Annals of Internal Medicine䊛

In the Clinic®

Type 2 Diabetes
D
iabetes is one of the most common ill-
nesses encountered by internists. Cur- Screening and Prevention
rently, an estimated 29.1 million people
in the United States have diabetes, and only
21.0 million of these cases have been diag- Diagnosis and Evaluation
nosed (1). The incidence of diabetes is increas-
ing because of the aging and changing ethnic
mix of the population and because of increasing Treatment
obesity. Based on current trends, it is expected
that the prevalence of diabetes will nearly dou-
ble by 2050 (2). Practice Improvement

Tool Kit

Patient Information

The CME quiz is available at www.annals.org/intheclinic.aspx. Complete the quiz to earn up to 1.5 CME credits.

Physician Writer CME Objective: To review current evidence for prevention, screening, diagnosis, treatment,
Sandeep Vijan and patient information of type 2 diabetes.
Funding source: American College of Physicians.
Disclosures: Dr. Vijan, ACP Contributing Author, has disclosed no conflicts of interest.
Disclosures can also be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms
.do?msNum=M14-2388.
With the assistance of additional physician writers, Annals of Internal Medicine editors
develop In the Clinic using resources of the American College of Physicians, including
ACP Smart Medicine and MKSAP (Medical Knowledge and Self-Assessment Program).
© 2015 American College of Physicians

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017


Screening and Prevention
1. Centers for Disease Con-
trol and Prevention. Na-
tional Diabetes Statistics
Report, 2014 [Internet].
[cited 2014 Oct 14]. Avail-
able from: www.cdc.gov
/diabetes/pubs/statsre-
port14.htm
2. Boyle JP, Thompson TJ,
Gregg EW, Barker LE,
Williamson DF. Projection
of the year 2050 burden
of diabetes in the US adult
population: dynamic mod-
eling of incidence, mortal-
ity, and prediabetes preva-
lence. Popul Health Metr.
2010;8:29. [PMID:
20969750]
3. Gregg EW, Williams DE,
Geiss L. Changes in
diabetes-related complica-
tions in the United States
[Letter]. N Engl J Med.
2014;371:286-7. [PMID:
25014698]
4. American Diabetes Associ-
ation. Standards of medi-
cal care in diabetes---
2014. Diabetes Care.
2014;37 Suppl 1:S14-80.
[PMID: 24357209]
5. American Diabetes Associ-
ation. Standards of medi-
cal care in diabetes---
2013. Diabetes Care.
2013;36 Suppl 1:S11-66.
[PMID: 23264422]
6. Simmons RK, Echouffo-
Tcheugui JB, Sharp SJ,
Sargeant LA, Williams KM,
Prevost AT, et al. Screen-
ing for type 2 diabetes
and population mortality
over 10 years (ADDITION-
Cambridge): a cluster-
randomised controlled
trial. Lancet. 2012;380:
1741-8. [PMID:
23040422]
7. Kahn R, Alperin P, Eddy D,
Borch-Johnsen K, Buse J,
Feigelman J, et al. Age at
initiation and frequency of
screening to detect type 2
diabetes: a cost-
effectiveness analysis.
Lancet. 2010;375:1365-
74. [PMID: 20356621]
8. The cost-effectiveness of
screening for type 2 dia-
betes. CDC Diabetes Cost-
Effectiveness Study Group,
Centers for Disease Con-
trol and Prevention.
JAMA. 1998;280:1757-
63. [PMID: 9842951]
9. Hofer TP, Vijan S, Hayward
RA. Estimating the micro-
vascular benefits of
screening for type 2 dia-
betes mellitus. Int J Tech-
nol Assess Health Care.
2000;16:822-33. [PMID:
11028137]
姝 2015 American College of Physicians
