious long-term consequences such as erectile dysfunction (Compton and
Miller 2001).

Women may experience sexual dysfunction across any area of sexual
response, including libido, arousal, and orgasm. Women can also experience dyspareunia associated with vaginal atrophy and dryness, and
are at risk of suffering menstrual disturbances and galactorrhea, particularly if they are taking prolactin-elevating antipsychotics. Many of the
reports in the literature cite high rates (>60%) of loss of libido and
arousal, including problems with vaginal lubrication and pleasure (Fan
et al. 2007; Teusch et al. 1995; ok et al. 2007). Additionally, women
with schizophrenia report low rates of orgasm, with papers suggesting
that over half of women with schizophrenia have never experienced an
orgasm (Friedman and Harrison 1984).
The reported prevalence of galactorrhea in women treated with antipsychotics ranges from 10% to 90% but differs according to the type of
antipsychotic (Byerly et al. 2007). Although not directly related to sexual function, menstrual abnormalities are frequently co-occurring antipsychotic adverse effects that often go unrecognized in clinical trials

Sexual Dysfunction


because this information is not volunteered and clinical investigators

may not specifically evaluate this side effect. Numerous cross-sectional
studies of menstrual irregularities in antipsychotic-treated women with
schizophrenia indicate that the propensity to elevate prolactin remains
a common mechanism for this complaint. In these studies, the occurrence of menstrual abnormalities ranges from 25% to 100%, with the average prevalence around 50% (Kelly and Conley 2004).

Impact of Sexual Dysfunction on

Patient Outcomes
Long-term quality-of-life outcomes in schizophrenia are often less than
ideal, with reduced life expectancy, high suicide rates, and low rates of
marriage and employment. The mental health field is increasingly
aware of the need to strive for recovery and to incorporate both medical
and psychological care along with psychiatric symptom control into
treatment planning. Sexuality, which is a vital component of the self
and provides a sense of personal satisfaction, remains a significant unmet need hindering full recovery for patients with schizophrenia. Importantly, sexual dysfunction has negative effects on critical aspects of
care necessary to achieving sustained psychiatric outcomes, namely
medication adherence and self-rated quality of life and well-being.

Adherence to antipsychotic treatment is the cornerstone of schizophrenia management and relapse prevention, yet nonadherence rates typically range from 40% to 50% (Lacro et al. 2002). A myriad of factors
contribute to nonadherence, including the following:
Lack of illness insight
Poor relationship to health care provider
Belief that treatment may not improve symptoms or prevent recurrence of symptoms
Negative attitude toward medications
Medication side effects
Sexual dysfunction is distressing to patients, particularly when perceived as a side effect of psychotropic medications, and the limited research on this topic indicates that impairment in sexual function can
negatively affect adherence to antipsychotic medications. In a survey of


Medical Illness and Schizophrenia

outpatients with severe mental illness, Rosenberg et al. (2003) found

that 43% of those who experienced sexual dysfunction said they had
considered stopping medications because of sexual side effects, and
27.5% claimed that they had actually stopped medications one or more
times due to sexual side effects. As discussed in the later section on etiology, sexual dysfunction has many possibly etiologies, yet several
studies report that patients often attribute their sexual dysfunction to
antipsychotic treatment (Lambert et al. 2004; Olfson et al. 2005).

Quality of Life
Quality of life is a broad term that includes subjective well-being and both
mental and physical function, and has become an outcome of great interest
in the treatment of schizophrenia. Those with schizophrenia experiencing
sexual dysfunction may report any of the following (Olfson et al. 2005):

Poorer quality of life

Lower levels of enjoyment in life
Low satisfaction with the quality of romantic relationships
Fewer romantic partners

Although limited formal data have been reported in this area, the few
existing studies document that sexual dysfunction is associated with
lower quality of life among patients with schizophrenia. These patients
describe sexual dysfunction as a troublesome side effect associated with
a worse subjective quality of life, with some men complaining that medication-related impotence is more bothersome than their psychotic
symptoms (Finn et al. 1990). Not surprisingly, one study found that
men with schizophrenia who reported sexual dysfunction also reported
a lower global quality of life compared with male patients without sexual dysfunction (Olfson et al. 2005). Moreover, in a large study assessing satisfaction and quality of life in patients with schizophrenia, lack
of sexual activity was one of the top four reasons for low ratings of life
satisfaction (S. Chan and Yu 2004).

Multiple factors contribute to sexual dysfunction in patients with
schizophrenia, and the etiologies for any one patient can be complex.
The symptoms of schizophrenia, including psychosocial difficulties, a
lack of personal relationships, negative symptoms, anhedonia, and
impulsivity, are potential contributing factors. The following physical

Sexual Dysfunction


effects of antipsychotic treatment may also contribute to sexual dysfunction (Kelly and Conley 2004):

Obesity and weight gain

Tardive dyskinesia
Extrapyramidal side effects

Over 100 medications or classes of medications have been associated

with sexual dysfunction, although the quality of evidence is not uniform (Thomas 2003) (see Table 121). A lengthier discussion of medication effects follows in the next section, but it is worthwhile noting here
that other physical and mental illnesses can affect sexual functioning,
including the following:

Cardiovascular disease
Peripheral neuropathy

Human sexual physiology is complex, involving endocrine, central

nervous, peripheral nervous, and vascular systems. The endocrine system is thought to be responsible for libido, so reductions in sex hormones such as testosterone may impair erectile function and libido in
males. In females, low testosterone levels have been theorized to be associated with low sexual desire and sexual dysfunction; however, this
has not been supported by direct evidence, and some conflicting reports failed to find low androgen levels to be predictive of low sexual
function in women (Davis et al. 2005). The limbic system mediates
arousal mechanisms, and neurohormones in the central nervous system
(CNS) also mediate sexual function, with the monoamine neurotransmitter serotonin acting primarily in an inhibitory role and dopamine in
a stimulatory role. Because dopamine is related to motivation and reward in the CNS, dopaminergic receptor antagonism may contribute to
decreased libido. Additionally, dopamine antagonism may lead to prolactin elevation and associated consequences of sexual dysfunction, as
described later in this section.
The peripheral nervous system participates in sexual function
through sensory input mediated by adrenergic (sympathetic) and cholinergic (parasympathetic) stimulation (Thomas 2003), with ejaculation


Medical Illness and Schizophrenia

TABLE 121. Medications associated with sexual dysfunction (not a

comprehensive list)
Medications (U.S. brand name),
by class

Reported effect on
sexual function

Medications inducing
cytochrome P450 enzymes
Carbamazepine (Tegretol)
Phenytoin (Dilantin)

Increased metabolism of
androgens, possibly affecting
sexual function

Monoamine oxidase inhibitors
Isocarboxazid (Marplan)
Phenelzine (Nardil)
Tranylcypromine (Parnate)
Selective serotonin reuptake
Citalopram (Celexa)
Escitalopram (Lexapro)
Fluoxetine (Prozac)
Fluvoxamine (Luvox)
Paroxetine (Paxil)
Sertraline (Zoloft)
Other antidepressants
Mirtazapine (Remeron)

-Adrenergic blockers
Prazosin (Minipress)
Terazosin (Hytrin)
Mixed --adrenergic blocker
Labetalol (Normodyne,
-Adrenergic antagonists
Atenolol (Tenormin)
Metoprolol (Lopressor)
Propranolol (Inderal)

Differences in risk of sexual side

effects may exist between
medications in class (inhibited
ejaculation, decreased libido)
Differences in risk of sexual side
effects may exist between
medications within class
(decreased libido, delayed
orgasm/ejaculation, anorgasmia,
decreased vaginal lubrication)

Decreased libido, delayed

orgasm/ejaculation, anorgasmia,
erectile dysfunction, decreased
vaginal lubrication
Ejaculatory disturbances
Inhibited ejaculation

Erectile dysfunction, decreased


Sexual Dysfunction


TABLE 121. Medications associated with sexual dysfunction (not a

comprehensive list) (continued)
Medications (U.S. brand name),
by class

Reported effect on
sexual function

Centrally acting agents

Clonidine (Catapres)
Methyldopa (Aldomet)
Reserpine (Reserpine)
Spironolactone (Aldactone)
Thiazides (various)

Erectile dysfunction, decreased

libido (methyldopa and

Antihypertensives, other
Guanethidine (sympathetic
nerve blocker) (Ismelin)
Cardiac medications
Digoxin (Lanoxin)
Gastrointestinal medication
Cimetidine (Tagamet)
Hormonal contraceptives

Spironolactone: decreased libido,

breast swelling, erectile
Thiazides: erectile dysfunction
Erectile dysfunction, ejaculatory
disturbances (retrograde)

Erectile dysfunction,
Decreased libido, impotence,
Decreased libido

Source. Adapted from American Association of Clinical Endocrinologists

Male Sexual Dysfunction Task Force: American Association of Clinical Endocrinologists Medical Guidelines for Clinical Practice for the Evaluation and
Treatment of Male Sexual Dysfunction: A Couples Problem2003 Update.
Endocrine Practice 9:7795, 2003.

in particular mediated by -adrenergic and -adrenergic systems, as

well as motor nerves. Potent -adrenergic blockade has been associated
with ejaculatory disturbances and priapism (Compton and Miller 2001;
Wang et al. 2006). Cholinergic receptor antagonism may affect erectile
function, and calcium channel blockade may also contribute to sexual
dysfunction (Cutler 2003; Gitlin 1994; Haddad and Wieck 2004). Serotonergic neurotransmission in the peripheral system has vasoconstriction/vasodilatory effects and plays a role in the normal sexual response
cycle, and serotonin agonists (typically in the form of selective seroto-


Medical Illness and Schizophrenia

nin reuptake inhibitor [SSRI] antidepressants) are notorious inducers of

sexual dysfunction. Finally, sedation and diminished arousal due to
histamine receptor antagonism may also contribute to sexual dysfunction.
Cigarette smoking status should be obtained during the inquiry into
sexual complaints, because up to 80% of persons with schizophrenia
smoke (Kelly et al. 2007), and smoking is associated with erectile dysfunction. Smoking is also a risk factor for coronary heart disease and peripheral vascular disease, with the result that smoking-induced
inflammation and dyslipidemia become important contributors to vascular erectile dysfunction and must be considered in all males with
erectile complaints, regardless of the type of antipsychotic treatment.
Given the complex neurobiology of sexual function, antipsychotic
medications can be associated with sexual dysfunction through any of
the mechanisms described, including effects on the following systems:


As noted in the later section on medication effects, certain classes of

antipsychotics are associated with specific risks for sexual dysfunction,
but prolactin elevation is a common problem with many antipsychotics.
Prolactin elevation from antipsychotic treatment primarily occurs by
dopamine blockade at the pituitary level and is the cause of the majority
of sexual dysfunction specifically attributable to antipsychotic treatment. A recent article indicated that prolactin elevation explained 40%
of all sexual dysfunction present in patients with schizophrenia (Knegtering et al. 2008). Prolactin inhibits gonadotropin-releasing hormone
and, by function of feedback loop between pituitary and hypothalamus,
subsequently inhibits luteinizing hormone and follicle-stimulating hormone. Hyperprolactinemia may be associated with a decrease in testosterone levels, which is thought to lead to reduction in sexual activity
(Barnes and Harvey 1993). Direct correlations between sexual dysfunction and prolactin levels have not been routinely found in all studies,
but those individuals with higher prolactin levels generally have higher
rates of sexual dysfunction (Byerly et al. 2007).
The effects of hyperprolactinemia include both sexual and hormonal
consequences. Hyperprolactinemia is associated with galactorrhea, gynecomastia, oligomenorrhea, amenorrhea, and libido changes in

Sexual Dysfunction


women; with galactorrhea, gynecomastia, changes in libido, and erectile

and ejaculatory dysfunction in men; and possibly with effects on bone
mineral density and risk of cancer (Haddad and Wieck 2004). Although
not a sexual side effect, galactorrhea can be very distressing and has
been reported in 10%90% of women treated with prolactin-elevating
antipsychotics (Byerly et al. 2007). As noted earlier, gynecomastia and
galactorrhea occur only rarely in men exposed to antipsychotics.

Prolactin and Osteoporosis

Accumulating evidence suggests that hyperprolactinemia may increase
the loss of bone mineral density in patients with schizophrenia during
long-term antipsychotic drug treatment, either through secondary
hypogonadism or hyperprolactinemia itself. Previous studies have
demonstrated that between 32% and 65% of patients treated with antipsychotic drugs suffer from bone mineral loss, leading to osteoporosis
(Meaney et al. 2004). One study found that among women with schizophrenia treated with haloperidol, 18 of 21 (86%) had significant bone
mineral density loss compared with 9 of 23 (39%) normal control subjects (Jung et al. 2006). Bone fractures in patients with schizophrenia
taking antipsychotics also occur more frequently than in the nonpsychiatric population. Other evidence suggests that amenorrhea and prolactin-elevating antipsychotic drugs are associated with a higher risk
for osteoporosis and fracture (Becker et al. 2003; Bilici et al. 2002;
Csermely et al. 2007; Meaney and OKeane 2007).

Prolactin and Cancer Risk

Long-term treatment with antipsychotics in mice has led to the development of pituitary adenomas and adenocarcinomas. Pharmacovigilance studies in humans have found a significantly greater proportion
of pituitary tumors reported with risperidone than with the other available antipsychotics (Doraiswamy et al. 2007), although the increased
level of prolactin monitoring (and subsequent computed tomographic
or magnetic resonance imaging) in risperidone-treated patients may account for a significant aspect of this discrepancy. Another prospective
study found that pituitary volume significantly increases over 1 year in
people treated with prolactin-elevating antipsychotics (MacMaster et al.
2007), although the true risk of pituitary tumor with prolactin-elevating
medications remains unclear. Among the few studies that have addressed whether dopamine receptor antagonists increase risk of breast
cancer, results have been conflicting (Wang et al. 2002). Some recent evidence suggests that endometrial cancer is related to a hormonal imbal-


Medical Illness and Schizophrenia

ance and that endometrial thickening and cancer may be related to

prolactin elevations secondary to antipsychotic treatment (Yamazawa
et al. 2003). One report estimated a fivefold increase in endometrial cancer risk with use of prolactin-raising antipsychotics in premenopausal
women (Yamazawa et al. 2003).
In summary, the etiology of sexual dysfunction in schizophrenia is
complex, although elevated serum prolactin levels contribute to a majority, but not all, of the impairments noted in the schizophrenia population. Hormonal side effects are clearly related to prolactin elevation;
however, other effects of hyperprolactinemia on osteoporosis and cancer risk remain somewhat speculative and warrant further study.

Medication Effects and Sexual Dysfunction

As mentioned above, medications can be associated with sexual dysfunction through a variety of mechanisms, and Table 121 lists common
drugs that have been associated with impairments in sexual functioning. In this chapter, we focus primarily on antipsychotic medications
but also comment on sexual dysfunction associated with other commonly used treatments in this population.

