Review article
art ic l e i nf o a b s t r a c t
Article history: Objective: Over the years studies have shown conflicting results about the risk of suicide in psychotic
Received 9 May 2015 depression (MD-psych). To understand this association, we undertook a comprehensive review of the
Received in revised form literature to ascertain whether individuals with MD-psych have higher rates of completed suicides,
12 February 2016
suicide attempts or suicidal ideation compared to those with non-psychotic depression (MD-nonpsych).
Accepted 9 March 2016
Methods: We searched Pubmed, PsycINFO and Ovid in English language, from 1946-October 2015. Stu-
Available online 11 March 2016
dies were included if suicidal ideation, attempts or completed suicides were assessed.
Keywords: Results: During the acute episode of depression, patients with MD-psych have higher rates of suicide,
Major depression suicide attempts, and suicidal ideation than patients with MD-nonpsych, especially when the patient is
Psychotic depression
hospitalized on an inpatient psychiatric unit. Studies done after the acute episode has resolved are less
Suicide
likely to show this difference, likely due to patients having received treatment.
Limitations: Diagnostic interviews were not conducted in all studies. Many studies did not report
whether psychotic symptoms in MD-psych patients were mood-congruent or mood-incongruent; hence
it is unclear whether the type of delusion increases suicide risk. Studies did not describe whether MD-
psych patients experienced command hallucinations encouraging them to engage in suicidal behavior.
Only 24 studies met inclusion criteria; several of them had small sample size and a quality score of zero,
hence impacting validity.
Conclusions: This review indicates that the seemingly conflicting data in suicide risk between MD-psych
and MD-nonpsych in previous studies appears to be related to whether one looks at differences during
the acute episode or over the long-term.
& 2016 Elsevier B.V. All rights reserved.
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.1. Aims of the study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1. Study population and study design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.1.1. Study selection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.1. Completed suicide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.1.1. Studies on inpatient psychiatric units. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.1.2. Population-based studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.1.3. Follow-up studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.2. Suicide attempts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.2.1. Studies on inpatient psychiatric units. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.2.2. Studies including both inpatient and outpatient referrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.2.3. Outpatient studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
n
Corresponding author.
E-mail addresses: izalpuri@stanford.edu (I. Zalpuri), anthony.rothschild@umassmemorial.org (A.J. Rothschild).
http://dx.doi.org/10.1016/j.jad.2016.03.035
0165-0327/& 2016 Elsevier B.V. All rights reserved.
24 I. Zalpuri, A.J. Rothschild / Journal of Affective Disorders 198 (2016) 23–31
1. Introduction 2. Methods
Major depression with psychotic features (MD-psych) is de- We searched Pubmed, PsycINFO/Ovid in English language only,
fined as patients meeting criteria for major depression and then from 1946-October 2015, using keywords suicide, major depres-
having features of psychosis, either hallucinations or delusions, sion and psychotic depression. All studies were included in this
commonly manifesting as nihilistic delusions, overly self-critical or report if suicide risk (ideation, attempts) or completed suicides
guilty beliefs. It remains a sub-classification of major depression, were assessed in subjects with MD-psych. Studies were cross
although some authors have argued that a separate designation sectional, cohort, case control, retrospective and randomized
from major depression is supported by scientific studies (Schatz- controlled trials.
berg and Rothschild, 1992). Patients with MD-psych have higher
scores on the Hamilton Rating Scale for Depression (Glassman and 2.1. Study population and study design
Roose, 1981, Nelson et al., 1984, Lykouras et al., 1986) higher rates
of mortality from medical causes (Vythilingam et al., 2003) and We considered studies where subjects met DSM III, DSM IV or
more frequent relapses (Lykouras et al., 1986, Robinson and Spiker, ICD-8, ICD-9, ICD-10 criteria for MD-psych and MD-nonpsych. All
1985, Aronson et al., 1987, Coryell et al., 1996) when compared to of these studies had participants of both genders, older than 18
patients with non-psychotic depression (MD-nonpsych), although years of age, except Quinlan et al. (1997) where subjects were
not all studies are in agreement (Coryell et al., 1987, Lykouras et al., adolescents, 12–18 years of age, Suominen et al. (2009) where
1994). Clinicians frequently miss the diagnosis of MD-psych individuals were 16 years or older, Park et al. (2014) where pa-
(Rothschild et al., 2008), which can lead to sub-optimal treatment, tients were over 7 years old and Kelleher et al. (2014) where
participants were 12–16 years of age. Exclusion criteria included
potentially increasing the risk of suicide. It is well known that
substance use and patients meeting DSM III, DSM IV or ICD-8, ICD-
patients with depression have a higher rate of suicide relative to
9 or ICD-10 criteria for schizophrenia, schizoaffective disorder or
general population (Hagnell et al., 1981, Henriksson et al., 1993).
bipolar disorder. Hospital-based, follow up and population-based
However, over the years, studies have shown mixed results about
studies were included. Hospital-based studies comprised of both
whether the risk of suicide in patients with MD-psych is different
inpatient and outpatient subjects. Case reports, book chapters and
than patients with MD-nonpsych.
letters to editors were excluded.