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017
Although care and complication
rates are clearly improving (3),
complications are still common,
and diabetes is among the lead-
ing causes of vision loss, amputa-
tion, and end-stage renal disease
Should we screen for type 2
diabetes?
Current data suggest that about
1 in 4 persons with diabetes are
unaware of their disease (1). Dia-
betes has a fairly long asymptom-
atic phase, during which some
patients will develop early dis-
ease complications, such as
background retinopathy or mi-
croalbuminuria. Some groups
have therefore suggested that
screening should be done every
third year in persons older than
45 years as well as in those
younger than 45 who have dia-
betes risk factors (see the Box:
Risk Factors for Type 2 Diabetes)
(4, 5).
However, at present no definitive
evidence shows that screening
improves health outcomes. In a
single, large-scale screening trial
in the United Kingdom, screen-
ing for diabetes among high-risk
persons did not lead to changes
in outcomes in 10 years of
follow-up (6). Evidence from
modeling studies is inconsistent,
and it is unclear whether screen-
ing is likely to substantially im-
prove outcomes or to be cost-
effective when applied broadly
(7–9). There is thus a lack of con-
sensus on who should be
screened, the magnitude of ben-
efit (if any), and how often
screening should be done.
In a cluster randomized trial in 33 practices in
England, 15 089 patients who were at high
risk for diabetes based on questionnaires were
invited for screening; 73% agreed, and 3%
were ultimately diagnosed with previously un-
known diabetes. After 9.6 years of follow-up,
there was no difference in mortality between
patients who were screened and those who
ITC2 In the Clinic
in the United States (4). In addi-
tion, it is a substantial risk factor
for atherosclerotic disease, which
is the leading cause of morbidity,
mortality, and expenditures in
diabetic persons.
were not (hazard ratio [HR], 1.06 [95% CI,
0.90 –1.25]), nor in cardiovascular (HR, 1.02
[CI, 0.75–1.38) or diabetes-related (HR, 1.26
[CI, 0.75–2.10]) mortality.
Which patients are likely to
benefit from screening?
Risk Factors for Type 2
Diabetes
Age >45 years
First-degree relative with type 2
diabetes
African American, Hispanic, Asian,
Pacific Islander, or
Native-American ethnicity
History of gestational diabetes or
delivery of infant weighing
≥9 lb
The polycystic ovary syndrome
Overweight, especially abdominal
obesity
Cardiovascular disease,
hypertension, dyslipidemia,
or other features of the
metabolic syndrome
Diabetes screening is most likely
to improve outcomes in patients
with risk factors for cardiovascu-
lar disease, particularly if treat-
ment goals differ for those with
and without diabetes. While pre-
vious recommendations sug-
gested screening for diabetes in
persons with hypertension, re-
cent studies suggest that blood
pressure treatment goals should
be the same for those with and
without diabetes (10), limiting the
rationale for screening in persons
with hypertension. Recent guide-
lines for the management of lip-
ids suggest a risk-based ap-
proach to initiation of lipid-
lowering therapy, using a risk
Annals of Internal Medicine 3 March 2015
 Table 1. Diagnostic Criteria for Type 2 Diabetes
Diagnosis Hemoglobin A1c Fasting Plasma
Level, % Glucose Level