First-Generation Antipsychotics
First-generation antipsychotics (FGAs) have been associated with a variety of symptoms of sexual dysfunction. Diminished libido, orgasm
changes and/or anorgasmia, erectile dysfunction, delayed or retrograde ejaculation, oligomenorrhea, amenorrhea, galactorrhea, and gynecomastia have all been reported in patients treated with FGAs.
Overall estimates of sexual dysfunction in patients treated with FGAs
range from 30% to 70% (Dossenbach et al. 2006). Smith et al. (2002) reported that males treated with FGAs were 6.3 times more likely to complain of sexual dysfunction and have 3.7 times greater rates of erectile
dysfunction and 16.4 times greater rates of ejaculatory dysfunction than
the general male population, whereas females treated with FGAs were
9.6 times more likely to complain of orgasmic dysfunction than the general female population.
Haloperidol and other high-potency dopamine antagonists are associated with greater prolactin elevation and hormonal side effects compared with low-potency agents, with data suggesting a correlation
between serum haloperidol concentrations and serum prolactin levels
(Byerly et al. 2007; Shim et al. 2007). Studies have reported that threefourths of females and one-third of males experience hyperprolactin-

Sexual Dysfunction


emia when taking FGAs (Smith et al. 2002). The prevalence of sexual
dysfunction among patients receiving haloperidol may reach 70%
and includes reduced libido in up to 68% and erectile dysfunction in
40%50% (Dossenbach et al. 2006; Olfson et al. 2005). The lowerpotency phenothiazines are associated with ejaculatory problems and
retrograde ejaculation, the result of -adrenergic blockade. Retrograde
ejaculation may occur in up to 60% of males treated with thioridazine,
and priapism is also reported with certain low-potency antipsychotics,
most notably chlorpromazine and thioridazine.

Second-Generation Antipsychotics
Compared with the FGAs, second-generation antipsychotics (SGAs)
are associated with fewer overall sexual disturbances, with the exception of risperidone and paliperidone, which have a relatively high affinity for the dopamine D 2 receptor and which are associated with
hyperprolactinemia to a greater degree than the more potent dopamine
antagonist haloperidol. With the exception of these two agents, comparative studies have shown lower rates of sexual dysfunction with the
SGA class. The following text details information regarding prolactin
elevations and sexual dysfunction with use of the SGAs available in the
United States.
Risperidone. Risperidone, a serotonin-dopamine antagonist, is associated with a greater risk for hyperprolactinemia compared with most
antipsychotics (Kelly and Conley 2006). Mean prolactin levels may
reach 3060 ng/mL at therapeutic dosages of risperidone (Knegtering
et al. 2003); however, prolactin levels are typically higher in women and
have been reported to be 60150 ng/mL in women at routine risperidone dosages (Kelly and Conley 2006). In the Clinical Antipsychotic
Trials of Intervention Effectiveness (CATIE) schizophrenia trial, risperidone was associated with an exposure-adjusted prolactin increase of
13.8 ng/mL during treatment; however, the data were not broken
down by gender, and the population was 74% male (Lieberman et al.
Risperidone treatment has been reported to cause dysfunction in all
areas of sexual responselibido, arousal (erectile dysfunction, ejaculatory difficulties, decreased vaginal lubrication), and orgasmas well as
to result in hormone-related effects such as menstrual irregularities,
amenorrhea, galactorrhea, and gynecomastia (Dossenbach et al. 2006;
Kelly and Conley 2004). In observational studies, sexual dysfunction
during risperidone treatment occurs in 60%70% of patients with


Medical Illness and Schizophrenia

schizophrenia (Dossenbach et al. 2006). Results from comparative studies also indicate that risperidone is associated with higher rates of sexual dysfunction compared with other antipsychotics (Kelly and Conley
2006). In the CATIE trial, 27% of subjects taking risperidone had decreased sex drive, arousal, or ability to reach orgasm; 4% experienced
gynecomastia or galactorrhea; and 18% of the women reported menstrual irregularities. These rates may be underestimates due to the lack
of specific rating instruments for assessing these side effects (Lieberman et al. 2005). Of 860 subjects treated with risperidone for 12 months
in the Intercontinental Schizophrenia Outpatient Health Outcomes
study, 60% experienced loss of libido and 46% had impotence and other
types of sexual dysfunction (Dossenbach et al. 2006). Other reports on
risperidone have noted rates of menstrual abnormalities as high as
50%100% (Kelly and Conley 2006). Risperidones -adrenergic antagonism is also thought to play a role in the occurrence of sexual side effects, and this may partially explain why prolactin elevations with
risperidone are not always correlated with sexual dysfunction in males.
Olanzapine. Olanzapine causes transient elevations in plasma prolactin levels due to its antagonist properties at the D 2 receptor, but the inhibitory potential (Ki) is approximately one-third that of risperidone.
The mean prolactin level of patients during daily treatment with olanzapine at dosages of 1030 mg is 17 ng/mL, which is higher than that of
normal controls, drug-free patients, and clozapine-treated patients
(Markianos et al. 2001). In most studies, olanzapine has been shown to
cause less prolactin elevation than haloperidol and risperidone, and its
impact on serum prolactin levels appears to be a dosage-related phenomenon (Crawford et al. 1997). With prolonged olanzapine exposure,
prolactin levels generally return to normal in most patients (Kelly and
Conley 2004); however, in a 28-week study one-third of olanzapinetreated patients continued to have measurable elevations in prolactin
levels (Tran et al. 1997). In the CATIE schizophrenia trial, prolactin
change during olanzapine treatment was 8.1 ng/mL (exposureadjusted mean) (Lieberman et al. 2005). Because olanzapine use in treatment-naive subjects is associated with modest increases, this decrease
in serum prolactin represents the removal of prolactin-elevating effects
of previous treatments.
Dossenbach et al. (2006) reported in one comparative clinical trial
that the odds of patients perceiving sexual problems or being unable to
perform sexually was greater in risperidone and haloperidol groups
than in the olanzapine group, and that females taking olanzapine (14%)
or quetiapine (8%) experienced fewer menstrual irregularities than

Sexual Dysfunction


those taking risperidone (23%) or haloperidol (29%). Other researchers

have also reported fewer sexual impairments in patients taking olanzapine than in patients taking FGAs or risperidone (Bobes et al. 2003;
Knegtering et al. 2006). In the CATIE schizophrenia trial, decreased sexual drive, arousal, or ability to reach orgasm occurred in 27% of patients
(Lieberman et al. 2005), and menstrual abnormalities occurred in 12% of
women. Prescribing information derived from olanzapine clinical trials
cites dysmenorrhea rates of 2%, which are equal to or less than placebo,
and infrequent (<1%) complaints of abnormal ejaculation, amenorrhea,
breast pain, decreased menstruation, female lactation, gynecomastia,
impotence, increased menstruation, menorrhagia, and metrorrhagia
(Zyprexa Prescribing Information 2008). However, as has been reported
with other medications, these percentages are likely to be underestimates of the true sexual dysfunction incidence in a real-world population. Priapism associated with olanzapine use is rare but has been
reported in the literature, as has one case report of olanzapine-associated
clitoral priapism (Bucur and Mahmood 2004; Childers et al. 2003;
Compton and Miller 2001; Jagadheesan et al. 2004; Songer and Barclay
Quetiapine. Quetiapine is a serotonin-dopamine antagonist with a low
affinity for the D 2 receptor and a quick rate of dissociation. Quetiapine
has not been shown to have a propensity to increase prolactin levels,
which suggests that the risk of hyperprolactinemia-related adverse effects is less than with other antipsychotic agents such as risperidone
and paliperidone. Prolactin levels have been reported to decrease during quetiapine treatment, and did not differ from placebo in clinical trials (Arvanitis and Miller 1997). In the CATIE schizophrenia trial,
exposure-adjusted mean prolactin changes decreased 10.6 ng/mL during quetiapine treatment, although this probably represents a removal
of prior drug effect rather than any direct prolactin-lowering property
of quetiapine.
Despite the prolactin-sparing effects of quetiapine, sexual side effects
have been reported, and published studies have noted sexual dysfunction occurring in 8%50% of patients treated with quetiapine (Kelly and
Conley 2006). In a study of 36 patients with schizophrenia receiving
quetiapine for over 4 weeks, Atmaca et al. (2005) found that diminished
libido occurred in about one-third of male and female patients. In a
study of schizophrenia patients, Bobes et al. (2003) reported that sexual
dysfunction occurred in 18% of quetiapine-exposed subjects compared
with almost 40% of patients taking haloperidol. Byerly et al. (2006a), in
a one-time rating of a large group of patients with schizophrenia or


Medical Illness and Schizophrenia

schizoaffective disorder, also found the severity of sexual dysfunction

to be lower in the quetiapine group than in the risperidone or olanzapine group. In a randomized, open-label, 6-week trial, Knegtering et al.
(2004) found that 4 of 25 (16%) quetiapine-treated patients and 12 of 24
(50%) risperidone-treated patients experienced sexual dysfunction. Byerly et al. (2004) performed an open-label switch to quetiapine and
found that eight patients (seven switched from risperidone and one
from haloperidol) who described significantly improved sexual function had decreased prolactin levels. Fewer menstrual irregularities also
occur in females taking this medication compared with risperidone and
FGAs, with the prevalence in the CATIE schizophrenia trial reported at
6% (Kelly and Conley 2006; Lieberman et al. 2005). Finally, a 12-week
double-blind trial comparing quetiapine to risperidone found that quetiapine was associated with better improvement in sexual function
compared with risperidone or fluphenazine (Kelly and Conley 2006),
but that approximately one-half of subjects taking quetiapine were experiencing sexual dysfunction, indicating that prolactin-independent
mechanisms may be at work in the etiology of sexual complaints.
Prescribing information for quetiapine lists the following as adverse
effects: infrequently reported adverse effects in clinical trials included
(<1%) dysmenorrhea, menorrhagia, impotence, abnormal ejaculation,
amenorrhea and leucorrhea, while gynecomastia was rarely (<0.1%) reported (Seroquel Prescribing Information 2008). Several case reports of
quetiapine-associated priapism have been reported, including one reported in quetiapine overdose (Casiano et al. 2007; Davol and Rukstalis
2005; Harrison et al. 2006; Pais and Ayvazian 2001).
Ziprasidone. Ziprasidone is a serotonin dopamine antagonist with a D2
affinity lower than that of risperidone and higher than that of olanzapine. It is associated with only minor prolactin-elevating effects, which
are measurable but are generally transient and not clinically problematic. In a 52-week double-blind study, prolactin levels at endpoint were
19 ng/mL for ziprasidone, compared with 60 ng/mL for risperidone
(Ananth et al. 1998). Scattered case reports of prolactin elevation have
appeared in the literature, but in the CATIE schizophrenia trial, the exposure-adjusted mean prolactin decreases were 5.6 ng/mL during
ziprasidone treatment (Lieberman et al. 2005). The effects of ziprasidone on sexual dysfunction have not been extensively reported in the
published literature, but the CATIE study found decreased sex drive,
arousal, and ability to reach orgasm in 19% of ziprasidone-treated subjects and menstrual irregularities in 14% of women. Prolonged erections and priapism have rarely been associated with ziprasidone use,

Sexual Dysfunction


but in clinical trials, impotence, abnormal ejaculation, amenorrhea,

menorrhagia, female lactation, metrorrhagia, male sexual dysfunction,
and anorgasmia were listed as infrequent (<1%) adverse effects, and gynecomastia was listed as a rare (<0.1%) adverse event (Geodon Prescribing Information 2007). Again, these reported rates in pivotal
clinical trials likely underestimate the true prevalence of sexual dysfunction.
Aripiprazole. Aripiprazole differs in mechanism from other SGAs in
that it acts as a partial dopamine agonist at D2 receptors. The mechanism of action suggests a lack of prolactin elevation, which has been
confirmed in the published literature. Aripiprazole exerts dopamine
agonism in pituitary cells at the molecular level; thus, lactotroph cells
are not inhibited by aripiprazole treatment. Serum prolactin levels have
been found to decrease significantly from baseline during treatment trials with aripiprazole. In a 6-week, double-blind, randomized trial by
Kane et al. (2007), mean prolactin levels decreased in the aripiprazole
group from 33.4 ng/mL to 5.2 ng/mL, whereas prolactin levels in the
perphenazine group remained unchanged (35.8 ng/mL to 35.5 ng/mL).
In a 4-week study, prolactin levels decreased significantly in the aripiprazole group (9.0 ng/mL), whereas the risperidone group experienced a mean increase of 55.4 ng/mL (H.Y. Chan et al. 2007).
The prevalence of sexual dysfunction with aripiprazole has not been
widely studied. In a recent review of antipsychotic trials published between 2002 and 2008, Baggaley (2008) reported aripiprazole to be associated with the lowest rate of sexual dysfunction. An open-label study
found improvements in prolactin levels, hormonal side effects, and sexual dysfunction in patients who switched to aripiprazole or who had
aripiprazole added to current regimens (Mir et al. 2008). In a randomized, double-blind trial, adding adjunctive aripiprazole to haloperidol
significantly decreased prolactin levels and improved hormonal side
effects in the majority of women (Shim et al. 2007), although this study
did not measure sexual dysfunction. In premarketing trials, less than
1% of patients complained of irregular menstruation, amenorrhea,
breast pain, or erectile dysfunction, and less than 0.1% of patients complained of gynecomastia and priapism (Abilify Prescribing Information
2008), although two cases of priapism have been reported in patients
treated with aripiprazole (Mago et al. 2006; Negin and Murphy 2005).
Paliperidone. Paliperidone is the active metabolite of risperidone
(9-hydroxyrisperidone), with similar affinity for D 2 receptors, but is
marketed as a unique drug entity. Data from three 6-week industry-