698 articles identiied 598 articles were excluded The search identified 698 articles. Of these 598 were excluded
through database searching based on title and abstract based on title and abstract. 100 articles were assessed for elig-
ibility. Of these, 76 were excluded based on exclusion criteria and
24 articles were included in the review. 22 were hospital based
(inpatient, outpatient and follow ups) and 2 were population
based. We present the results of our review below in the following
100 full-text articles were 76 articles excluded based sections: Completed Suicide, Suicide Attempts, and Suicidal
assessed for eligibility on exclusion criteria Ideation.
Study Sample Description Study population Findings Risk of completed suicide Score
in psychotic vs. non-psy-
chotic depression
Coryell et al. Inpatient 40 year follow up study using chart reviews to de- 122 patients with MD-psych and 8 suicides in MD-psych group, 8.4%of the de- No difference 0
(1982) scribe short and long term outcomes in patients 103 patients with MD-nonpsych ceased; 7 suicides in MD-nonpsych group, 10% of
hospitalized 5 years prior to ECT availability the deceased no difference between groups in
25
26
Table 2
Studies comparing risk of suicide attempts in MD-psych and MD-nonpsych.
Study Sample Description Study Population Findings Risk of suicide at- Score
tempts in psychotic
vs. non-psychotic
depression
Frangos Inpatient Retrospective chart review; the two groups were compared 145 with MD-psych and 119 with MD-psych is a severe variant of major depression Higher but not sta- 0
et al. to assess whether MDpsych represents a distinct subtype MD-nonpsych tistically significant
(1983) or a severe variant of major depression (no significant
difference)
Robinson Inpatient 1-year follow up study following discharge from hospital 52 patients met criteria for MD- Patients in MD-psych group had higher rates of major depression No difference 0
and Spi- psych; control group consisted of or delusions lasting longer than 9 months and of being in a major
ker (1985) 52 patients with MD-nonpsych depressive episode at the end of the follow up period
Miller and Inpatient Retrospective analysis of MDpsych patients admitted to 45 consecutive patients with MD- MD-psych patients with delusions of sinfulness, guilt, deserved 1.5 fold higher, not 0
Chabrier Payne Whitney Clinic with medically serious suicide psych discharged from the hospital punishment, or persecution were more likely to make more ser- statistically
(1987) attempts ious suicide attempts than those with delusions of bodily disease, significant
between MD-psych and MD-nonpsych groups. suicide risk in the two groups. There was no statistically significant
difference between the two groups. Both in this study and the
3.1.1. Studies on inpatient psychiatric units study by Black et al. (1988), psychiatric status of suicide victims at
One study included a retrospective analysis of all suicides at the the time of suicide was unknown.
New York State Psychiatric Institute over a twenty-five year period A study (Suominen et al., 2009) reported a 4-year follow up in
from 1955 to 1980 (Roose et al., 1983). Patients were diagnosed Finland where the study population comprised of individuals 16
retrospectively according to Research Diagnostic Criteria (Spitzer years or older with ICD-10 diagnosis of MD-psych, MD-nonpsych
et al., 1978). 39 patients had committed suicide during the study and other or unspecified (F32.8, F32.9, F33.8, F33.9). They found
period; 22 had affective disorder and 14 of those met criteria for that psychotic symptoms increased the risk of completed suicide
major depressive disorder. Of those 14, 10 patients had MD-psych. following a suicide attempt by 3-fold in individuals with MD-
Patients with unipolar MD-psych were 5 times more likely to psych (Hazard Ratio ¼3.32[Confidence Interval:1.95–5.67]). One
commit suicide than a patient with a MD-nonpsych (relative risk: limitation of this study was that diagnoses were not made by
5.3, p o0.05). Delusions were the most powerful predictors of standardized diagnostic interview schedules but were made by
suicide potential in patients with endogenous depression. 5 out of physicians who treated these patients on the unit. A registry based
10 patients with MD-psych committed suicide when they were Danish prospective cohort study (Leadholm et al., 2014) was
believed to be better by family and medical staff. The study raised conducted to determine suicide risk factors in depressed adults
an interesting question as to whether the psychotic process hospitalized at Danish hospitals between 1994 and 2010. This in-
transforms suicidal thoughts in depressed patients to suicidal acts. cluded 755 patients who committed suicide during follow up. Of
In 1987, a retrospective study of inpatients in the depression these 551 (2.1%) had MD-nonpsych and 280 (2.3%) were diagnosed
ward of Weissenau State Psychiatric Hospital was conducted with MD-psych by ICD-10 criteria. MD-psych was not found to be
(Wolfersdorf et al., 1987). During the period of their study, 6 pa- an independent risk factor for suicide. All subjects in this study
tients committed suicide. 4/46 (8.7%) patients with MD-psych were severely depressed, whereas previous studies had reported
committed suicide compared to 2/46 (4.3%) patients with MD- on a range of depression severity.