mmol/L mg/dL
Prediabetes 5.7–6.4 5.55–6.94 100–125
Diabetes ≥6.5 ≥7.0 ≥126
10. Cushman WC, Evans
GW, Byington RP, Goff
DC Jr, Grimm RH Jr,
Cutler JA, et al; ACCORD
calculator that includes the pres- 30% and 10% of energy consumed, respec- Study Group. Effects of
tively; an increase in fiber intake; and moder- intensive blood-pressure
ence of diabetes as a risk factor control in type 2 diabe-
(11). Knowledge of diabetes sta- ate exercise for at least 30 minutes per day tes mellitus. N Engl J
(12). Med. 2010;362:1575-
tus alters the likelihood of recom- 85. [PMID: 20228401]
mending treatment and may ar- 11. Stone NJ, Robinson JG,
The Diabetes Prevention Project, a random- Lichtenstein AH, Bairey
gue for screening the population ized, controlled trial that involved 3234 U.S. Merz CN, Blum CB, Eckel
RH, et al; American Col-
who would otherwise not be can- patients with prediabetes (mean age, 51 lege of Cardiology/Amer-
didates for lipid-lowering ther- years; mean body mass index, 34 kg/m2), ican Heart Association
Task Force on Practice
apy; however, at present there showed that a lifestyle modification program Guidelines. 2013 ACC/
are no formal evaluations of aimed at a 7% weight loss reduced the cumu- AHA guideline on the
treatment of blood cho-
the effects of diabetes lative incidence of diabetes over 3 years, from lesterol to reduce athero-
sclerotic cardiovascular
screening on lipid treatment 29% to 14% (relative risk [RR], 0.42 [CI, 0.34 – risk in adults: a report of
recommendations. 0.52]) compared with placebo (13). Ten-year the American College of
Cardiology/American
follow-up found persistence of the initial effect Heart Association Task
Diabetes is more likely to be de- of lifestyle, although after the study period the Force on Practice Guide-
tected in persons with risk factors lines. Circulation. 2014;
rates in the lifestyle and placebo groups were 129:S1-45. [PMID:
for the disease (Table 1). How- similar, implying that the intervention must be 24222016]
12. Tuomilehto J, Lindström
ever, beyond the increased prev- maintained for benefit to continue (14). J, Eriksson JG, Valle TT,
alence of disease, there is no Hämäläinen H, Ilanne-
A randomized, controlled trial that involved Parikka P, et al; Finnish
consistent evidence supporting Diabetes Prevention
improved clinical outcomes with 577 Chinese adults with impaired glucose tol- Study Group. Prevention
erance assigned to diet, exercise, both, or nei- of type 2 diabetes melli-
screening, and recommenda- tus by changes in life-
ther found that the incidence of diabetes over style among subjects
tions are based largely on expert with impaired glucose
6 years was 68% among persons in the “nei-
opinion. tolerance. N Engl J Med.
ther” group, 44% in the diet group, 41% in the 2001;344:1343-50.
[PMID: 11333990]
Can type 2 diabetes be exercise group, and 46% in the “both” group. 13. Knowler WC, Barrett-
prevented? All 3 interventions resulted in statistically sig- Connor E, Fowler SE,
Hamman RF, Lachin JM,
Several high-quality randomized nificant reductions in the progression to diabe- Walker EA, et al; Diabe-
trials show that lifestyle changes tes (15). tes Prevention Program
Research Group. Reduc-
in diet and exercise lead to sub- Some medications can prevent
tion in the incidence of
type 2 diabetes with
stantial reductions in the inci- diabetes onset in patients with lifestyle intervention or
metformin. N Engl J
dence of type 2 diabetes in per- prediabetes. Med. 2002;346:393-
sons with “prediabetes,” defined 403. [PMID: 11832527]
14. Knowler WC, Fowler SE,
as having impaired fasting glu- In the medication group of the Diabetes Pre- Hamman RF, Christophi
cose or impaired glucose toler- vention Project, metformin (850 mg twice CA, Hoffman HJ, Brenne-
man AT, et al; Diabetes
ance. These programs achieved daily) reduced the cumulative incidence of di- Prevention Program
modest weight loss (generally abetes from 29% to 22% over 3 years (RR, 0.69 Research Group. 10-year
follow-up of diabetes
5%–7% of body weight) but were [CI, 0.57– 0.83], a significant but smaller reduc- incidence and weight
markedly effective. tion than that observed with the lifestyle inter- loss in the Diabetes
Prevention Program
vention in this trial (13). Ten-year follow-up Outcomes Study. Lancet.
In a randomized, unblinded, controlled trial of again showed persistence of initial effect, al- 2009;374:1677-86.
[PMID: 19878986]
522 overweight Finnish patients with im- though after the study period the rates in the 15. Pan XR, Li GW, Hu YH,
paired glucose tolerance (mean age, 55 years), metformin and placebo group were similar (14). Wang JX, Yang WY, An
ZX, et al. Effects of diet
an intervention aimed at a 5% reduction in and exercise in prevent-
weight decreased the incidence of newly diag- In the randomized, double-blind, international ing NIDDM in people
with impaired glucose
nosed type 2 diabetes over 3 years, from 23% Study to Prevent NonInsulin-Dependent Diabetes tolerance. The Da Qing
to 11%. The intervention involved personal Mellitus, which involved 1429 patients with im- IGT and Diabetes Study.
Diabetes Care. 1997;20:
counseling sessions to encourage a reduction paired glucose tolerance, acarbose (100 mg 3 537-44. [PMID:
in total and saturated fat intake to less than times daily) reduced the incidence of diabetes 9096977]