Medical Illness and Schizophrenia

sponsored trials showed that median prolactin elevations were 81 ng/

mL for women and 24 ng/mL for men in a pooled analysis of all dosages. Four percent of patients receiving extended-release paliperidone
15 mg/day reported potential prolactin-related adverse events,
whereas 1%2% of patients receiving placebo or extended-release paliperidone 312 mg/day spontaneously reported potential prolactinrelated events (Meltzer et al. 2008).
No controlled trials specifically addressing sexual dysfunction in patients using paliperidone have been published to date, but paliperidones propensity to increase serum prolactin is nearly identical to that
of risperidone, and most of the information regarding this compound
has been reported in studies using the parent compound, risperidone.
Knegtering et al. (2005) and Melkersson (2006) demonstrated in small
studies of risperidone that plasma concentration of 9-hydroxyrisperidone correlated with increases in prolactin and may be the main contributor to increased levels of serum prolactin in many patients treated
with risperidone. One double-blind, 6-week, randomized, placebocontrolled study (Marder et al. 2007) of 444 subjects (222 receiving paliperidone) found that prolactin-related adverse effects were reported in
two men and one woman taking paliperidone (lactation and decreased
libido) and two patients taking placebo. No systematic assessment of
sexual function was utilized, however.
Clozapine. Clozapine has weak affinity for D2 receptors and dissociates
fairly quickly, resulting in only transient prolactin elevations (Volavka
et al. 2004). These transient changes are measurable in the laboratory,
but clinical hyperprolactinemia generally does not occur with this antipsychotic. Not surprisingly, most studies suggest that clozapine has
less impact on sexual functioning than do other antipsychotics, especially risperidone and FGAs. Nonetheless, clozapine does have high affinity for muscarinic cholinergic and -adrenergic receptors, and may
induce sexual dysfunction through these mechanisms. An open-label
study of 103 subjects taking clozapine found that 57 (55%) patients reported sexual side effects. Interestingly, only three of the patients reported these complaints based on an open question as to whether they
experienced side effects from clozapine, whereas the other 54 patients
reported sexual dysfunction during systematic inquiry (Yusufi et al.
2007). Although prolactin-related adverse effects are unlikely to occur,
anticholinergic and antiadrenergic activity of clozapine can be associated with ejaculatory or erectile problems in males. Case reports of priapism and impotence have been reported with clozapine treatment
(Compton and Miller 2001). According to prescribing information, 1%

Sexual Dysfunction


of patients may experience abnormal ejaculation (Clozaril Prescribing

Information 2005). In the United States, clozapine is reserved for use in
patients with schizophrenia that is refractory to other antipsychotics;
thus, it is difficult to determine whether illness severity or other potential differences within this patient population may impact sexual dysfunction prevalence.
Other Frequently Used Medications. People with schizophrenia are often treated with other psychotropic medications that have the potential
to cause sexual dysfunction. Antidepressants may be associated with
sexual side effects in differing degrees, based on their mechanisms of
action. Although most published data regarding sexual side effects in
patients treated with antidepressants have come from the depression
literature (e.g., Clayton and Montejo 2006; Segraves 2007; Werneke et al.
2006), it is expected that these medication effects would generalize to
other populations, such as patients with anxiety or schizophrenia.
SSRIs are associated with anorgasmia and delayed orgasm in up to
30%40% of treated patients, decreased libido in up to 20% of patients, and erectile difficulties in up to 10% of patients.
Tricyclic antidepressants, due to varied noradrenergic, serotonergic,
cholinergic, and histaminic effects, have been associated with sexual
Monoamine oxidase inhibitors may be associated with sexual dysfunction through effects on the serotonergic system.
Trazodone, at dosages 150 mg, has been associated with priapism in
rare cases, thought to be at least partially explained by -adrenergic
Other antidepressants, such as mirtazapine, are thought to exert
fewer sexual side effects; however, erectile dysfunction may occur
as a result of noradrenergic activity.
Venlafaxine, duloxetine, and bupropion have been associated with
less sexual dysfunction than have SSRIs.
Valproate, although not extensively reported to cause sexual dysfunction, has also been associated with reproductive endocrine abnormalities (Morris and Vanderkolk 2005). Limited data are available
regarding sexual dysfunction risk from other mood-stabilizing agents
and benzodiazepines; however, reports suggest that treatment with the
following may contribute in varying degrees to sexual dysfunction:
lithium; anticonvulsants such as carbamazepine, valproate, or lamotrigine; and benzodiazepines (Ghadirian et al. 1982; Herzog et al. 2005).


Medical Illness and Schizophrenia

The assessment of sexual dysfunction among patients with schizophrenia has not been uniform in clinical research or practice. Few standardized and validated instruments have been used in studies of sexual
dysfunction for this patient population, and definitions of sexual dysfunction and the parameters measured within these instruments are
quite heterogeneous. Nonetheless, the higher prevalence rates reported
following focused inquiry suggest that in clinical practice, directly addressing sexual function may be the most effective way to identify patient concerns about sexual disturbances. According to several reports,
patients may be unlikely to spontaneously report sexual dysfunction for
a number of reasons, and women may be less likely to report sexual dysfunction or suspected sexual side effects than men (Knegtering et al.
1999). Knegtering et al. (2006) demonstrated that although only 8.7% of
patients complained of sexual dysfunction spontaneously after 6 weeks
of antipsychotic treatment, 30.4% of patients in the study reported sexual dysfunction in response to a semistructured interview. In a survey of
individuals with severe mental illness (not exclusively schizophrenia),
patients reported various reasons they had not discussed sexual function with their clinicians, including they thought it would be embarrassing for them (28%), they thought nothing could be done (20%), and their
sexual function did not bother them (16%) (Rosenberg et al. 2003).
Recommendations from consensus panels, such as the Mt. Sinai Conference on Health Monitoring of Patients With Schizophrenia, indicate
that mental health providers should directly question patients about
changes in menstruation, libido, galactorrhea, and erectile and ejaculatory function. If patients are receiving a prolactin-elevating antipsychotic, direct patient questioning should occur at each visit until a
stable dosage is reached, then at least yearly thereafter (Marder et al.
2004). Clinicians should understand that broaching the subject of sexual
behavior and questioning patients about changes in sexual function are
integral to the treatment of patients with schizophrenia, because of
quality-of-life reasons and the impact that sexual side effects may have
on medication adherence.
Measuring sexual dysfunction in patients with schizophrenia has received far less attention than in the general population, and few biological indices may be reliably measured to screen for sexual dysfunction.
Serum prolactin levels can be assessed, but these do not reliably predict
sexual dysfunction (Knegtering et al. 2008). Most assessment techniques have employed rating instruments, but Kelly and Conley (2004)
noted that for studies published prior to 2002, a total of 15 different

Sexual Dysfunction


scales were employed in the 16 published studies. Since then, a few

scales have become validated and more widely used in this patient population, yet no gold-standard instruments exist. Patients with schizophrenia have fewer partners and relationships than do people in the
general population; therefore, not all rating instruments reliably assess
sexual complaints of schizophrenia patients.
Much of the existing literature for sexual assessment in patients with
mental illness comes from studies in depression. Clayton (2001) put
forth criteria for assessment tools to measure sexual dysfunction in patients with depression, and subsequent authors (Kelly and Conley 2004)
suggested that these criteria may also be useful in the schizophrenia
population. Among those elements thought to be most useful for any
assessment of sexual dysfunction in patients with schizophrenia are the
Assess premorbid and lifelong function compared to current
Track changes across time
Assess phase-specific sexual functioning
Separate medication effects from illness
Include gender-specific questions
Be brief
Be nonintrusive
Table 122 summarizes some of the most commonly used rating scales
and information to help guide clinicians in the assessment of these

Psychosocial Issues to Discuss
Both men and women have reported that one area with the highest
proportion of unmet needs is counseling about intimate relationships.
Additionally, studies addressing sexual issues have generally concluded
that patients with schizophrenia are prepared and open to discuss issues
relating to sexual activity, and the majority of patients with schizophrenia believe that discussing sexual issues may actually be beneficial for
their outcomes (Lewis and Scott 1997). Furthermore, concern about sexual dysfunction may exacerbate psychiatric symptoms (Sullivan and
Lukoff 1990).


TABLE 122. Rating instruments for sexual function used in schizophrenia



Number of


Bous et al. Semistructured Based on UKU sex- Based on UKU; libido,

ual dysfunction
orgasm, erection/
items (4 for men, ejaculation, vaginal
5 for women)
McGahuey Self-report
5 for men
Drive, arousal, penile
et al. 2000
5 for women
erection, vaginal
lubrication, orgasm
Changes in Sexual Clayton et Semistructured 36 for men
Sexual desire/frequency,
al. 1997
34 for women
desire/interest, sexual
pleasure, arousal/interest,
Dickson Glazer
Dickson et Self-report,
32 for men
Sexual desire, arousal,
al. 2001
computerized 40 for women
frequency of activity,
orgasmic function,
satisfaction, sexually
related problems; questions
on perceptions of sexual
side effects

Used in schizophrenia or
antipsychotic studies?

Yes. Validated, found

reliable (Byerly et al.


Medical Illness and Schizophrenia

in Sexual
Arizona Sexual
Experience Scale


TABLE 122. Rating instruments for sexual function used in schizophrenia (continued)



Udvalg for
(sexual items)

15 for men

Lingjaerde Clinician rated 48 items related

et al. 1987
to adverse
10 items on

UKU=Udvalg for Kliniske Undersogelser.

Used in schizophrenia or
antipsychotic studies?

Erectile function, orgasm,
desire, intercourse,
satisfaction, overall
Frequency of sexual thoughts, Yes
erections, masturbation over
past 2 years and immediate
2 weeks; changes in specific
aspects of sexual function
3 components: adverse effects Yes
(psychic, neurological,
autonomic, other), global
assessment of presence/
absence of interference in
daily performance due to
adverse effects, and
indication of consequences
of clinician taking action to
address side effects



15 for men


Sexual Dysfunction

Rosen et al. Self-report

Index of Erectile
Sexual Functioning Burke et al. Structured

Number of


Medical Illness and Schizophrenia

Patients who are switched to less sexually adverse medications must

be counseled regarding the expected impact on all aspects of sexual and
hormonal functioning. Because symptoms may improve, libido and the
desire to engage in intimate relationships may increase. Several cases
have been reported of patients being switched to an SGA from an FGA
with resultant unplanned pregnancies, potentially related to improvements in sexual function as well as the return of fertility (Dickson and
Hogg 1998; Neumann and Frasch 2001; Tnyi et al. 2002). Patients with
schizophrenia often have poor judgment and report fairly frequent sexual activity with people who are known to be injection drug users. Fifty
percent of patients with schizophrenia in one study were reported to be
involved in sex exchange behavior (sex bought or sold for money,
drugs, or goods), and condom use was particularly low, with less than
10% of these individuals using protective measures (Cournos et al.
1994). Moreover, women with schizophrenia not wishing to become
pregnant do not commonly use contraception (Miller 1997). Sexual
activity also occurs with those known to be infected with human immunodeficiency virus (HIV) or who are at risk, with documented HIV
infection rates that are higher in patients with severe mental illness
compared with the general population (Blank et al. 2002). Furthermore,
rates of HIV infection have been increasing substantially in adults with
serious mental illness (Otto-Salaj and Stevenson 2001). (For a more extensive discussion of HIV, see Chapter 10, HIV and Hepatitis C in
Patients With Schizophrenia.)
Women with schizophrenia also have more abortions and are more
often victims of violence during pregnancy than those without mental
illness (Miller and Finnerty 1996). Therefore, clinicians should be ready
to deal with these issues and refer patients for additional counseling if
needed. To help with the sexual dysfunction of patients with schizophrenia, clinicians should arrange for family planning, education, and
contraceptive counseling as an integral part of the comprehensive treatment plan.

Prolactin Monitoring
Despite variations in serum prolactin levels, both within and across individuals, a reasonable consensus exists regarding the upper limit of
the normal range. Conservative reports consider >1820 ng/mL in men
and >24 ng/mL in women as the standard for the upper limit (Jung et
al. 2005, 2006; Kinon et al. 2003, 2006). Not all prolactin elevations are
symptomatic, however, and levels associated with sexual dysfunction
and hormonal side effects vary widely. Nevertheless, women have sig-

Sexual Dysfunction


nificantly greater elevations from dopamine antagonists than do males,

and prolactin levels of 80150 ng/mL are in most cases correlated with
clinically evident hormonal side effects. The recommendations from the
2004 Mt. Sinai Conference on Health Monitoring of Patients With
Schizophrenia include directly asking patients about changes in menstruation, libido, erectile or ejaculatory function, and galactorrhea,
which may be prolactin-related adverse effects, and if patients are receiving antipsychotics with greater propensity to elevate prolactin, the
authors recommend more frequent assessment. With a positive
screen for these adverse events, prolactin levels should be drawn and,
if possible, other potential causes for these symptoms should be ruled
out (Marder et al. 2004). In the Practice Guideline for the Treatment of Patients With Schizophrenia, the American Psychiatric Association (2004)
suggests initial or baseline screening for symptoms of hyperprolactinemia, and, if indicated on the basis of clinical history, a baseline serum
prolactin level. These guidelines recommend follow-up screening for
symptoms of hyperprolactinemia at each visit until stable, then yearly
if the patient is taking an antipsychotic known to increase prolactin.
Prolactin levels at follow-up are recommended if indicated on the basis
of clinical history.

Medication Management
No universal guidelines exist for managing sexual dysfunction in
schizophrenia. Once sexual symptoms are identified, the degree of discomfort and dissatisfaction associated with the symptoms should be assessed, especially because sexual side effects are often of concern to
patients and may affect medication adherence if the patient attributes
sexual dysfunction to his or her treatment regimen. Interventions for the
management of sexual dysfunction in patients with schizophrenia
should be undertaken after careful consideration of all etiologies because sexual disturbances are complex and may arise from various
sources. In addition to antipsychotic treatment, symptoms of schizophrenia itself and the psychosocial problems it entails are likely to impact sexuality. Other mental and physical conditions and diseases (e.g.,
depression, diabetes, nicotine dependence, cardiovascular disease) and
medications (e.g., antidepressants, anticholinergic agents, antihypertensives) may contribute to sexual dysfunction. When possible, elucidation of the likely source of sexual dysfunction will assist in preparing a
plan for management, beginning with the identification and elimination
of potential nonpsychiatric causes. Management of sexual dysfunction
related to schizophrenia also depends, in part, on the presenting symp-


Medical Illness and Schizophrenia

toms. When addressing symptoms thought to be related to hyperprolactinemia, other reasonable clinical alternatives should be discarded
prior to focusing on the potential to alleviate the prolactin elevation.
Dose Lowering and Antipsychotic Switching. As recognized previously
in this chapter, different antipsychotic agents lead to different sexual
dysfunction risks, with FGAs as a group contributing greater sexual
side effect burden than SGAs. Among the SGAs, risperidone and extended-release paliperidone are more likely to be associated with increases in prolactin and the risk of sexual and neuroendocrine
dysfunction. Lowering the dosage of the antipsychotic is a consideration, but clinicians must use caution to avoid decreasing an antipsychotic below its therapeutic threshold for a specific patient; moreover,
antipsychotic agents with high levels of dopamine antagonism are associated with prolactin elevation even at low dosages, and no formal
studies have been published to support a dosage reduction for sexual
side effects (Costa et al. 2006).
Switching to another antipsychotic with less propensity to cause sexual side effects entails some risk of psychotic relapse during the switch,
but this option may be effective for certain patients (Marder et al. 2004),
with some positive data indicating the benefits of successful antipsychotic switching. Kinon et al. (2006) openly switched 27 stable schizophrenia patients with hyperprolactinemia to olanzapine, and found
reductions in mean serum prolactin, increased free testosterone levels,
and improvements in sexual functioning. Nakajima et al. (2005) examined the effects of openly switching 25 female schizophrenia patients
with hyperprolactinemia to quetiapine. Although eight patients
dropped out due to psychotic exacerbation, the 17 patients who completed the 8-week study (68%) showed significant decrease in serum
prolactin without significant change in illness severity. Byerly et al.
(2008) reported the differential effects of switching patients with schizophrenia and sexual dysfunction to quetiapine versus continuing risperidone. Forty-two patients with risperidone-associated sexual
dysfunction were randomized to double-blind quetiapine versus risperidone for 6 weeks. This study did not find significant differences in
rated levels of sexual dysfunction between the quetiapine and risperidone groups, although total scores were slightly lower in the switch
group at endpoint. Mir et al. (2008) reported data from an open-label
switch to aripiprazole from other antipsychotics (N= 27), and found reduced prolactin levels, improved libido, reduction of menstrual and
ejaculatory difficulties, and reduced menstrual dysfunction, accompanied by overall improvement in satisfaction in sexual functioning. By