nonpsych. The authors reported no statistically significant differ-
ences in suicide rates; however, their sample size was very small. 3.2. Suicide attempts
On calculating odds ratio, they found that the relative risk that
MD-psych patients would die by suicide was 2.1 times greater than The majority of studies, which looked at attempted suicides,
MD-nonpsych patients. have shown a higher rate of suicide attempts in MD-psych com-
pared to MD-nonpsych. Table 2 summarizes pertinent studies
3.1.2. Population-based studies comparing the two groups.
In 1994, a psychological autopsy was conducted to study dif-
ferences between MD-psych and MD-nonpsych as part of the 3.2.1. Studies on inpatient psychiatric units
National Suicide Prevention Project in Finland (Isometsä et al., In one study (Frangos et al., 1983), authors conducted a retro-
1994). Authors retrospectively analyzed suicides committed in spective chart review of a total of 264 patients: 145 MD-psych and
Finland between April 1987-March 1988. Information was ob- 119 MD-nonpsych, discharged from 6 psychiatric inpatient units of
tained by interviewing relatives and health care providers of sui- the State Mental Hospital in Athens. They reported that suicide
cide victims and provisional diagnoses were assigned retro- attempts were more frequent in MD-psych group, 27.6% vs. 22.7%,
spectively using DSM-III-R criteria. The sample consisted of 24 but this finding was not statistically significant. Another study
MD-psych and 46 MD-nonpsych patients. The authors reported (Wolfersdorf et al., 1987) also reported that more suicide attempts
that the two groups did not differ in previous suicide attempts or were seen in previous case histories of MD-psych patients (37%) as
frequency of suicidal ideation or intent. However, they found that opposed to MD-nonpsych patients (20%, p ¼ 0.07). Over a 12-year
patients with MD-psych were more likely to use violent methods period, researchers conducted a retrospective chart review (Hori
of suicide (88% vs. 59%, p¼ 0.028) than patients with MD-non- et al., 1993) of hospitalized patients with MD-psych or MD-non-
psych. A higher proportion of MD-psych patients committed sui- psych at University of Tsukuba, Japan, using DSM-III criteria, over a
cide during hospital treatment (21% vs. 11%, p¼ NS), but it was not 12-year period. Any self-inflicted injury where the subject thought
found to be statistically significant. Only 21% of the MD-psych that there was a risk of dying was considered as a suicide attempt.
sample was inpatient at the time of suicide. 18/38 (47%) and 10/55 (18.2%) patients attempted suicide in the
MD-psych and MD-nonpsych group respectively (p o0.05). Pa-
3.1.3. Follow-up studies tients with MD-psych were more likely to use more violent
In 1988, a follow up study did not find any significant differ- methods of suicide than the MD-nonpsych group such as pene-
ences in rates of suicide between MD-psych and MD-nonpsych trating or cutting deep into their bodies (9.4%), jumping (9.4%) or
patients in a 0–14 year follow up study of patients admitted to self-immolation by fire (6.3%). Hanging was slightly more common
University of Iowa Psychiatric Hospital between January 1970 and in the MD-nonpsych group (36.8% vs. 31.2% in MD-psych group).
December 1981 (Black et al., 1988). They found that 10.9% of Another study (Lykouras et al., 2002) compared the two groups in
subjects with MD-psych committed suicide as compared to 11.5% an inpatient study in Greater Athens Area and did not find any
subjects with MD-psych. In a 1-year follow up study following significant difference in the rate of suicide attempts during pre-
discharge from an inpatient unit, a study (Robinson and Spiker, vious (27.5% for MD-psych vs. 26.5% for MD-nonpsych) and current
1985) matched 52 MD-psych to 52 MD-nonpsych patients and episodes (12.5% for both). Patients and an informant were inter-
found no significant differences between the two groups. How- viewed on previous suicide attempts. The small number of suicide
ever, this was not the primary focus of their study, their power was attempters in this study (11 in MD-psych and 17 in MD-nonpsych
low and the study period was short. They found 1 suicide in MD- group) could reduce the power of comparison and makes results
psych and no suicides in MD-nonpsych group. A 40 year follow up liable to type II error. They found that half of the subjects in each
study (Coryell et al., 1982) using chart reviews was conducted on group used violent methods (stabbing, shooting or hanging
inpatients in Iowa, where 10% (7/70) died by suicide in the MD- themselves, severing wrists, drowning), which differed from the
psych group compared to 8.4% (8/95) in the MD-nonpsych group, findings of Hori et al. (1993) where patients with MD-psych used
although again the primary focus of this study was not to assess more violent methods of suicide than patients with MD-nonpsych.