3 March 2015 Annals of Internal Medicine In the Clinic ITC3 姝 2015 American College of Physicians

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017

 from 42% to 32% compared with placebo. The
relative risk reduction over 3 years was 25% (16).
The DREAM trial (Diabetes Reduction Assess-
ment with ramiripril and rosiglitazone Medica-
tion) randomly assigned 5269 adults without
previous cardiovascular disease but with im-
paired fasting glucose, impaired glucose toler-
ance, or both to rosiglitazone 8 mg per day or
placebo and to rosiglitazone up to 15 mg per
day or placebo. After a median 3 years, 11.6%
of patients who received rosiglitazone devel-
oped diabetes or died compared with 26.0%
of patients who received placebo (HR, 0.40 [CI,
0.35 to 0.46]). Cardiovascular event rates were
statistically similar in both groups (17).
The implications of prevention for diabetes
screening have not been fully elucidated,
but screening is generally necessary to
identify the high-risk prediabetes popula-
tion. However, because lifestyle and dietary
modification are likely to benefit everyone
regardless of diabetes status, the advan-
tages of labeling patients as “prediabetic”
is uncertain. The most likely option is to
consider screening in persons who are at
particularly high-risk (that is, those with
multiple factors as discussed in Table 1)
and to implement prevention, perhaps with
medication therapy as well as lifestyle mod-
ification, in persons with high-risk predia-
betes. There are national resources being
put in place to help disseminate lifestyle
changes for at-risk patients; more detail is
available at www.cdc.gov/diabetes/prevention
/index.htm.

Screening and Prevention... Little direct evidence shows clinical bene-

fit from broad-based screening programs for type 2 diabetes. The sin-
gle large-scale trial did not show mortality benefits at 10 years, and
modeling studies have yielded inconsistent results. Diabetes can clearly
be prevented in persons who have prediabetes with programs aimed at
modest weight loss, and medication may be indicated for those who
cannot achieve lifestyle goals. However, because diet and exercise pro-
grams tend to be universally beneficial, screening may best be preserved
for persons at particularly high risk, largely to identify those who may bene-
fit from medications to prevent diabetes. Guidelines for lifestyle change
suggest that loss of about 7% of body weight and 150 minutes of exercise
per week are enough to substantially reduce diabetes risk.

CLINICAL BOTTOM LINE

Diagnosis and Evaluation

16. Chiasson JL, Josse RG,
Gomis R, Hanefeld M,
Karasik A, Laakso M;
STOP-NIDDM Trail Re-
search Group. Acarbose
for prevention of type 2
diabetes mellitus: the
STOP-NIDDM ran-
domised trial. Lancet.
2002;359:2072-7.
[PMID: 12086760]
17. Gerstein HC, Yusuf S,
Bosch J, Pogue J, Sheri-
dan P, Dinccag N, et al;
DREAM (Diabetes REduc-
tion Assessment with
ramipril and rosiglita-
zone Medication) Trial
Investigators. Effect of
rosiglitazone on the
frequency of diabetes in
patients with impaired
glucose tolerance or
impaired fasting glucose:
a randomised controlled
trial. Lancet. 2006;368:
1096-105. [PMID:
16997664]
What are the diagnostic criteria
for type 2 diabetes in
nonpregnant adults?
Clinicians should confirm the di-
agnosis of diabetes in persons
with classic symptoms (polyuria,
polydipsia, polyphagia, and
weight loss) or in those with evi-
dence of diabetes complications
(retinopathy, nephropathy, neu-
ropathy, impotence, acanthosis
nigricans, or frequent infections).
There are many tests that can be
used to diagnose type 2 diabetes;
however, due to ease of use and
reliability, the current recommen-
dation is to measure hemoglobin
A1c (HbA1c) levels, with a threshold
of ≥6.5% being diagnostic for dia-
betes (4). Other tests can also be
used, including measuring fasting
plasma glucose levels, with a level
of ≥126 mg/dL confirmed by test-
ing on a different day being diag-
nostic for diabetes. Alternatively,
diabetes can be diagnosed in per-
sons with classic symptoms and a
nonfasting glucose ≥200 mg/dL,
again confirmed by a second test.
Finally, an oral glucose tolerance
test (OGTT) could be used, with a
2-hour plasma glucose level of 200
mg/dL considered diagnostic for
diabetes.
Prediabetes can be diagnosed in
persons with an HbA1c level
5.7%– 6.4%, fasting glucose levels
100 to 125 mg/dL, or an OGTT

姝 2015 American College of Physicians ITC4 In the Clinic Annals of Internal Medicine 3 March 2015