Sexual Dysfunction


the end of the study, over half of the patients were still taking their original antipsychotic in addition to aripiprazole.
Adjunctive Treatment. The use of adjunctive medication for the treatment of sexual dysfunction in patients with schizophrenia has not been
as widely studied as the treatment of sexual dysfunction in the general
population. The lack of large, well-designed, controlled longitudinal
trials for the treatment of sexual dysfunction in patients with schizophrenia gives rise to difficulties in recommending any intervention
with a high degree of confidence. Most of the information regarding
management of sexual dysfunction in schizophrenia is focused on those
symptoms thought to be adverse effects of the antipsychotic agents and
consists mainly of small open-label studies and case reports, although a
few small, randomized, placebo-controlled studies have been published.
Sildenafil, an agent that improves erectile dysfunction via selective
inhibition of cyclic guanosine monophosphatespecific phosphodiesterase type 5 (PDE5), has been used in schizophrenia in one doubleblind, placebo-controlled, crossover study (Gopalakrishnan et al. 2006);
two open-label trials (Atmaca et al. 2002; Aviv et al. 2004); and several
case reports. In the double-blind trial, sildenafil 2550 mg was given to
32 patients with schizophrenia or delusional disorder and antipsychotic-induced erectile dysfunction. Sildenafil was superior to placebo
on all measures, including erections and satisfaction with sexual intercourse. In a subgroup with elevated prolactin (n =22), sildenafil differed
from placebo in satisfactory erections, duration of erections, and satisfaction with sexual intercourse. Adverse events were consistent with
trials in the population without schizophrenia and included nasal stuffiness and headache (Gopalakrishnan et al. 2006). One open-label study
reported that the use of sildenafil 50100 mg in 10 patients with olanzapine-induced erectile dysfunction resulted in significant differences in
erectile dysfunction scores (Atmaca et al. 2002). The second study found
that sildenafil 2575 mg in 12 male patients with schizophrenia taking
risperidone resulted in either partial or great improvement in 67% of
patients during open-label treatment (Aviv et al. 2004).
Vardenafil is another PDE5 inhibitor with data for use in patients with
schizophrenia. One open-label 12-week trial of 21 completing outpatients
found that vardenafil 1020 mg was associated with significant improvements in orgasmic function, sexual desire, intercourse satisfaction, and
overall satisfaction. Improvements in sexual function were also associated with improvement in quality of life as measured by the Quality of
Life Scale (Mitsonis et al. 2008). Adverse events, which were similar to


Medical Illness and Schizophrenia

those reported with other PDE5 inhibitors, most commonly included

headache, flushing, nasal congestion, dyspepsia, and orthostatic hypotension. One subject withdrew from the study due to headache.
Bromocriptine is a dopamine agonist that has been reported to improve libido in patients with hyperprolactinemia (Kelly and Conley
2004). Although use of the dopamine agonists has been concerning due
to the potential for exacerbating psychotic symptoms, three small studies have investigated bromocriptine for sexual dysfunction. A small,
open-label study of bromocriptine 510 mg/day resulted in a resumption of normal menstrual cycles in 55% of women and relief of galactorrhea in 33%, but smaller effects were seen with erectile and ejaculatory
dysfunction (Beau and Guillard 1980). In an open-label trial of bromocriptine 57.5 mg/day, of the 35 patients (24 female, 11 male) studied, improvement in impotence was seen in 66% of male subjects, and
return of menstrual cycle and relief of galactorrhea were seen in over
50% of the females with these complaints (Matsouka et al. 1986). A third
study compared bromocriptine to the herbal supplement PeonyGlycyrrhiza Decoction (PGD) in the treatment of risperidone-induced
hyperprolactinemia. In this randomized study, 20 women with schizophrenia taking risperidone received PGD or bromocriptine 5 mg/day
for 4 weeks, and were then crossed over to other treatment after a
4-week washout. Neither treatment was associated with worsening of
psychotic symptoms, and similar decreases in serum prolactin were
found with both agents; however, 56% of subjects taking PGD but only
17% of subjects taking bromocriptine had improvement in side effects
(Yuan et al. 2008).
Cabergoline, a dopamine agonist, has been found effective in a few
cases for reducing prolactin elevation in patients with antipsychoticassociated hyperprolactinemia without worsening of psychiatric symptoms (Cohen and Biederman 2001; Tollin 2000). One open-label pilot
study reported the results of 19 patients with schizophrenia and risperidone-associated hyperprolactinemia who were treated with cabergoline 0.1250.250 mg/week for 8 weeks. Treatment with cabergoline
was associated with a statistically significant decrease in prolactin, and
11 patients reported alleviation of clinical signs of hyperprolactinemia.
Cabergoline was not associated with a change in psychopathology
(Cavallaro et al. 2004). A black-box warning has been issued, however,
due to significant concern regarding cardiopulmonary toxicity with use
of cabergoline and, to a lesser extent, bromocriptine. Fifteen cases of
valvular heart disease in patients taking cabergoline had been reported
to the U.S. Food and Drug Administration by the end of 2002 (Flowers
et al. 2003).

Sexual Dysfunction


Selegiline, a monoamine oxidase inhibitor, was examined in one randomized trial for its potential use for sexual dysfunction in schizophrenia. Ten men with schizophrenia reporting sexual dysfunction while
being treated with an FGA (perphenazine or haloperidol) were randomly assigned to receive selegiline 15 mg/day or placebo for 3 weeks,
with a 2-week washout before a crossover to 3 weeks of the other treatment. No significant differences were found between the selegiline and
placebo phases on measures of sexual function, although prolactin levels decreased during selegiline exposure (Kodesh et al. 2003). Side effects were not reported in the manuscript.
Amantadine has been explored for the treatment of neuroendocrine
side effects and/or sexual dysfunction because it purportedly causes
dopamine release at neuronal terminals, and therefore may be effective
in decreasing prolactin levels that are elevated as a result of antipsychotic use. Two open-label studies have been reported. Correa et al.
(1987) found reductions in prolactin, gynecomastia and galactorrhea,
breast tenderness, decreased libido, and amenorrhea in 10 patients with
schizophrenia who were treated with open-label amantadine 200300
mg/day in a 7-week study. In the other study, Valevski et al. (1998) reported the effects of amantadine on sexual dysfunction in males with
schizophrenia who were treated with antipsychotics. Open-label amantadine 100 mg/day for 6 weeks was shown to improve patient scores on
desire, erection, and satisfaction from sexual performance. Amantadine,
however, is associated with a theoretical risk of worsening psychosis or
other psychiatric side effects due to its effects on dopamine, as well as
other dopaminergic adverse effects such as orthostatic hypotension.
Aripiprazole has not been systematically studied for its effects on improving sexual dysfunction; however, adjunctive aripiprazole has been
shown in a randomized, double-blind, 8-week trial to normalize elevated prolactin levels and improve hormonal side effects. In the aripiprazole group, 88.5% of patients at week 8 had normalized prolactin
levels compared with 3.6% of patients in the placebo group. Among the
11 female patients with menstrual disturbances randomized to aripiprazole, seven regained menstruation during the study (63.6%),
whereas no female patients did in the placebo group (Shim et al. 2007).
A number of other open-label studies (Mir et al. 2008) and case reports
of adjunctive aripiprazole with existing antipsychotics have also been
published that support the results of larger trials.
Several case reports and small studies have been published, mostly
in the Japanese literature, describing the use of herbal supplements in
the treatment of hyperprolactinemia associated with antipsychotics.
One study and a case report using Shakuyaku-kanzo-to, or TJ-68, an


Medical Illness and Schizophrenia

herbal product, have found significant reductions in prolactin levels

and hormonal side effects (Yamada et al. 1997, 1999). In one small, randomized, crossover study, the preparation Peony-Glycyrrhiza Decoction (PGD) was compared to bromocriptine in 20 women with
schizophrenia and risperidone-associated hyperprolactinemia experiencing oligomenorrhea or amenorrhea. In this small study, PGD was
more beneficial than bromocriptine for improving side effects, although
prolactin decreases were similar (Yuan et al. 2008). Little information is
available regarding the pharmacology or side effects of these agents.

Sexual dysfunction is a frequently occurring problem in patients with
schizophrenia, yet its recognition and significance have been overlooked
by clinicians. Multiple etiological factors (e.g., schizophrenia itself, vascular disease, diabetes, smoking) are likely to exist beyond the effects of
psychotropic medications, and these should not be overlooked. Many
antipsychotic medications are associated with sexual dysfunction, and
prolactin elevations contribute to a large proportion of sexual complaints, but hyperprolactinemia is not always associated with hormonal
and sexual side effects. Anticholinergic and -adrenergic mechanisms
may also be particularly important for certain patients, with males especially vulnerable to erectile problems from the latter.
The management of patients with sexual dysfunction involves optimal communication between the patient and clinician, careful exploration of potential causes and contributing factors, and assessment of
impact on the patients attitude toward treatment and quality of life.
Identifying and eliminating nonpsychiatric causes may be helpful in
addressing this problem. Treatment changes based on sexual dysfunction should be carefully considered and, when possible, targeted to the
specific symptoms. Switching an antipsychotic medication to one less
likely to cause sexual dysfunction, or in some cases reducing the dosage, has been proposed but may not always be the practical or recommended approach for certain patients. The best approaches for the
treatment of hyperprolactinemia and sexual dysfunction currently appear to be either initiating treatment with prolactin-sparing antipsychotics or decreasing prolactin levels through the use of dopamine
agonists or aripiprazole adjunctively, if the side effects are definitely
secondary to prolactin elevations. The use of PDE5 inhibitors may be of
particular value in cases with a less clear association between sexual
dysfunction and elevated prolactin, particularly in males who, for in-

Sexual Dysfunction


stance, smoke or have diabetes; however, studies are needed to help

guide evidence-based treatment. As clinicians strive to reach recovery
in patients, better communication on sexuality and sexual dysfunction
will assist greatly in this endeavor.

Key Clinical Points

Sexuality and sexual dysfunction in patients with schizophrenia have received little attention in the past.

Sexual dysfunction has been difficult to measure accurately in patients

with schizophrenia, but its prevalence is generally considered to be
higher in these patients than in the general population. Reported prevalence rates vary widely, with most publications citing rates of 50%75%.

The types and prevalence of sexual dysfunction are likely to differ

between men and women. Males with schizophrenia who have sexual
dysfunction typically complain of decreases in libido and erectile,
ejaculatory, and orgasmic functions, whereas women may experience
disturbances in libido, vaginal lubrication, and orgasm, as well as dyspareunia and menstrual disturbances.

Sexual dysfunction, especially if it is perceived by the patient to be an

effect of antipsychotic medication, may be a factor that impacts treatment adherence.

Multiple factors may contribute to sexual dysfunction, including lifestyle factors, symptoms of schizophrenia and other comorbidities, medications or substances used, and cigarette smoking.

The mechanisms by which medications affect sexual function likely

differ. Although many neurotransmitter systems (dopaminergic, serotonergic, adrenergic, cholinergic, histaminic) that are affected by antipsychotics may contribute to sexual dysfunction, elevation of serum
prolactin is thought to account for a significant proportion of sexual
dysfunction seen in patients.

The effects of prolactin elevation should be considered when screening

patients at clinic visits, especially those patients taking prolactin-elevating

First- and second-generation antipsychotics have both been associated

with sexual dysfunction, although differences exist both within and
among the different classes.


Medical Illness and Schizophrenia

Risperidone, paliperidone, and some FGAs have been associated with

the greatest increases in prolactin levels among patients treated with

The assessment of sexual dysfunction has not been widely studied. Several instruments exist, but no gold standard has been identified. Because
patients may be hesitant to broach the subjects themselves, directly questioning patients about sexual function and dysfunction should be incorporated into the treatment plan and may result in detecting dysfunction.

The management of sexual dysfunction in patients with schizophrenia

is complex. Psychosocial issues should be addressed. Medication management may be necessary, and adjunctive medications have sometimes
been reported in the literature. Only limited data exist to evaluate the
efficacy of these interventions.