28
Table 3
Studies comparing risk of suicidal ideation in MD-psych and MD-nonpsych groups.
Nelson et al. Inpatient The study examined clinical variables, family his- 13 patients with MD-psych com- MD-psy group had more family history of depression and less of al- Increased, po 0.05 2
(1984) tory, cortisol secretion and 4-week tricyclic anti- pared to 12 MD-nonpsych patients coholism, poorer response to tricyclics, greater frequency of cortisol
depressant response in both groups hypersecretion
Wolfersdorf Inpatient Retrospective study that assessed the significance of 46 patients with MD-psych com- MDpsych patients were less suicidal than MD non-psy. The two Decreased, p o 0.01 0
et al. delusional symptoms for suicidality in MDpsych pared to 46 MD-nonpsych patients groups did not differ in frequency of suicide.
(1987) patients
Johnson Population- Cross-sectional and 1-year prospective data. This 114 MD-psych and 662 MD-non- Patients with MDpsych had a more SA, more multiple episodes, more No significant 0
et al. Based study examined the rates and associated features of psych patients persistent illness over 1 year, higher comorbidity with OCD, somati- difference
(1991) MDpsych in a community sample zation, phobia. Findings support its inclusion in DSM-IV as a distinct
subtype
This could be due to the age differences between the two studies, 3.2.4. Population-based studies
since Lykouras et al. (2002) had enrolled a geriatric population, In 1991, using community samples, Johnson et al. (1991) found
whereas patients in the study conducted by Hori et al. (1993) were a significant difference in the rate of suicide attempts between
relatively younger. MD-psych and MD-nonpsych. They used the community sample
An inpatient cross-sectional study (Lee et al., 2013) was con- from the NIMH Epidemiologic Catchment Area study (Johnson
ducted to compare the suicide attempt rates in the MD-psych and et al., 1991), where two sets of interviews were conducted a year
MD-nonpsych groups. Their sample consisted of patients admitted apart at five sites. Patients met DSM III criteria for major depres-
to a geriatric psychiatry ward of Veterans General Hospital- Taipei, sion and reported delusions or hallucinations. The risk for suicide
Taiwan, aged 65 years or older, who met DSM-IV criteria for major attempts was more than double for MD-psych patients compared
depressive disorder. In MD-psych patients, 14 out of 34 (41.2%) to MD-nonpsych patients (27.2% vs. 12.6%, p o 0.05).
patients had current suicide attempts, whereas in 98 MD-non-
psych patients, only 10 (10.2%) had current suicide attempts 3.2.5. Follow-up studies
(p ¼0.001). The odds ratio for suicide attempts in the MD-psych One study (Robinson and Spiker, 1985) found no difference
group was 4.04 times higher than their MD-nonpsych between the two groups; however only 2 suicide attempts in MD-
counterparts. psych and 1 attempt in MD-nonpsych group were reported. In
another inpatient study (Miller and Chabrier, 1987), authors found
3.2.2. Studies including both inpatient and outpatient referrals the risk of suicide attempts 1.5 times higher in MD-psych group,
Incorporating both inpatients and outpatients as part of NIMH but this was not statistically significant. In 1992, a retrospective
sponsored multisite Study of the Pharmacotherapy of Psychotic chart review of elderly suicide attempters admitted to an inpatient
Depression (STOP-PD) study, researchers (Schaffer et al., 2008) unit was conducted (Lyness et al.,1992). In this review, authors
reported that of the first 183 patients with MD-psych: 59.6% re- reported that patients who had made more severe attempts were
ported suicidal ideation or made a suicide attempt during the more likely to be diagnosed as MD-psych, although this trend was
current episode; 20.8% had attempted suicide during the current not significant.
episode; and, 35.5% had a made a lifetime suicide attempt. In
2009, authors (Meyers et al., 2009) reported on the full sample of 3.3. Suicidal ideation
the 259 subjects from the NIMH STOP-PD study: 18.5% subjects
had attempted suicide during the current episode of psychotic Nearly all hospital based studies examining and comparing
depression. In 2010, another group of researchers (Holma et al., rates of suicidal ideation between MD-psych and MD-nonpsych
2010) conducted a prospective 5-year study in patients with major have reported a higher degree of suicidal ideation in the MD-psych
depression. They found no difference in the rate of suicide at- group. Table 3 summarizes studies comparing suicidal ideation in
tempts between MD-psych and MD non-psych groups. the two groups.