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017

 with a 2-hour plasma glucose
level140 –199 mg/dL (Table 1).
What should the initial
evaluation of patients with
newly diagnosed type 2
diabetes include?
Providers should conduct a
detailed history and physical ex-
amination, including review of
diet; physical activity; and assess-
ment of cardiovascular, cerebro-
vascular, and erectile dysfunc-
tion. The initial evaluation should
include blood pressure measure-
ment and inspection for possible
diabetes complications via car-
diovascular, neurologic, skin,
and foot examinations. Labora-
tory tests should assess levels of
glucose control (HbA1c level),
cholesterol levels, and ne-
phropathy (urine microalbumin–
creatinine ratio and serum cre-
atinine), along with liver
function testing for persons who
are likely to need lipid-lowering
therapy. At diagnosis, ophthal-
mologic assessment should be
done to evaluate for
retinopathy.

Diagnosis and Evaluation... Type 2 diabetes is common and should be

considered when patients present with suggestive symptoms (e.g.,
polyuria or polydipsia), signs (e.g., acanthosis nigricans), or complica-
tions of disease (e.g., retinopathy). The diagnosis can be confirmed by
HbA1c levels measuring ≥6.5% or higher or by fasting plasma glucose
levels >7.0 mmol/L (126 mg/dL) on 2 occasions at least 1 day apart.
Random plasma glucose levels and OGTT can also be used to diagnose
type 2 diabetes. Newly diagnosed patients should be examined for hy-
pertension, as well as neurologic, ophthalmologic, and podiatric com-
plications. The initial laboratory evaluation should include an assess-
ment of glucose control, a lipid profile, and measurement of the urine
microalbumin– creatinine ratio.

CLINICAL BOTTOM LINE

What are the components of
nondrug therapy for patients
with type 2 diabetes?
Lifestyle changes, primarily diet
and exercise, are the corner-
stones of managing type 2 diabe-
tes and should be considered
first-line therapy for patients un-
less severe hyperglycemia re-
quires immediate medical treat-
ment. No one diet or exercise
regimen applies to all patients
with diabetes, and an individual-
ized assessment should be used
to develop a feasible strategy.
The American Diabetes Associa-
tion nutrition guidelines can be
accessed at http://care
.diabetesjournals.org/content
/37/Supplement_1/S120.full.
Treatment
In a study of patients with newly diagnosed
type 2 diabetes, diet initially reduced HbA1c
levels by 2.25 percentage points. However,
control deteriorated over time and most pa-
tients eventually required drug therapy (18).
A meta-analysis of 14 randomized trials that
compared exercise with no exercise and in-
volved a total of 377 patients with type 2 dia-
betes showed that exercise significantly im-
proved glycemic control, reduced visceral
adipose tissue, and reduced plasma triglycer-
ides even in the absence of weight loss (19).
What is the role of home
18. Intensive blood-glucose
glucose monitoring? control with sulphony-
lureas or insulin com-
Home glucose monitoring allows pared with conventional
patients and providers to assess treatment and risk of
complications in patients
glucose control longitudinally with type 2 diabetes
(UKPDS 33). UK Prospec-
and can provide real-time feed- tive Diabetes Study
back on the effect of glucose (UKPDS) Group. Lancet.
1998;352:837-53.
treatments. Home monitoring is [PMID: 9742976]