Abilify Prescribing Information. Princeton, NJ, Bristol-Myers Squibb, 2008
Aizenberg D, Zemishlany Z, Dorfman-Etrog P, et al: Sexual dysfunction in male
schizophrenia patients. J Clin Psychiatry 56:137141, 1995
Aizenberg D, Modai I, Landa A, et al: Comparison of sexual dysfunction in
male schizophrenic patients on treatment with classical antipsychotics versus clozapine. J Clin Psychiatry 62:541544, 2001
American Association of Clinical Endocrinologists Male Sexual Dysfunction
Task Force: American Association of Clinical Endocrinologists medical
guidelines for clinical practice for the evaluation and treatment of male sexual dysfunction: a couples problem2003 update. Endocrine Practice
9:7795, 2003
American Psychiatric Association: Practice Guideline for the Treatment of Patients With Schizophrenia, 2nd Edition. Arlington, VA, American Psychiatric Association, 2004
Ananth J, Burgoyne K, Smith M, et al: Prolactin levels and new antipsychotics.
Paper presented at the 21st Annual Congress, European College of Neuropsychopharmacology, Glasgow, Scotland, July 1998
Arvanitis L, Miller BG: Multiple fixed doses of Seroquel (quetiapine) in patients with acute exacerbation of schizophrenia: a comparison with haloperidol and placebo. The Seroquel Trial 13 Study Group. Biol Psychiatry
42:233246, 1997
Atmaca M, Kuloglu M, Tezcan E: Sildenafil use in patients with olanzapineinduced erectile dysfunction. Int J Impot Res 14:547549, 2002
Atmaca M, Kuloglu M, Tezcan E: A new atypical antipsychotic: quetiapineinduced sexual dysfunctions. Int J Impot Res 17:201203, 2005

Sexual Dysfunction


Aviv A, Shelef A, Weizman A: An open-label trial of sildenafil addition in

risperidone-treated male schizophrenia patients with erectile dysfunction.
J Clin Psychiatry 65:97103, 2004
Baggaley M: Sexual dysfunction in schizophrenia: focus on recent evidence.
Hum Psychopharmacol 23:201209, 2008
Barnes TRE, Harvey CA: Psychiatric drugs and sexuality, in Sexual Pharmacology. Edited by Riley AJ, Peet M, Wilson C. Oxford, UK: Oxford University
Press, 1993, pp 176196
Beau Y, Guillard P: Trial therapy with bromocriptine of secondary endocrine effects after psychotropic treatment [in French]. Ann Med Psychol (Paris)
138:179186, 1980
Becker D, Liver O, Mester R, et al: Risperidone, but not olanzapine, decreases
bone mineral density in female premenopausal schizophrenia patients.
J Clin Psychiatry 64:761766, 2003
Bilici M, Cakirbay H, Guler M, et al: Classical and atypical neuroleptics and
bone mineral density in patients with schizophrenia. Int J Neurosci
112:817828, 2002
Bitter I, Basson BR, Dossenbach MR: Antipsychotic treatment and sexual functioning in first-time neuroleptic-treated schizophrenic inpatients. Int Clin
Psychopharmacol 20:1921, 2005
Blank MD, Mandell DS, Aiken L, et al: Co-occurrence of HIV and serious mental
illness among Medicaid recipients. Psychiatr Serv 53:868873, 2002
Bobes J, Garca-Portilla MP, Rejas J, et al: Frequency of sexual dysfunction and
other reproductive side-effects in patients with schizophrenia treated with
risperidone, olanzapine, quetiapine, or haloperidol: the results of the ERIE
study. J Sex Marital Ther 29:125147, 2003
Bous J, Castelein S, Kluiter H, et al: The Antipsychotics and Sexual Functioning
Questionnaire (ASFQ): a pilot study evaluating reliability and validity in
patients treated with antipsychotics (letter). Schizophr Res 60:353, 2003
Bucur M, Mahmood T: Olanzapine-induced clitoral priapism. J Clin Psychopharmacol 24:572573, 2004
Burke MA, McEvoy JP, Ritchie JC: A pilot study of a structured interview addressing sexual function in men with schizophrenia. Biol Psychiatry 35:32
35, 1994
Byerly MJ, Lescouflair E, Weber MT, et al: An open-label trial of quetiapine for antipsychotic-induced sexual dysfunction. J Sex Marital Ther 30:325332, 2004
Byerly MJ, Nakonezny PA, Bettcher BM, et al: Sexual dysfunction associated
with second-generation antipsychotics in outpatients with schizophrenia
or schizoaffective disorder: an empirical evaluation of olanzapine, risperidone, and quetiapine. Schizophr Res 86:244250, 2006a
Byerly MJ, Nakonezny PA, Fisher R, et al: An empirical evaluation of the Arizona Sexual Experience Scale and a simple one-item screening test for assessing antipsychotic-related sexual dysfunction in outpatients with
schizophrenia and schizoaffective disorder. Schizophr Res 81:311316,


Medical Illness and Schizophrenia

Byerly M, Suppes T, Tran Q, et al: Clinical implications of antipsychoticinduced hyperprolactinemia in patients with schizophrenia spectrum or
bipolar spectrum disorders. J Clin Psychopharmacol 27:639661, 2007
Byerly MJ, Nakonezny PA, Rush AJ: Sexual functioning associated with quetiapine switch vs risperidone continuation in outpatients with schizophrenia or schizoaffective disorder: a randomized double blind pilot trial.
Psychiatry Res 159:115120, 2008
Casiano H, Globerman D, Enns MW: Recurrent priapism during treatment with
clozapine, quetiapine and haloperidol. J Psychopharmacol 21:898899,
Cavallaro R, Cocchi F, Angelone SM, et al: Cabergoline treatment of risperidone-induced hyperprolactinemia: a pilot study. J Clin Psychiatry 65:187
190, 2004
Chan HY, Lin WW, Lin SK: Efficacy and safety of aripiprazole in the acute treatment of schizophrenia in Chinese patients with risperidone as an active
control: a randomized trial. J Clin Psychiatry 68:2936, 2007
Chan S, Yu IW: Quality of life of clients with schizophrenia. J Adv Nurs 45:72
83, 2004
Childers JB, Schwartz AC, Compton MT: Olanzapine-associated priapism. Psychosomatics 44:351352, 2003
Clayton AH: Recognition and assessment of sexual dysfunction associated with
depression. J Clin Psychiatry 62 (suppl 3):59, 2001
Clayton AH, Montejo AL: Major depressive disorder, antidepressants, and sexual dysfunction. J Clin Psychiatry 67 (suppl 6):3337, 2006
Clayton AH, McGarvey EL, Clavet GJ: The Changes in Sexual Functioning
Questionnaire (CSFQ): development, reliability, and validity. Psychopharmacol Bull 33:3145, 1997
Clozaril Prescribing Information. East Hanover, NJ, Novartis Pharmaceuticals,
Cohen LG, Biederman J: Treatment of risperidone-induced hyperprolactinemia
with a dopamine agonist in children. J Child Adolesc Psychopharmacol
11:435440, 2001
Compton MT, Miller AH: Priapism associated with conventional and atypical
antipsychotic medications: a review. J Clin Psychiatry 62:362366, 2001
Correa N, Opler LA, Kay SR, et al: Amantadine in the treatment of neuroendocrine side effects of neuroleptics. J Clin Psychopharmacol 7:9195, 1987
Costa AM, Lima MS, Mari J de J, et al: A systematic review on clinical management of antipsychotic-induced sexual dysfunction in schizophrenia. Sao
Paulo Med J 124:291297, 2006
Cournos F, Guido JR, Coomaraswamy S, et al: Sexual activity and risk of HIV
infection among patients with schizophrenia. Am J Psychiatry 151:228232,
Crawford AM, Beasley CM Jr, Tollefson GD: The acute and long-term effect of
olanzapine compared with placebo and haloperidol on serum prolactin
concentrations. Schizophr Res 26:4154, 1997

Sexual Dysfunction


Csermely T, Halvax L, Vizer M, et al: Relationship between adolescent amenorrhea and climacteric osteoporosis. Maturitas 56:368374, 2007
Cutler AJ: Sexual dysfunction and antipsychotic treatment. Psychoneuroendocrinology 28:6982, 2003
Davis SR, Davison SL, Donath S, et al: Circulating androgen levels and selfreported sexual function in women. JAMA 294:9196, 2005
Davol P, Rukstalis D: Priapism associated with routine use of quetiapine: case
report and review of the literature. Urology 66(880e):1314, 2005
Dickson RA, Hogg L: Pregnancy of a patient treated with clozapine. Psychiatr
Serv 49:10811083, 1998
Dickson RA, Glazer WM, Violato C: A computerized self-report questionnaire
for assessing sexual functioning in psychotic patients: the DGSF scale.
Schizophr Res 49 (suppl 1):283, 2001
Doraiswamy PM, Schott G, Star K, et al: Atypical antipsychotics and pituitary
neoplasms in the WHO database. Psychopharmacol Bull 40:7476, 2007
Dossenbach M, Hodge A, Anders M, et al: Prevalence of sexual dysfunction in
patients with schizophrenia: international variation and underestimation.
Int J Neuropsychopharmacol 8:195201, 2005
Dossenbach M, Dyachkova Y, Pirildar S, et al: Effects of atypical and typical antipsychotic treatments on sexual function in patients with schizophrenia:
12-month results from the Intercontinental Schizophrenia Outpatient
Health Outcomes (IC-SOHO) study. Eur Psychiatry 21:251258, 2006
Fan X, Henderson DC, Chiang E, et al: Sexual functioning, psychopathology
and quality of life in patients with schizophrenia. Schizophr Res 94:119
127, 2007
Finn SE, Bailey JM, Schultz RT, et al: Subjective utility ratings of neuroleptics in
treating schizophrenia. Psychol Med 20:843848, 1990
Flowers CM, Racoosin JA, Lu SL, et al: The U.S. Food and Drug Administrations registry of patients with pergolide-associated valvular heart disease.
Mayo Clin Proc 78:730731, 2003
Friedman S, Harrison G: Sexual histories, attitudes, and behavior of schizophrenic and normal women. Arch Sex Behav 13:555567, 1984
Geodon Prescribing Information. New York, NY, Pfizer, 2007
Ghadirian AM, Chouinard G, Annable L: Sexual dysfunction and plasma prolactin levels in neuroleptic-treated schizophrenic outpatients. J Nerv Ment
Dis 170:463467, 1982
Gitlin MJ: Psychotropic medications and their effects on sexual function: diagnosis,
biology, and treatment approaches. J Clin Psychiatry 55:406413, 1994
Gopalakrishnan R, Jacob KS, Kuruvilla A, et al: Sildenafil in the treatment of
antipsychotic-induced erectile dysfunction: a randomized, double-blind,
placebo-controlled, flexible-dose, two-way crossover trial. Am J Psychiatry
163:494499, 2006
Grinshpoon A, Ponizovsky AM: The relationship between need profiles, clinical symptoms, functioning and the well-being of inpatients with severe
mental disorders. J Eval Clin Pract 14:218225, 2008


Medical Illness and Schizophrenia

Haddad PM, Wieck A: Antipsychotic-induced hyperprolactinaemia: mechanisms, clinical features and management. Drugs 64:22912314, 2004
Harrison G, Dilley JW, Loeb L, et al: Priapism and quetiapine: a case report. Psychopharmacol Bull 39:117119, 2006
Herzog AG, Drislane FW, Schomer DL, et al: Differential effects of antiepileptic
drugs on sexual function and hormones in men with epilepsy. Neurology
65:10161020, 2005
Howes OD, Wheeler MJ, Pilowsky LS, et al: Sexual function and gonadal hormones in patients taking antipsychotic treatment for schizophrenia or
schizoaffective disorder. J Clin Psychiatry 68:361367, 2007
Jagadheesan K, Thakur A, Akhtar S: Irreversible priapism during olanzapine
and lithium therapy (letter). Aust N Z J Psychiatry 38:381, 2004
Jung DU, Seo YS, Park JH, et al: The prevalence of hyperprolactinemia after
long-term haloperidol use in patients with chronic schizophrenia. J Clin
Psychopharmacol 25:613615, 2005
Jung DU, Conley RR, Kelly DL, et al: Prevalence of bone mineral density loss in
Korean patients with chronic schizophrenia: a cross-sectional study. J Clin
Psychiatry 67:13911396, 2006
Kane JM, Meltzer HY, Carson WH, et al: Aripiprazole for treatment resistant
schizophrenia: results of a multicenter, randomized, double-blind comparison study versus fluphenazine. J Clin Psychiatry 68:213223, 2007
Kelly DL, Conley RR: Sexuality and schizophrenia: a review. Schizophr Bull
30:767779, 2004
Kelly DL, Conley RR: A randomized double-blind 12-week study of quetiapine,
risperidone, or fluphenazine on sexual functioning in people with schizophrenia. Psychoneuroendocrinology 31:340346, 2006
Kelly DL, Boggs DL, Conley RR: Reaching for wellness in schizophrenia. Psychiatr Clin North Am 30:453479, 2007
Kinon BJ, Gilmore JA, Liu H, et al: Prevalence of hyperprolactinemia in schizophrenic patients treated with conventional antipsychotic medications or
risperidone. Psychoneuroendocrinology 28 (suppl 2):5568, 2003
Kinon BJ, Ahl J, Liu-Seifert H, et al: Improvement in hyperprolactinemia and reproductive comorbidities in patients with schizophrenia switched from
conventional antipsychotics or risperidone to olanzapine. Psychoneuroendocrinology 31:577588, 2006
Knegtering H, Blijd C, Boks MPM: Sexual dysfunctions and prolactin levels in patients using classical antipsychotics, risperidone or olanzapine. Schizophr
Res 36:355356, 1999
Knegtering H, van der Moolen AEGM, Castelein S, et al: What are the effects of
antipsychotics on sexual dysfunctions and endocrine functioning? Psychoneuroendocrinology 28:109123, 2003
Knegtering R, Castelein S, Bous H, et al: A randomized open-label study of the
impact of quetiapine versus risperidone on sexual functioning. J Clin Psychopharmacol 24:5661, 2004

Sexual Dysfunction


Knegtering R, Baselmans P, Castelein S, et al: Predominant role of the 9-hydroxy metabolite of risperidone in elevating blood prolactin levels. Am J
Psychiatry 162:10101012, 2005
Knegtering H, Boks M, Blijd C, et al: A randomized open-label comparison of
the impact of olanzapine versus risperidone on sexual functioning. J Sex
Marital Ther 32:315326, 2006
Knegtering H, van den Bosch R, Castelein S, et al: Are sexual side effects of prolactin-raising antipsychotics reducible to serum prolactin? Psychoneuroendocrinology 33:711717, 2008
Kockott G, Pfeiffer W: Sexual disorders in nonacute psychiatric outpatients.
Compr Psychiatry 37:5661, 1996
Kodesh A, Weizman A, Aizenberg D, et al: Selegiline in the treatment of sexual
dysfunction in schizophrenic patients maintained on neuroleptics: a pilot
study. Clin Neuropharmacol 26:193195, 2003
Lacro JP, Dun LB, Dolder CR, et al: Prevalence of and risk factors for medication
nonadherence in patients with schizophrenia: a comprehensive review of
recent literature. J Clin Psychiatry 63:892909, 2002
Lambert M, Conus P, Eide P, et al: Impact of present and past antipsychotic side
effects on attitude toward typical antipsychotic treatment and adherence.
Eur Psychiatry 19:415422, 2004
Laumann EO, Paik A, Rosen RC: Sexual dysfunction in the United States: prevalence and predictors. JAMA 281:537544, 1999
Lewis J, Scott E: The sexual education needs of those disabled by mental illness.
Psychiatr Rehabil J 21:164167, 1997
Lieberman JA, Stroup TS, McEvoy JP, et al: Effectiveness of antipsychotic
drugs in patients with chronic schizophrenia. N Engl J Med 353:12091223,
Lingjaerde O, Ahlfors UG, Bech P, et al: The UKU side effect rating scale: a new
comprehensive rating scale for psychotropic drugs and a cross-sectional
study of side effects in neuroleptic-treated patients. Acta Psychiatr Scand
Suppl 334:1100, 1987
MacMaster FP, El-Sheikh R, Upadhyaya AR, et al: Effect of antipsychotics on
pituitary gland volume in treatment-naive first-episode schizophrenia: a
pilot study. Schizophr Res 92:207210, 2007
Mago R, Anolik R, Johnson RA, et al: Recurrent priapism associated with use of
aripiprazole. J Clin Psychiatry 67:14711472, 2006
Marder SR, Essock SM, Miller AL, et al: Physical health monitoring of patients
with schizophrenia. Am J Psychiatry 161:13341349, 2004
Marder SR, Kramer L, Ford E, et al: Efficacy and safety of paliperidone extended-release tablets: results of a 6-week, randomized, placebo-controlled
study. Biol Psychiatry 62:13631370, 2007
Markianos M, Hatzimanolis J, Lykouras L: Neuroendocrine responsivities of
the pituitary dopamine system in male schizophrenic patients during treatment with clozapine, olanzapine, risperidone, sulpiride, or haloperidol.
Eur Arch Psychiatry Clin Neurosci 251:141146, 2001