In 2014, Park et al., (2014) investigated a cohort of Korean pa-
tients (n¼ 966) of age over 18 years, with MD-psych and MD- 3.3.1. Studies on inpatient psychiatric units
nonpsych to determine which clinical variables were significantly Inpatient studies have found a higher degree of suicidal idea-
different between the two groups. Of 24 MD-psych patients, 41.7% tion in patients with MD-psych as compared to their MD-non-
(N ¼10) attempted suicide, whereas of 942 MD-nonpsych patients, psych counterparts (Nelson et al., 1984, Miller and Chabrier, 1987).
22.1% (N ¼208) attempted suicide (p ¼ 0.023). Nelson et al. (1984) studied family history, cortisol secretion,
clinical variables and 4-week tricyclic antidepressant response in
3.2.3. Outpatient studies 13 MD-psych and 12 MD-nonpsych patients admitted to West
Several outpatient studies have also reported differences in the Virginia University Medical Center. They found that the Hamilton
frequency of suicide attempts between MD-psych and MD-non- Depression Rating scale item for suicidal ideation was significantly
psych patients. An outpatient study (Gaudiano et al., 2008) en- higher in the MD-psych group than the MD-nonpsych group
rolled outpatient psychiatric patients in a study at the Rhode Is- (2.871.0 vs. 1.4 71.4, p o0.05). Another study (Wolfersdorf et al.,
land Hospital Department of Psychiatry and reported a higher 1987) found that at the time of admission, half of the MD-psych
degree of suicide attempts in MD-psych compared to MD-non- patients were suicidal as compared to 76% in the MD-nonpsych
psych patients (p o0.05), although the number of suicide at- group, p o 0.01. In 1993, one study (Hori et al., 1993) found the rate
tempters was small: 6/60 (10%) in MD-psych and 21/1052 (2%) in of suicidal ideation in MD-psych to be 84.2 vs. 67.3 (p o0.05) in
MD-nonpsych. In a recent study (Adeosun and Jeje, 2013), authors MD-nonpsych. The only inpatient adolescent study conducted
enrolled 129 MD-psych and 117 MD-nonpsych outpatients over a (Quinlan et al., 1997) studied adolescents 12–18 years of age, who
four-month period at the largest mental health care facility in were admitted to the University Medical Center in Michigan, over
Nigeria. Diagnoses were made using Structured Clinical Interview a 2-year period. The Suicidal Ideation Questionnaire-JR was used
for DSM-IV. The rates of suicide attempts were significantly higher to assess presence and severity of suicidal ideation. Patients with
in MD-psych subjects compared to MD-nonpsych patients (17.8 vs. MD-psych had statistically significantly higher scores (48.6 722.7)
6.8%, p o0.009). In 2014, Kelleher et al. (2014) conducted a clinical compared to the MD-nonpsych group (28.3 7 24.2) p ¼0.017 and
case-clinical control study in the Republic of Ireland, where par- were found to express considerable more frequent and severe
ticipants were 108 adolescents, 12–16 years of age, with and suicidal ideation.
without psychotic experiences. Authors aimed to assess the pre-
valence of attenuated psychotic experiences and their relationship 3.3.2. Outpatient studies
with affective, anxiety and behavioral disorders, multimorbid Several outpatient studies also found similar results. One study
psychopathology as well as with global functioning and suicidal (Gaudiano et al., 2008) reported increased thoughts of death
behavior. 37 (34%) individuals were diagnosed with Major De- (OR¼ 3.7) thoughts of suicide (OR¼ 2.8) suicide plan (OR ¼2.4) and
pressive Disorder. They found that of those with psychotic ex- suicide attempt (OR ¼5.5) in MD-psych patients compared with
periences (MD-psych ¼25), 64% attempted suicide, whereas of MD-nonpsych patients. Of 60 MD-psych patients, 78% (47) had
those with no psychotic experiences (MD-nonpsych ¼12), 17% at- thoughts of death, 48% (29) had thoughts of suicide and 23% (14)
tempted suicide [Odds Ratio¼ 8.89; 95%Confidence Interval (1.59– had a suicide plan compared to MD-nonpsych patients where 49%
49.83)]. (520) had thoughts of death, 25% (265) had thoughts of suicide
30 I. Zalpuri, A.J. Rothschild / Journal of Affective Disorders 198 (2016) 23–31
and 12% (126) had a plan. Another study (Adeosun and Jeje, 2013) with missed diagnoses were considered to have a psychotic dis-
found a higher prevalence of suicidal ideation in MD-psych vs. order. This finding suggests that the physicians were missing the
MD-nonpsych (58.1% vs. 40.2%, p o 0.005). psychosis rather than the mood disorder.
A missed diagnosis can lead to less than optimal treatment
3.3.3. Studies including both inpatient and outpatient referrals being rendered (Andreescu et al., 2007) and thus prolong the pa-
In 1999, a large case control study (Thakur et al., 1999) re- tient's recovery time, which in turn may increase the risk of sui-
cruiting inpatients as well as outpatients at Duke University cide. In addition, many patients with psychotic depression feel
Medical Center was conducted. Based on the Carroll Self Rating paranoid and will tend to keep their thoughts to themselves. This
Depression scale (Carroll et al.,1981), the MD-psych group includes suicidal thoughts and ideation making it more difficult for
(n ¼189) reported more suicidal ideation as compared to the MD- the clinician to assess the patient's risk for suicide.
nonpsych group (n¼ 485) (57% vs. 42%, p ¼0.004). They did not A limitation of our review is that some of the studies included
specify what percentage of this was suicidal ideation and what in the review were retrospective analyses and did not use formal
percentage were attempts. diagnostic interviews to distinguish MD-psych from MD-non-
psych. Hence, it is possible that some of the patients in these
3.3.4. Population-based study studies did not receive the correct diagnosis. Another limitation is
In their community sample, Johnson et al. (1991) found no that most studies did not report whether the psychotic symptoms
difference in suicidal ideation between MD-psych and MD-non- in the MD-psych patients were mood-congruent or mood-incon-
psych (47.8% in MD-nonpsych vs. 55.8% in MD-psych), although gruent. Consequently, we cannot make any conclusions as to
they had found a higher rate of suicide attempts in the MD-psych whether the type of delusion increases the risk of suicide. The
group when compared to the MD-nonpsych group. papers we reviewed did not describe whether the auditory hal-
lucinations observed in the MD-psych patients were command
hallucinations encouraging the patients to engage in suicidal be-
4. Conclusions havior and enhancing their risk of suicide. Ultimately, only 24
studies met the inclusion criteria; several had small sample size
The purpose of this study was to ascertain, by undertaking a and a quality score of zero, thereby impacting validity.
comprehensive and critical review of the psychiatric literature, Studies that reported suicide attempts or completed suicides in
whether patients with MD-psych have a higher rate of completed the follow up period did not list the psychiatric status of victims at
suicides, suicide attempts or suicidal ideation than patients with the time. Since suicidal behavior in depression or bipolar disorders
MD-nonpsych. The answer to this question and an explanation for is state and severity dependent, it is possible that the attempts or
the seemingly conflicting data in previous studies appears to be deaths occurred during major depressive episodes. Suicidal beha-
related to whether one looks at differences during the acute epi- vior in patients with major mood disorder is extremely rare during
sode or over the long-term. euthymia/remission (Holma et al., 2010). Whereas on the other
Our review indicates that during the acute episode of depres- hand, studies show that psychotic/nonpsychotic nature of de-
sion, particularly when the patient is hospitalized on an inpatient pression is quite consequent across consecutive depressive
psychiatric unit, patients with MD-psych have higher rates of episodes.
suicide, suicide attempts, and suicidal ideation than patients with In conclusion, although studies in the literature appear to be
MD-nonpsych. This makes sense because in the acute episode the conflicting, our comprehensive review indicates that patients with
patient with psychotic depression has the combination of a se- psychotic depression tend to have higher rates of suicide, suicide
verely depressed mood and delusions, which are frequently of a attempts, and suicidal ideation when compared to patients with
nihilistic and depressing quality. non-psychotic depression. This appears to be particularly true
Studies done after the acute episode has resolved (e.g. long- during the acute episode of the illness.
term follow-up studies) are less likely to show a difference be-
tween psychotic and non-psychotic depressed patients on rates of
Acknowledgments
suicide, suicide attempts, or suicidal ideation. We believe this is Supported in part by the Irving S. and Betty Brudnick Endowed Chair of Psy-
due to the fact that the patients may have received treatment and chiatry at University of Massachusetts Medical School (Dr. Rothschild).
the patients with MD-psych are less likely to be suffering from the
combination of depression and nihilistic, negativistic delusions.
The findings from our comprehensive literature review have References
practical implications for the clinician treating patients with psy-
chotic depression; specifically, that the risk of suicide remains Adeosun, I.I., Jeje, O., 2013. Symptom profile and severity in a sample of nigerians
increased until the patient has had remission of their depressive with psychotic versus nonpsychotic major depression. Depression Res. Treat.