Medical Illness and Schizophrenia

Matsouka I, Nakai T, Miyake M, et al: Effects of bromocriptine on neurolepticinduced amenorrhea, galactorrhea and impotence. Jpn J Psychiatry Neurol
40:639646, 1986
McGahuey CA Gelenberg AJ, Laukes CA, et al: The Arizona Sexual Experience
Scale (ASEX): reliability and validity. J Sex Marital Ther 26:2540, 2000
Meaney AM, OKeane V: Bone mineral density changes over a year in young females with schizophrenia: relationship to medication and endocrine variables. Schizophr Res 93:136143, 2007
Meaney AM, Smith S, Howes OD, et al: Effects of long-term prolactin-raising
antipsychotic medication on bone mineral density in patients with schizophrenia. Br J Psychiatry 184:503508, 2004
Melkersson KI: Prolactin elevation of the antipsychotic risperidone is predominantly related to its 9-hydroxy metabolite. Hum Psychopharmacol 21:529
532, 2006
Meltzer HY, Bobo WV, Nuamah IF, et al: Efficacy and tolerability of oral paliperidone extended-release tablets in the treatment of acute schizophrenia:
pooled data from three 6-week, placebo-controlled studies. J Clin Psychiatry 69:817829, 2008
Miller LJ: Sexuality, reproduction, and family planning in women with schizophrenia. Schizophr Bull 23:623635, 1997
Miller LJ, Finnerty M: Sexuality, pregnancy, and childrearing among women
with schizophrenia-spectrum disorders. Psychiatr Serv 47:502506, 1996
Mir A, Shivakumar K, Williamson RJ, et al: Change in sexual dysfunction with
aripiprazole: a switching or add-on study. J Psychopharmacol 22:244253,
Mitsonis CI, Mitropoulos PA, Dimopoulos NP, et al: Vardenafil in the treatment
of erectile dysfunction in outpatients with chronic schizophrenia: a flexible-dose, open-label study. J Clin Psychiatry 69:206212, 2008
Morris GL 3rd, Vanderkolk C: Human sexuality, sex hormones, and epilepsy.
Epilepsy Behav 7 (suppl 2):S22S28, 2005
Nakajima M, Terao T, Iwata N, et al: Switching female schizophrenic patients to
quetiapine from conventional antipsychotic drugs: effects on hyperprolactinemia. Pharmacopsychiatry 38:1719, 2005
Negin B, Murphy TK: Priapism associated with oxcarbazepine, aripiprazole,
and lithium. J Am Acad Child Adolesc Psychiatry 44:12231224, 2005
Neumann NU, Frasch K: [Olanzapine and pregnancy: two case reports] (in German). Nervenarzt 72:876878, 2001
Olfson M, Uttaro T, Carson WH, et al: Male sexual dysfunction and quality of
life in schizophrenia. J Clin Psychiatry 66:331338, 2005
Otto-Salaj LL, Stevenson LY: Influence of psychiatric diagnoses and symptoms
on HIV risk behavior in adults with serious mental illness. AIDS Read
11:197204, 206208, 2001
Pais VM, Ayvazian PJ: Priapism from quetiapine overdose: first report and proposal of mechanism (letter). Urology 58:462, 2001

Sexual Dysfunction


Pinderhughes CA, Grace EF, Reyna LJ: Psychiatric disorders and sexual functioning. Am J Psychiatry 128:12761283, 1972
Rosen RC, Riley A, Wagner G, et al: The International Index of Erectile Function
(IIEF): a multidimensional scale for assessment of erectile dysfunction.
Urology 49:822830, 1997
Rosenberg KP, Bleiberg KL, Koscis J, et al: A survey of sexual side effects among
severely mentally ill patients taking psychotropic medications: impact on
compliance. J Sex Marital Ther 29:289296, 2003
Segraves RT: Sexual dysfunction associated with antidepressant therapy. Urol
Clin North Am 34:575579, 2007
Seroquel Prescribing Information. Wilmington, DE, AstraZeneca Pharmaceuticals, 2008
Shim J-C, Shin J-G, Kelly DL, et al: Adjunctive treatment with a dopamine partial agonist, aripiprazole, for antipsychotic-induced hyperprolactinemia: a
placebo-controlled trial. Am J Psychiatry 164:14041410, 2007
Smith SM, OKeane V, Murray R, et al: Sexual dysfunction in patients taking
conventional antipsychotic medication. Br J Psychiatry 181:4955, 2002
Songer DA, Barclay JC: Olanzapine-induced priapism. Am J Psychiatry
158:20872088, 2001
Sullivan G, Lukoff D: Sexual side effects of antipsychotic medication: evaluation and intervention. Hosp Community Psychiatry 41:12381241, 1990
Tnyi T, Trixler M, Keresztes Z: Quetiapine and pregnancy (letter). Am J Psychiatry 159:674, 2002
Teusch L, Scherbaum N, Bohme H, et al: Different patterns of sexual dysfunctions associated with psychiatric disorders and psychopharmacological
treatment: results of an investigation by semistructured interview of
schizophrenic and neurotic patients and methadone-substituted opiate addicts. Pharmacopsychiatry 28:8492, 1995
Thomas DR: Medications and sexual function. Clin Geriatr Med 19:553562, 2003
Tollin SR: Use of the dopamine agonists bromocriptine and cabergoline in the
management of risperidone-induced hyperprolactinemia in patients with
psychotic disorders. J Endocrinol Invest 23:765770, 2000
Tran PV, Hamilton SH, Kuntz, et al: Double-blind comparison of olanzapine
versus risperidone in the treatment of schizophrenia and other psychotic
disorders. J Clin Psychopharmacol 17:407418, 1997
ok A, Incesu C, Aker T, et al: Sexual dysfunction in patients with schizophrenia on antipsychotic medication. Eur Psychiatry 22:328333, 2007
Valevski A, Modai I, Zbarski E, et al: Effect of amantadine on sexual dysfunction in neuroleptic-treated male schizophrenic patients. Clin Neuropharmacol 21:355357, 1998
Volavka J, Czobor P, Cooper TB, et al: Prolactin levels in schizophrenia and
schizoaffective disorder patients treated with clozapine, olanzapine, risperidone, or haloperidol. J Clin Psychiatry 65:5761, 2004
Wang CS, Kao WT, Chen CD, et al: Priapism associated with typical and atypical antipsychotic medications. Int Clin Psychopharmacol 21:245248, 2006


Medical Illness and Schizophrenia

Wang PS, Walker AM, Tsuang MT, et al: Dopamine antagonists and the development of breast cancer. Arch Gen Psychiatry 59:11471154, 2002
Werneke U, Northey S, Bhugra D: Antidepressants and sexual dysfunction.
Acta Psychiatr Scand 114:384397, 2006
Yamada K, Kanba S, Yagi G, et al: Effectiveness of herbal medicine (shakuyakukanzo-to) for neuroleptic-induced hyperprolactinemia. J Clin Psychopharmacol 17:234235, 1997
Yamada K, Kanba S, Yagi G, et al: Herbal medicine (shakuyaku-kanzo-to) in the
treatment of risperidone-induced amenorrhea. J Clin Psychopharmacol
19:380381, 1999
Yamazawa K, Matsui H, Seki K, et al: A case-control study of endometrial cancer after antipsychotics exposure in premenopausal women. Oncology
64:116123, 2003
Yuan HN, Wang CY, Sze CW, et al: A randomized, crossover comparison of
herbal medicine and bromocriptine against risperidone-induced hyperprolactinemia in patients with schizophrenia. J Clin Psychopharmacol
28:264270, 2008
Yusufi B, Mukherjee S, Flanagan R, et al: Prevalence and nature of side effects
during clozapine maintenance treatment and the relationship with clozapine dose and plasma concentration. Int Clin Psychopharmacol 22:238
243, 2007
Zyprexa Prescribing Information. Indianapolis, IN, Eli Lilly, 2008

Managing the Health
Outcomes of
Treatment in Children
and Adolescents
Christoph U. Correll, M.D.

In children and adolescents, antipsychotics are used in increasing quantities for schizophrenia, other psychotic disorders, and various
nonpsychotic conditions (Olfson et al. 2006). For youth, as for adults,
the majority of antipsychotics used are second-generation agents,
partly because children and adolescents are especially sensitive to the
adverse neuromotor effects of first-generation antipsychotics (Correll et
al. 2006). At the time of this writing, the U.S. Food and Drug Administration (FDA) has approved only four antipsychoticshaloperidol and

This work was supported in part by the Zucker Hillside Hospital National Institute of Mental Health Advanced Center for Intervention and Services Research for the Study of Schizophrenia grant MH 074543-01, and the North
ShoreLong Island Jewish Health System Research Institute National Institutes
of Health General Clinical Research Center grant M01 RR018535.



Medical Illness and Schizophrenia

thioridazine, which are first-generation antipsychotics (FGAs), and risperidone and aripiprazole, which are second-generation antipsychotics
(SGAs)for use in pediatric schizophrenia, with approval for the two
FGAs being based predominantly on adult data and small studies in
However, with the increasing use of SGAs, their effects on body
weight and metabolic health have become increasing concerns (Correll
and Carlson 2006). Weight gain and obesity have known associations
with diabetes, dyslipidemia, and hypertension, all of which are leading
risk factors for future cardiovascular morbidity and mortality (Ebbeling
et al. 2002). Moreover, data suggest that obesity during childhood predicts poor metabolic outcomes for adults even better than obesity during adulthood (Must et al. 1992). In this chapter, I review the available
pediatric data on the side effects of antipsychotics with adverse health
impact, aiming to inform antipsychotic prescribing and monitoring
practices in this vulnerable population. Data from studies of children
and adolescents with schizophrenia are highlighted, augmented by pediatric data from studies of other disorders where appropriate. Based
on the far-reaching implications of age-inappropriate weight gain and
metabolic abnormalities, adequate monitoring, management, and,
whenever possible, prevention of adverse health outcomes in youths is
crucial to promoting long-term physical and psychological well-being,
sustained treatment adherence, and adequate role functioning and attainment of developmental milestones.

Developmental Considerations
Childhood through adolescence is a period of unparalleled development, biologically, psychologically, and socially. Although data regarding the bioavailability and bioactivity of medications in pediatric
populations are relatively slim, drug uptake, distribution, and metabolism in youth are affected by several factors that differ from those in
adults. Some of these factors relevant for medication treatment include
active tissue growth, reproductive hormone release beginning in adolescence, a higher ratio of liver organ-to-tissue mass, greater intracellular and extracellular tissue water and glomerular filtration rate, lower
protein binding, and reduced fat tissue mass in youths compared with
adults (Paxton and Dragunow 1993). Clinically, these differences translate into the general need for higher dosages per kilogram of weight in
children and adolescents compared with adults in order to reach comparable serum levels and achieve therapeutic efficacy; also, more fre-

Managing Health Outcomes in Children and Adolescents


quent doses during the day may be required in prepubertal children

(Woods et al. 2002).

Efficacy of Antipsychotics in Children and

Adolescents With Schizophrenia
Twelve, mostly recent, randomized, controlled antipsychotic trials
(N= 1,014) have shown efficacy as monotherapy in pediatric schizophrenia (Kumra et al. 2008b). Superior efficacy of antipsychotics versus
placebo in reducing scores on the Positive and Negative Syndrome
Scale (PANSS) and/or the Brief Psychiatric Rating Scale (BPRS) was
demonstrated in three randomized, placebo-controlled trials of patients
ages 1317 with schizophrenia. The SGAs used in these studies were
aripiprazole (N =294, using fixed dosages of 10 mg/day, mean 9.56 mg/
day, and 30 mg/day, mean 27.8 mg/day) (Robb et al. 2007), olanzapine
(N = 107, using flexible dosages of 2.520 mg/day, mean 11.1 mg/day)
(Kryzhanovskaya et al. 2005), and risperidone (N= 158, using flexible
dosage range of 0.53 mg/day, mean 2.6 mg/day, and 46 mg/day,
mean 5.3 mg/day) (Haas et al. 2007). In one additional, currently unpublished study, risperidone (N= 279) was dosed at 16 mg/day (mean
4.0 mg/day) and compared to a pseudo-placebo with a dosage range of
0.10.6 mg/day (mean 0.5 mg/day) in pediatric patients with schizophrenia (Janssen Pharmaceutica, unpublished data). In all four trials,
compared with patients taking the placebo or the pseudo-placebo dose
of risperidone, the patients taking the SGA had superior primary outcome, consisting of the total mean score change on the PANSS or BPRS
from baseline to endpoint. In an older, 4-week, placebo-controlled trial
(Pool et al. 1976), both haloperidol (n= 25, mean 9.8 mg/day) and loxapine (n = 26, mean 87.5 mg/day) were associated with significantly
greater reductions in BPRS scores compared with placebo, but there
were no differences between the two active medication groups.
The remaining antipsychotic studies in pediatric schizophrenia were
all active-controlled trials, comparing different antipsychotics and lacking a placebo arm, thereby precluding a conclusive assessment of the efficacy of the tested antipsychotics. In the three studies not involving
clozapine that lasted between 4 and 8 weeks, no statistically significant
differences in efficacy were found between the tested antipsychotics,
including comparisons of thiothixene (mean 16.2 mg/day) and
thioridazine (mean 178 mg/day) (N= 21; Realmuto et al. 1984); haloperidol (mean 5.0 mg/day), olanzapine (mean 12.3 mg/day), and risperidone (mean 4.0 mg/day) (N= 50; Sikich et al. 2004); and molindone


Medical Illness and Schizophrenia

(range 10140 mg/day, target 65 mg/day), olanzapine (range 2.520

mg/day, target 12.5 mg/day), and risperidone (range 0.56 mg/day,
target 3 mg/day) (N =119; Sikich et al. 2008). On the other hand, in relatively small, active controlled trials, clozapine at mean dosages between 176 mg/day and 403 mg/day was superior on several efficacy
measures compared with haloperidol (N=21, mean 16 mg/day; Kumra
et al. 1996), olanzapine (N=13, mean 18.1 mg/day; Shaw et al. 2006), and
high-dose olanzapine (up to 30 mg/day) (N=39, mean 26.2 mg/day;
Kumra et al. 2008a).
Taken together, findings from these studies show that antipsychotics
are superior to placebo in youth with schizophrenia, as in adults, and
that the differences between antipsychotics other than clozapine may
be relatively small. Therefore, differences in adverse effects, and especially adverse effects that can impair health outcomes and longevity,
are to be taken seriously when prescribing this class of medications to
pediatric patients, who often suffer from a severe form of the disorder
and who are likely to require antipsychotic treatment for long periods
of time.