2013.
and psychotic symptomatology. The first important step in pre-
Andreescu, C., Mulsant, B.H., Peasley-Micklus, C., Rothschild, A.J., Flint, A.J., Heo, M.,
venting suicide in patients with MD-psych is recognition of the Caswell, M., Whyte, E.M., Meyers, B.S., 2007. Persisting low use of anti-
illness. Unfortunately, psychotic depression is frequently not re- psychotics in the treatment of major depressive disorder with psychotic fea-
tures. J. Clin. Psychiatry 68, 194–200.
cognized (Rothschild et al., 2008) which can lead to an under
Aronson, T.A., Shukla, S., Hoff, A., 1987. Continuation therapy after ECT for delu-
appreciation of the risk of suicide. It is likely that the missed di- sional depression: a naturalistic study of prophylactic treatments and relapse. J.
agnosis rate observed in this study was a conservative estimate of ECT 3 (4), 251–259.
the rate in the general clinical population, as patients with co- Black, D.W., Winokur, G., Nasrallah, A., 1988. Effect of psychosis on suicide risk in
1593 patients with unipolar and bipolar affective disorders. Am. J. Psychiatry
morbid conditions such as a history of substance abuse in the past 145 (7), 849–852.
3 months or unstable medical conditions were excluded. MD- Carroll, B.J., Feinberg, M., Smouse, P.E., Rawson, S.G., Greden, J.F., 1981. The Carroll
psych was most commonly misdiagnosed as either major de- rating scale for depression. I. Development, reliability and validation. Br. J.
Psychiatry 138 (3), 194–200.
pressive disorder without psychotic features, depression NOS, or Coryell, W., Tsuang, M.T., McDaniel, J., 1982. Psychotic features in major depression:
mood disorder NOS. It was quite striking that none of the patients is mood congruence important? J. Affect. Disord. 4 (3), 227–236.
I. Zalpuri, A.J. Rothschild / Journal of Affective Disorders 198 (2016) 23–31 31
Coryell, W., Endicott, J., Keller, M., 1987. The importance of psychotic features to Lykouras, L., Gournellis, R., Fortos, A., Panagiotis, O., Christodoulou, G.N., 2002.
major depression: course and outcome during a 2‐year follow-up. Acta Psy- Psychotic (delusional) major depression in the elderly and suicidal behaviour. J.
chiatr. Scand. 75 (1), 78–85. Affect. Disord. 69 (1), 225–229.
Coryell, W., Leon, A., Winokur, G., Endicott, J., Keller, M., Akiskal, H., Solomon, D., Lyness, J.M., Conwell, Y., Nelson, J.C., 1992. Suicide attempts in elderly psychiatric
1996. Importance of psychotic features to long-term course in major depressive inpatients. J. Am. Geriatr. Soc. 40 (4), 320–324.
disorder. Am. J. Psychiatry 153 (4), 483–489. Meyers, B.S., Flint, A.J., Rothschild, A.J., Mulsant, B.H., Whyte, E.M., Peasley-Miklus,
Gaudiano, A.B., Young, D., Chelminski, I., Zimmerman, M., 2008. Depressive C., Papademetriou, E., Leon, A.C., Heo, M., 2009. A double-blind randomized
symptom profiles and severity patterns in outpatients with psychotic vs. controlled trial of olanzapine plus sertraline vs. olanzapine plus placebo for
nonpsychotic major depression. Compr. Psychiatry 49 (5), 421–429. psychotic depression: the study of pharmacotherapy of psychotic depression
Glassman, A.H., Roose, S.P., 1981. Delusional depression: a distinct clinical entity? (STOP-PD). Arch. Gen. Psychiatry 66 (8), 838–847.
Arch. Gen. Psychiatry 38 (4), 424–427. Miller, F., Chabrier, L.A., 1987. The relation of delusional content in psychotic de-
Frangos, E., Athanassenas, G., Tsitourides, S., Psilolignos, P., Katsanou, N., 1983. pression to life-threatening behavior. Suicide Life Threat. Behav. 17 (1), 13–17.
Psychotic depressive disorder: a separate entity? J. Affect. Disord. 5 (3), Nelson, W.H., Khan, A., Orr, W.W., 1984. Delusional depression: phenomenology,
259–265. neuroendocrine function, and tricyclic antidepressant response. J. Affect. Dis-
Hagnell, O., Lanke, J., Rorsman, B., 1981. Suicide rates in the lundby study: mental ord. 6 (3), 297–306.
illness as a risk factor for suicide. Neuropsychobiology 7 (5), 248–253. Park, S.C., Lee, H.Y., Sakong, J.K., Jun, T.Y., Lee, M.S., Kim, J.M., Kim, J.B., Yim, H.W.,
Henriksson, M.M., Aro, H.M., Kuoppasalmi, K.I., Jouko, K.L., 1993. Mental disorders Park, Y.C., 2014. Distinctive clinical correlates of psychotic major depression:
and comorbidity in suicide. Am. J. Psychiatry 1 (50), 935–940.
the CRESCEND study. Psychiatry Invest. 11 (3), 281–289.
Holma, K.M., Melartin, T.K., Haukka, J., Holma, I.A.K., Sokero, T.P., Isometsa, E.T.,
Quinlan, P.E., King, C.A., Hanna, G.L., Ghaziuddin, N., 1997. Psychotic versus non-
2010. Incidence and predictors of suicide attempts in DSM-IV major depressive
psychotic depression in hospitalized adolescents. Depression Anxiety 6 (1),
disorder: a five-year prospective study. Am. J. Psychiatry 167 (7), 801–808.
40–42.
Hori, M., Shiraishi, H., Koizumi, J., 1993. Delusional depression and suicide. Psy-
Robinson, D.G., Spiker, D.G., 1985. Delusional depression: a one year follow-up. J.
chiatry Clin. Neurosci. 47 (4), 811–817.
Affect. Disord. 9 (1), 79–83.
Isometsä, E., Henriksson, M., Aro, H., Heikkinen, M., Kuoppasalmi, K., Lonnqvit, J.,
Roose, S.P., Glassman, A.H., Walsh, B.T., Woodring, S., Vital-Herne, J., 1983. De-
1994. Suicide in psychotic major depression. J. Affect. Disord. 31 (3), 187–191.
pression, delusions, and suicide. Am. J. Psychiatry 140 (9), 1159–1162.
Jadad, A.R., Moore, R.A., Carroll, D., Jenkinson, C., Reynolds, D.J.M., Gavaghan, D.J.,
Rothschild, A.J., Winer, J., Flint, A.J., Mulsant, B.H., Whyte, E.M., Heo, M., Fratoni, S.,
McQuay, H.J., 1996. Assessing the quality of reports of randomized clinical
Gabriele, M., Kasapinovic, S., Meyers, B.S., 2008. Missed diagnosis of psychotic
trials: is blinding necessary? Controll. Clin. Trials 17 (1), 1–12.
Johnson, J., Horwath, E., Weissman, M.M., 1991. The validity of major depression depression at 4 academic medical centers. J. Clin. Psychiatry 69 (8), 1293–1296.
with psychotic features based on a community study. Arch. Gen. Psychiatry 48 Schaffer, A., Flint, A.J., Smith, E., Rothschild, A.J., Mulsant, B.H., Szanto, K., Peasley-
(12), 1075–1081. Miklus, C., Heo, M., Papademetriou, E., Meyers, B.S., 2008. Correlates of sui-
Kelleher, I., Devlin, N., Wigman, J.T., Kehoe, A., Murtagh, A., Fitzpatrick, C., Cannon, cidality among patients with psychotic depression. Suicide Life Threat. Behav.
M., 2014. Psychotic experiences in a mental health clinic sample: implications 38 (4), 403–414.
for suicidality, multimorbidity and functioning. Psychol. Med. 44 (08), Schatzberg, A., Rothschild, A.J., 1992. Psychotic (delusional) major depression:
1615–1624. should it be included as a distinct syndrome in DSM-IV. Am. J. Psychiatry 149
Leadholm, A.K.K., Rothschild, A.J., Nielsen, J., Beck, P., Ostergaard, S.D., 2014. Risk (6), 733–745.
factors for suicide among 34,671 patients with psychotic and non-psychotic Spitzer, R.L., Endicott, J., Robins, E., 1978. Research diagnostic criteria: rationale and
severe depression. J. Affect. Disord. 156, 119–125. reliability. Arch. Gen. Psychiatry 35 (6), 773–782.
Lee, T.W., Tsai, S.J., Yang, C.H., Hwang, J.P., 2013. Clinical and phenomenological Suominen, K., Haukka, J., Valtonen, H.M., Lonnqvist, J., 2009. Outcome of patients
comparisons of delusional and non-delusional major depression in the chinese with major depressive disorder after serious suicide attempt. J. Clin. Psychiatry
elderly. Int. J. Geriatr. Psychiatry 18 (6), 486–490. 70 (10), 1372–1378.
Lykouras, E., Malliaras, D., Christodoulou, G.N., Papakostas, A., Voulgari, A., Tzonou, Thakur, M., Hays, J., Krishnan, K.R.R., 1999. Clinical, demographic and social char-
A., Stefanis, C., 1986. Delusional depression: phenomenology and response to acteristics of psychotic depression. Psychiatry Res. 86 (2), 99–106.
treatment. Acta Psychiatr. Scand. 73 (3), 324–329. Vythilingam, M., Chen, J., Bremner, J.D., Mazure, C.M., Maciejewski, P.K., Nelson, J.C.,
Lykouras, L., Christodoulou, G.N., Malliaras, D., Stefanis, C., 1994. The prognostic 2003. Psychotic depression and mortality. Am. J. Psychiatry 160 (3), 574–576.
importance of delusions in depression: a 6-year prospective follow-up study. J. Wolfersdorf, M., Keller, F., Steiner, B., Hole, G., 1987. Delusional depression and
Affect. Disord. 32 (4), 233–238. suicide. Acta Psychiatr. Scand. 76 (4), 359–363.