Adverse Effects Associated With

Antipsychotics in Children and
Children and adolescents seem to be more sensitive than adults to most
antipsychotic adverse effects, including sedation, extrapyramidal side
effects (except akathisia, which occurs at similar rates), withdrawal dyskinesia, prolactin elevation (especially in postpubertal patients), weight
gain, and related metabolic abnormalities (Correll et al. 2006). On the
other hand, adverse effects that require a longer time to develop physiologically (e.g., diabetes mellitus) and/or that are related to medication dosage and lifetime exposure (e.g., tardive dyskinesia) seem to
emerge less frequently in pediatric cohorts, at least during follow-up
durations that are relatively short (Correll et al. 2006). Actually, these
adverse events may not occur at a reduced frequency in youths, who
have a relatively large reserve of pancreatic beta cells and striatal
dopamine neurons, but may rather be delayed and show up earlier in
adulthood, depending on how early antipsychotic treatment was initiated during childhood and adolescence. Clearly, longer-term follow-up
studies of youths in whom antipsychotics are initiated are required to
clarify this issue further.

Managing Health Outcomes in Children and Adolescents


Weight Gain and Metabolic Adverse Effects

Although pediatric data are limited, children and adolescents with psychiatric disorders seem to be at increased risk for being overweight or
obese (Patel et al. 2007), especially when exposed to antipsychotics for
longer periods of time (Laita et al. 2007). Age-inappropriate weight gain
is of particular concern in children and adolescents because weight gain
is associated with health and psychosocial issues at a time in human development when self-esteem and body image are being formed and
consolidated. Weight gain can be associated with the development of
eating disorders and depression, and the medical consequences of
weight gain may adversely affect pulmonary, gastrointestinal, renal,
musculoskeletal, cardiovascular, and endocrine systems in a manner
that is both enduring and detrimental (Ebbeling et al. 2002). These medical comorbidities and complications can impair patients quality of life
and shorten life expectancy.
Antipsychotic-Related Weight Gain. Reasons for weight gain and obesity in patients with psychiatric illness are multifactorial and complex, including effects of the underlying psychiatric illness, consequences of
lifestyle behaviors (particularly sedentary lifestyle and unhealthy diet),
and direct and indirect effects of psychotropic treatment. A recent review
of data suggested that the weight gain potential of SGAs for pediatric patients follows roughly the same rank order as that for adults (American Diabetes Association et al. 2004) but that the magnitude is greater (Correll
and Carlson 2006). Exceptions may be a greater relative weight gain propensity of risperidone in youth (Safer 2004), a greater likelihood of aripiprazole and ziprasidone to not be weight neutral in subgroups of
pediatric patients, and possibly greater olanzapine-induced weight gain
compared with clozapine in treatment-refractory youth, although results
differ depending on the studies and patient groups (Correll 2008b; Correll
and Carlson 2006). Research studies have not yet resolved whether the
greater antipsychotic-related weight gain in pediatric patients than in
adults is due to physiological differences between growing children and
adults that are related to mechanisms regulating food intake and energy
homeostasis, whether the greater weight gain is due to less lifetime antipsychotic exposure, or whether it is due to a combination of both of these
factors. The latter is suggested by findings of substantial amounts of
weight gain in adults who are antipsychotic naive or in a first episode of
schizophrenia (Alvarez-Jimnez et al. 2008).
In an 8-week, randomized study of 25 patients ages 716 years, clozapine and olanzapine resulted in similar weight gain (3.8 6.0 kg and


Medical Illness and Schizophrenia

3.6 4.0 kg, respectively) (Shaw et al. 2006). In an earlier, randomized

trial of patients ages 1018 years with early-onset schizophrenia, clozapine and haloperidol produced similar and quite modest weight gain
(0.9 6.5 kg and 0.9 2.9 kg, respectively) (Kumra et al. 1996). On the
other hand, in a naturalistic study, olanzapine was associated with
greater weight gain (4.6 1.9 kg) than clozapine (2.5 2.9 kg), which had
comparable weight gain to that of risperidone (2.8 1.3 kg) (Fleischhaker et al. 2006). However, because clozapine was used in treatmentrefractory patients, weight gain in clozapine-treated children could
have been attenuated by prior antipsychotic exposure and resultant
weight gain.
Similar to the study by Fleischhaker et al. (2006), Sikich et al. (2004)
found a higher weight gain in youths ages 517 years with psychotic
disorders who were randomly assigned to receive olanzapine for
8 weeks (7.14.1 kg) than in youths randomized to either risperidone
(4.9 3.6 kg) or haloperidol (3.5 3.7 kg). After only 8 weeks of treatment, all treatment groups had mean observed weight gain that was severe and disproportionate to that expected from normal growth. In a
prior open-label study, Ratzoni et al. (2002) reported on weight change
in adolescents with schizophrenia treated for 12 weeks using the same
three antipsychotics. Both statistically and clinically significant weight
gain was observed with olanzapine (7.2 6.3 kg) and risperidone
(3.9 4.8 kg), but the mean weight change with haloperidol (1.1 3.3 kg)
was not significant. Finding weight gain of 7% from baseline to the
end of 12 weeks in 90.3% of patients taking olanzapine, 42.9% taking
risperidone, and 12.5% taking haloperidol, the authors concluded that
the weight gain in their adolescent sample was more extreme than that
observed in adult studies. Similarly, high weight gain has been reported in preschool-age children (ages 46 years); during only 8 weeks
of treatment, children treated with olanzapine and risperidone had a
mean increase in weight of 12.9% 7.1% and 10.1% 6.1%, respectively
(Biederman et al. 2005).
In a 6-month naturalistic, unmatched study of 66 patients with fewer
than 30 days of antipsychotic exposure and treated with olanzapine
(n =20), risperidone (n =22), or quetiapine (n =24), age- and sex-adjusted
weight gain, measured as body mass index (BMI) z-scores, was significant in the olanzapine (1.1 0.82) and risperidone (0.48 0.73) groups,
but not in the quetiapine group (0.27 0.86) (Fraguas et al. 2008). In this
small sample, a transition to an adverse health outcomethat is, >85th
BMI percentile plus at least one adverse health outcome related to
blood pressure or lipid or glucose metabolism (Correll and Carlson
2006)occurred significantly compared to baseline in the olanzapine

Managing Health Outcomes in Children and Adolescents


group (15.0%60.0%), but not in the risperidone (22.7%36.4%) or quetiapine (15.5%20.8%) groups.
Due to the confounding effects of nonrandom assignment and small
sample sizes in the studies mentioned previously, larger, placebocontrolled trials are more informative. Because none of the four 6-week,
randomized, placebo-controlled studies of SGAs in adolescents with
pediatric schizophrenia have been published and only three have been
presented at scientific meetings, at this point, data on absolute weight
gain are available for only two of these trials. In a study comparing
fixed-dosage aripiprazole 10 mg (n =102) and 30 mg (n =100) with placebo (n=100), a slight but significant difference (P<0.05) in weight gain
was reported in adolescents taking placebo (0.8 kg) versus aripiprazole 10 mg (0 kg) and 30 mg (0.2 kg) (Findling et al. 2007), although the
weight change with aripiprazole was minimal in this study and comparable to what has been reported in adults. On the other hand, treatment with olanzapine (n= 72) at dosages between 2.5 and 20 mg (mean
11.1 mg) was associated with a 4.3-kg weight gain, which was significantly greater (P<0.0001) than the 0.1-kg weight gain in patients on placebo (n =35) (Kryzhanovskaya et al. 2005).
Because mean weight change can easily obscure substantial weight
gain in a subset of patients due to weight loss in another subgroup with
different past antipsychotic exposure and other characteristics, an important consideration is the proportion of patients who gained at least
7% of weight in the recent placebo-controlled trials with SGAs. Figure
131 summarizes the data from the three double-blind, placebocontrolled studies of atypical antipsychotics in adolescents with schizophrenia for which data are available on the proportion of patients who
experienced at least 7% of weight gain during 6 weeks of treatment
(Correll 2008b). Results suggest that similar to adult studies, the greatest weight gain occurred in patients taking olanzapine (45.8%) versus
placebo (1.7%) (Kryzhanovskaya et al. 2005). Risperidone was associated with intermediate risk, with at least 7% weight gain occurring in
15% of the 1- to 3-mg group and 16% of the 4- to 6-mg group, compared
with 2% of the placebo group (Haas et al. 2007). Finally, aripiprazole
was associated with the lowest risk, with at least 7% weight gain in 4%
of patients on 10 mg and 5.2% of patients on 30 mg of aripiprazole, compared with 1% of patients on placebo (Findling et al. 2007). The resulting numbers needed to harm (NNH) in patients with adolescent
schizophrenia were 3.2 for olanzapine, 7.17.7 for risperidone, and
24.433.3 for aripiprazole (Correll 2008b). Data from the completed
treatment of early-onset schizophrenia spectrum disorders study
(Sikich et al. 2008), which compared risperidone, olanzapine, and


Medical Illness and Schizophrenia

molindone, will add to this picture, but more randomized head-to-head

studies of antipsychotics in youths are needed to directly compare riskbenefit ratios.
Of note, currently available data suggest that combined treatment
with an SGA plus a stimulant does not seem to substantially attenuate
SGA-induced weight gain (Aman et al. 2004; Calarge et al., in press). On
the other hand, at least with olanzapine, risperidone, and possibly quetiapine, combined treatment with an SGA plus a mood stabilizer seems
to be associated with greater weight gain than monotherapy with a
mood stabilizer or even combination treatment with lithium plus valproic acid, as suggested by trials in youths with bipolar disorder (Correll 2007b).
Adverse Metabolic Effects. With the widely reported age-inappropriate
weight gain in children and adolescents treated with antipsychotics, a
worsening of metabolic indices, such as triglycerides, cholesterol, and
insulin resistance, is expected to occur. However, methodologically
sound, conclusive, and long-term prospective data regarding metabolic
effects of antipsychotics in pediatric patients are largely lacking. In a
cross-sectional study with metabolic data from 80 to 95 pediatric patients, total cholesterol levels (P< 0.001) and low-density lipoprotein
(LDL) cholesterol levels (P=0.018) were significantly higher in youths
receiving antipsychotics for 12 months or longer than in similar patients
given antipsychotics for less than 1 month (Laita et al. 2007). Despite
large differences in BMI (i.e., 24.1 in youth with longer-term exposure
vs. 20.4 in youth with shorter-term exposure), triglyceride levels (which
correlate with insulin resistance) were not statistically different (94.7
48.3 mg/dL vs. 82.7 48.9 mg/dL). Although this lack of a significant
difference for triglycerides is somewhat surprising, these results could
have been due to any of the following: the sample size was relatively
small, groups were not randomized or closely matched, weight changes
may have occurred largely within the normal weight range, youths
might be able to draw upon intact and effective compensatory mechanisms, and some of the assessments were from nonfasting patients. Although the link between antipsychotic treatment and adverse metabolic
consequences such as dyslipidemia, hyperglycemia, diabetes, and metabolic syndrome has been established in adults (American Diabetes Association et al. 2004), the few published retrospective and prospective
pediatric studies (Biederman et al. 2005, 2007; Malone et al. 2007; Martin
and LEcuyer 2002; Sikich et al. 2004) have produced almost exclusively
negative results, in that antipsychotic exposure and weight gain were
associated with no significant metabolic abnormalities. However, the





n =100

n =54

n =35

Aripiprazole 10 mg
n =102

Risperidone 13 mg
n =55

Olanzapine 220 mg
n =72

Aripiprazole 30 mg
n =100

Risperidone 46 mg
n =51

FIGURE 131. Proportion of adolescents (ages 1317 years) with schizophrenia gaining >7% of baseline weight in 6-week, doubleblind, placebo-controlled trials with atypical antipsychotics.

Managing Health Outcomes in Children and Adolescents

Percentage of patients




Medical Illness and Schizophrenia

interpretation of these results is limited by small sample sizes, varying

treatment histories, and mostly the inclusion of random glucose assessments.
In a pooled analysis of 24 medication trials of antipsychotic-treated
youth with bipolar disorder, nonfasting glucose and lipid changes were
nonsignificant in the only two atypical antipsychotic trials with available data (N =61) (Correll 2007b). Moreover, as briefly reported (Correll
and Carlson 2006), interim results from a large-scale, naturalistic cohort
study presented at a scientific meeting suggested the emergence of relevant rates of new-onset dyslipidemia and increases in insulin resistance in several of the SGA groups after only 3 months of exposure in
the subgroup of antipsychotic-naive patients. This preliminary report is
consistent with another recent study of 66 pediatric patients (67% male,
mean age 15.2 years) with psychotic disorders, treated naturalistically
with either olanzapine (n= 20), risperidone (n =22), or quetiapine (n =24)
(Fraguas et al. 2008). Importantly, as in the preliminary report by Correll et al. (2006), patients had minimal prior antipsychotic useall had
fewer than 30 days lifetime exposure, and 38% were antipsychotic naive. After 6 months of antipsychotic exposure in this nonrandomized
and unmatched sample, total cholesterol levels increased significantly
in the olanzapine group (P=0.045) and the quetiapine group (P= 0.016),
but not in the risperidone group. Although weight gain was significant
in all three groups, triglyceride, LDL, high-density lipoprotein (HDL),
and glucose levels did not increase significantly compared with baseline levels over the 6-month period in this small group of patients.
As recently summarized (Correll 2008b), short-term, placebocontrolled trials with aripiprazole, risperidone, and quetiapine in
youths with schizophrenia or bipolar disorder found no significant
mean changes in lipid or glucose parameters. The same was true for a
recently presented placebo-controlled trial with ziprasidone in pediatric patients with bipolar disorder, who experienced minimal weight
gain (DelBello et al. 2008). Conversely, compared with placebo, olanzapine resulted in significant increases in triglycerides in adolescents with
schizophrenia (P= 0.029), as well as significant increases in blood glucose (P= 0.002), total cholesterol (P= 0.010), and uric acid (P =0.026) in
pediatric patients with bipolar disorder (Tohen et al. 2007). Moreover,
in patients with pediatric bipolar disorder, new-onset abnormal metabolic values that occurred at any time during the study were significantly more frequent with olanzapine than with placebo; these values
included abnormally elevated total cholesterol (19.1% vs. 2.1%,
P = 0.004), low HDL (10.9% vs. 0.0%, P = 0.016), and hypertriglyceridemia (49.1% vs. 14.8%, P=0.003) (Tohen et al. 2007).

Managing Health Outcomes in Children and Adolescents


Although individual case reports of diabetes in pediatric patients taking antipsychotics have been published (for specific reports, see Correll
and Carlson 2006), methodologically rigorous, prospective studies of
metabolic syndrome and diabetes in youths exposed to antipsychotics,
especially for longer periods of time, are almost absent. Despite the absence of such long-term studies, long-term prospective data in nonpsychiatric pediatric populations followed into adulthood indicate that ageinappropriate weight gain has particularly deleterious effects when occurring early in life. In follow-up studies, obesity, metabolic abnormalities, and weight gain during childhood have been shown to strongly
predict obesity, metabolic syndrome, hypertension, cardiovascular
morbidity, sleep apnea, osteoarthritis, and malignancy risk in adulthood (Baker et al. 2007; Dietz and Robinson 2005; Freedman et al. 2004;
Must et al. 1992; Srinivasan et al. 2002). Clearly, these data strengthen
the concern about the magnitude of the antipsychotic-related weight
gain that is observed in short-term studies of pediatric populations.

Prolactin-Related Side Effects

As previously summarized (Correll and Carlson 2006), FGAs and several SGAs are associated with clinically relevant prolactin elevations in
children and adolescents. Hyperprolactinemia can result in sexual and
reproductive system side effects, such as amenorrhea or oligomenorrhea, erectile dysfunction, decreased libido, hirsutism, and breast
symptoms, such as enlargement, engorgement, pain, or galactorrhea.
However, in open-label (Findling et al. 2003; Masi et al. 2003; Saito et al.
2004) and randomized (Aman et al. 2002) studies with youth, serum
prolactin levels are not tightly correlated with these side effects, and not
all patients with hyperprolactinemia develop these signs and symptoms (Masi et al. 2003). Like adults, many pediatric patients continue
to have normal gonadal function and no overt reproductive-system
side effects, despite moderately elevated serum prolactin (Findling
et al. 2003). Available data also suggest that hyperprolactinemia is
dosage dependent, may normalize over time in some patients, and
resolves after antipsychotic discontinuation. Similar to the pattern
seen in adults, the relative potency of antipsychotic drugs in inducing
hyperprolactinemia in pediatric patients is roughly the following:
paliperidone risperidone > haloperidol > olanzapine > ziprasidone
> quetiapine> clozapine >aripiprazole.
Due to prepubertal status and less sexual activity and familiarity
with their developing sexuality, many pediatric patients exposed to
antipsychotics may not experience, notice, or report sexual or reproductive system dysfunction, making it more difficult to determine if hyper-


Medical Illness and Schizophrenia

prolactinemia is present through history taking or even physical

examination (Correll and Carlson 2006). Studies suggesting that levels
of hyperprolactinemia that cause hypogonadism (i.e., that suppress
gonadotropin-releasing hormone and, thus, sex hormone levels) are
associated with osteoporosis and increased risk of bone fractures are
especially concerning, because adolescence is the prime time of bone
mineralization development (Byerly et al. 2007). Other potential but unconfirmed risks from childhood hyperprolactinemia might include a
negative effect on pubertal development (Correll and Carlson 2006) and
a risk for breast cancer or pituitary tumors (Byerly et al. 2007).
In a review of antipsychotic effects on prolactin levels in youths with
schizophrenia spectrum disorders (Kumra et al. 2008b), risperidone and
haloperidol were noted to have the greatest effect on prolactin levels; quetiapine, clozapine, and aripiprazole were noted to have the least effect on
prolactin levels; and olanzapine and ziprasidone appeared to have intermediate effects. In a 3-week study of 161 children and adolescents, olanzapine treatment was associated with a greater baseline-to-endpoint increase
in prolactin levels compared with placebo, as well as with surprisingly
high incidence rates of hyperprolactinemia, particularly among boys (in
boys, 62.5% with olanzapine vs. 5% with placebo, P<0.001; in girls, 25.7%
with olanzapine vs. 0% with placebo, P=0.007) (Tohen et al. 2007).
A pooled analysis of several pediatric risperidone studies in children
ages 515 years treated with mean dosages of 0.020.06 mg/kg/day
found that the biggest increase in prolactin levels occurred in the first 1
2 months (N=550) (Findling et al. 2003). However, results of this study
have to be interpreted within the limitations that they are based on a
sample with an inherently relatively low hyperprolactinemia risk, consisting of prepubertal individuals, predominantly boys, treated with low
dosages of risperidone, and cotreatment with stimulants was allowed.
One larger study (N =222) examined the effects of risperidone on
height and sexual maturation (Dunbar et al. 2004). Boys ages 1015 years
and girls ages 915 years were evaluated for sexual maturation by Tanner staging and followed for 1112 months. Encouragingly, this study
found no correlation between prolactin levels and either height or sexual
development; however, firm conclusions from this single study are precluded by sample heterogeneity in age and pubertal stage, combined
with follow-up of only 1 year.

QTc Prolongation
As discussed extensively in Chapter 7, The Spectrum of Cardiovascular Disease in Patients With Schizophrenia, antipsychotics can differ-

Managing Health Outcomes in Children and Adolescents


entially prolong the heart ratecorrected QT interval (QTc) of the

electrocardiogram. QTc prolongation may be associated with torsade
de pointes, a potentially fatal arrhythmia (Blair et al. 2004). Concerns in
pediatric patients include the notion that the developing cardiac conduction system may be especially vulnerable to these effects. In adults,
QTc prolongation is usually minimal with antipsychotics compared to
placebo, except with certain medications (e.g., thioridazine), but among
the SGAs, ziprasidone has been associated with the greatest QTc prolongation compared to baseline (Glassman and Bigger 2001), although
without known incidence of torsade de pointes.
Cardiac conduction effects of ziprasidone have been studied in children and adolescents, and QTc prolongation to >430 milliseconds has
been described in three of 20 pediatric patients treated prospectively
with ziprasidone (mean QTc prolongation of 28 26 milliseconds,
P< 0.01) (Blair et al. 2005). Although no relationship to ziprasidone dosage was found in this study, dosages were very low (mean 3013 mg/
day, range 3060 mg/day). In a small study in only 12 youths, a statistically significant increase in QTc (14.721.0 milliseconds, P= 0.04) was
reported at a mean ziprasidone dosage of 98.3 (range 40160) mg/day
(Malone et al. 2007). On the other hand, QTc changes were reported to
have been nonsignificant in studies of 12 (McDougle et al. 2002), 16
(Sallee et al. 2000), and 21 patients (Biederman et al. 2007) at ziprasidone
dosages of 57.3 (range 20120) mg/day, 28.2 (range 540) mg/day, and
59.2 (range 20120) mg/day, respectively. Moreover, no patient in any
of these studies reported cardiac side effects, such as dizziness, palpitations, or syncope. Thus, the clinical relevance of this degree of QTc prolongation, which did not reach the generally accepted pathological
threshold of >500 milliseconds or an increase in QTc over baseline of
>60 milliseconds (Glassman and Bigger 2001), is unclear, and theoretical concerns about QTc prolongations with ziprasidone need to be
weighed against the more certain benefits regarding the relative risk for
weight gain and metabolic abnormalities. (For more discussion, see
Chapter 7, The Spectrum of Cardiovascular Disease in Patients With

Monitoring and Management of

Adverse Effects
Pediatric patients taking antipsychotics must be monitored carefully and
proactively for adverse effects, particularly medical adverse effects that can
shorten life expectancy, decrease quality of life and functioning, or nega-


Medical Illness and Schizophrenia

tively affect treatment alliance and adherence. If results of monitoring

show clinically relevant adverse events, actions should be taken to minimize their impact on patient health and other outcomes. Importantly, these
adverse effects may have to be identified and managed by mental health
prescribers because young patients are most often seen by pediatricians
only once per year for routine checkups. Because physical health is very
important and also related to psychiatric treatments and outcomes, the
psychiatrist needs to take the initiative to monitor medical effects of psychiatric medications, initiate referrals for a consultation, and actively comanage the patients physical health in addition to the patients mental health.

Monitoring Strategies
Several assessments of adverse effects should be made at baseline and
at regular intervals to monitor the impact of antipsychotic treatments
on pediatric patients (see Table 131). Importantly, as children and adolescents undergo normal developmental changes, the adverse effect
assessment and monitoring has to take into consideration developmental norms and thresholds that can differ substantially from those for
adults (see Table 132).
At baseline and annually, personal and family histories of metabolic
and endocrine complications should be assessed. At each visit, the clinician should inquire about the status of health lifestyle behaviors related to
diet, activity, sleep, and substance use, unless the patients weight and
metabolic status are healthy and stable. The elicited information should be
compared with recommended behaviors in the general pediatric population (American Medical Association 2007) or those that were adapted for
youth with psychiatric illness (Correll and Carlson 2006) (see Table 133).

Height, Weight, and Body Composition

Being dynamic parameters in youths, height and weight should be measured and recorded at each visit. Height measurement can be difficult and
requires adequate toolsideally a wall-mounted stadiometerand
trained personnel who measure height several times and make sure patients stand erect, place their heels against the wall, and keep their head
straight. The clinical measures used most often to monitor body weight include absolute weight change (kilograms, pounds), percent weight
change (weight change/baseline weigh t), and ch an ge in BMI
(BMI=[weight in kilograms]/[height in meters2] or [weight in pounds]/
[height in inches2]703). Although all measures are easily obtained and
valid in adults, these measures are only useful in pediatric patients who
are followed for periods of 3 months or less because they do not take into

Managing Health Outcomes in Children and Adolescents


account age-appropriate development. Therefore, BMI values need to be

adjusted for age- and sex-dependent growth velocities, using freely available growth charts ( or calculators
( bmiz2.html). The adjusted
BMI values are the BMI z-score and the BMI percentile, where a z-score
(i.e., standard deviations) of zero and the 50th BMI percentile represent
the population mean, and where continuation on the same BMI z-score or
percentile over time indicates stable age- and sex-adjusted body weight in
relationship to height. Because BMI z-scores are not capped at the zero
or 99th percentile, z-scores are preferably used to track weight change.
BMI percentiles, on the other hand, are useful to determine the weight category that a pediatric patients belongs to, where a BMI percentile of <5 is
underweight, 584.9 is healthy weight, 8594.9 is overweight, and 95 is
obese (see Table 132). Although waist circumference is preferred over
BMI in adults as a metabolic syndrome criterion and is highly predictive
of metabolic syndrome (Straker et al. 2005), the American Medical Association Expert Committee did not recommend waist circumference measures in pediatric patients because cutoffs are less well established and
assessments are liable to measurement error (Correll 2008c).
Despite the importance of age-inappropriate weight gain in pediatric
patients, no consensus exists in the general pediatric literature regarding the cutoff for clinically meaningful weight change during development. Rather, a BMI in the 85th percentile is the accepted lower
intervention threshold in youths (American Medical Association 2007)
(see Table 132). However, in psychiatric care, where the underlying
disorder together with adverse treatment effects can lead to rapid and
often significant weight gain, clinicians require guidance at what point
to consider changing therapy or using adjunctive treatments to address
clinically relevant weight gain. In patients with psychiatric illness, clinically significant weight gain or abnormal weight status that requires
a reconsideration of the current treatment plan has recently been operationalized as 1) >5% weight gain during 3 months, or any of the following three conditions at any time during treatment: 2) an increase 0.5
in BMI z-score; 3) BMI percentile 8594.9 plus one adverse health consequence (hyperglycemia, dyslipidemia, hyperinsulinemia, hypertension, orthopedic, gall bladder, or sleep disorder); or 4) BMI 95th
percentile or abdominal obesity (> 90th percentile) (Correll et al. 2006).

Fasting Blood Glucose and Lipids

Although no generally accepted pediatric guidelines regarding fasting
glucose and lipid monitoring have been published, one proposed
schedule requires monitoring at baseline, after 3 months of treatment,


TABLE 131. Suggested monitoring strategies in children and adolescents treated with antipsychotic agents a


During titration
and at target

At 3 months

Every 3

Every 6


Medical Illness and Schizophrenia

Height, weight
(calculate BMI
percentile, BMI
Fasting glucose
and lipids
Blood pressure and
Personal and
family medical



TABLE 131. Suggested monitoring strategies in children and adolescents treated with antipsychotic agents a (continued)

Electrolytes, full
blood count, renal

Prolactin d


During titration
and at target

At 3 months

Every 3

Every 6

If on
or clozapine
Only if

If symptomatic,
mandatory CBC
for clozapine

If on
ziprasidone or
if symptomatic
on clozapine
Only if
Only if
symptomatic symptomatic

Only if

Only if

Only if


Note. BMI=body mass index; CBC = complete blood cell count.

More frequent assessments of abnormalities occur if patient is at very high risk for specific adverse events by personal or family history.
Lifestyle behaviors include diet, exercise, smoking, substance use, and sleep hygiene.
Medical history includes components of the metabolic syndrome (obesity, arterial hypertension, diabetes, dyslipidemia); past medical
history for coronary heart disease or coronary heart disease equivalent disorders (i.e., diabetes mellitus, peripheral arterial disease, abdominal aortic aneurysm, and symptomatic carotid artery disease); history of premature coronary heart disease in first-degree relatives
(males <55 years and females <65 years); and past efficacy and adverse effect experiences in patients and/or family members.
In case of abnormal sexual symptoms or signs, prolactin is drawn fasting in the A.M. and approximately 12 hours after the last antipsychotic dose.

Managing Health Outcomes in Children and Adolescents



Medical Illness and Schizophrenia


TABLE 132.

Clinically relevant thresholds for body weight and

metabolic parameters in children and adolescents


Children and adolescents

Body weight category


*BMI<5th percentile for sex and


Normal weight

*BMI 5th<85th percentile for sex

and agea

Overweight (previously at risk

for overweight in pediatric

*BMI 85th<95th percentile for sex

and agea

Obese (previously overweight

in pediatric patients)

*BMI 95th percentile for sex and


Blood lipid abnormalities

Total cholesterol

*170 mg/dL

LDL cholesterol

*130 mg/dL

HDL cholesterol

*<40 mg/dL in males and females


*110 mg/dL

Blood glucose abnormalities

Fasting hyperglycemia

100125 mg/dL

2-hour oral glucose tolerance

test result

140199 mg/dL

Fasting diabetes (needs to be


126 mg/dL

2-hour postglucose load diabetes

200 mg/dL

Insulin abnormalities and

insulin resistance
Fasting hyperinsulinemia

*>20 mol/L

Homeostatic model assessmentb


Triglycerides:HDL cholesterol ratio *? >3.5

Managing Health Outcomes in Children and Adolescents

TABLE 132.


Clinically relevant thresholds for body weight and

metabolic parameters in children and adolescents (continued)


Children and adolescents

Metabolic syndrome

three out of five criteria required

Abdominal obesity criterion

*Waist circumference 90th

percentile, or BMI 95th
percentile for sex and ag