3 March 2015 Annals of Internal Medicine In the Clinic ITC5 姝 2015 American College of Physicians

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017


19. Thomas DE, Elliott EJ,
Naughton GA. Exercise
for type 2 diabetes melli-
tus. Cochrane Database
Syst Rev. 2006:
CD002968. [PMID:
16855995]
20. Malanda UL, Welschen
LM, Riphagen II, Dekker
JM, Nijpels G, Bot SD.
Self-monitoring of blood
glucose in patients with
type 2 diabetes mellitus
who are not using insu-
lin. Cochrane Database
Syst Rev. 2012;1:
CD005060. [PMID:
22258959]
姝 2015 American College of Physicians
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017
considered part of the standard
of care for persons receiving in-
sulin therapy to allow sensible
dose adjustments and to help
determine whether symptoms
are due to hyperglycemia or hy-
poglycemia. The optimum fre-
quency of home monitoring has
not been formally evaluated and
is usually left to the discretion of
the patient and provider. The
role of home glucose monitoring
in guiding oral therapy is less
clear; a formal evidence review
found a small reduction in HbA1c
levels at 6 months, but this bene-
fit subsided by 12 months, sug-
gesting that self-monitoring has
no sustained effect (20).
Patients are generally advised to
monitor fasting and premeal glu-
cose levels. However, postpran-
dial measurement may be helpful
in persons with elevated HbA1c
levels despite normal fasting lev-
els. Some observational data
suggest that postmeal glucose
excursions may be tied to cardio-
vascular risk, leading some ex-
perts to recommend routine
postprandial monitoring. How-
ever, thus far no trials have
shown that treatment of these
excursions reduces cardiovascu-
lar risk.
What is the target HbA1c level?
There is no clear single HbA1c
target that applies to all patients
with type 2 diabetes. Most orga-
nizations and quality measure-
ment groups advocate a target
≤7% for most patients, based on
the results of the U.K. Prospective
Diabetes Study (UKPDS) (18);
however, this was a study of
newly diagnosed patients, who
typically have milder disease. By
the end of the study (10 years),
mean HbA1c levels were close to
8% in the intensive-therapy
group— only a minority of patients
was able to maintain a level <7%.
Persons maintaining better con-
trol had reduced risk for early,
asymptomatic microvascular
ITC6 In the Clinic
complications but did not have
clear benefits on either cardio-
vascular outcomes or symptom-
atic microvascular complications,
such as vision loss, amputation,
or end-stage renal disease.
In a 20-year follow-up of the UKPDS, the group
that was initially assigned to intensive control
had lower rates of myocardial infarction (16.8
vs. 19.6 per 1000 patient-years) and death
(26.8 vs. 30.3 per 1000 patient-years), even
though differences in glycemic control were
not maintained between groups (16).
This implies that early control
may have a “memory” effect and
may provide distant benefits, but
it also implies that significant
benefits take at least 15–20 years
to occur.
Some experts advocate more
aggressive targets for glycemic
control, or to treat where possi-
ble to near-normal glucose lev-
els. Three trials, all of which were
done in patients with existing dia-
betes (in contrast to the UKPDS,
which studied those with newly
diagnosed diabetes) evaluated
this approach.
In ACCORD, a study of 10 251 U.S. patients
(mean age, 62.2 years), the group assigned to
intensive therapy had target HbA1c levels less
than 6.0%, whereas the group assigned to
conventional therapy had target levels ranging
from 7.0%–7.9%. The achieved levels of con-
trol were 6.4% and 7.5%. The trial was stopped
early due to a 22% increase in total mortality
in the intensive-control group (5.0% vs. 4.0%;
P = 0.04). The overall main end point (nonfa-
tal myocardial infarction, nonfatal stroke, and
cardiovascular death) did not differ between
the groups. Hypoglycemia and weight gain
were also more prevalent in the intensive-
control group (21).
The Veterans Affairs Diabetes Trial (VADT) stud-
ied 1791 veterans (mean age, 60.4 years). The
group assigned to intensive control was as-
signed a target HbA1c level 1.5% lower than
the control group. The achieved levels of con-
trol were 6.9% in the intensive group and
8.4% in the control group. No significant ef-
fects on the primary end point (a composite of
cardiovascular events, heart failure, vascular
surgery, and amputation), total mortality, or
microvascular events were found (22).
Annals of Internal Medicine 3 March 2015
 ADVANCE, a multinational study of 11 140 pa-
tients, comprised a group assigned to inten-
sive care with a target HbA1c level ≤6.5% as
well as a control group. The HbA1c level in the
intensive-control group was 6.5% and 7.3% in
the control group. The intensive-control group
had reduced nephropathy (4.1% vs. 5.2%; P =
0.006), but there were no differences in car-
diovascular events or mortality (23).
Interpretation and reconciliation
of the results of the 4 major
glucose-lowering trials are diffi-
cult. It seems that moderate glu-
cose control early in the disease
course (mean HbA1c level 7%
over the first 10 years but with a
worsening trend over that time
frame) does eventually—after
about 2 decades—provide benefit
in lowering cardiovascular events
and mortality. More aggressive
control, at least in the shorter
term, does not seem to provide
substantial benefit and may in-
crease mortality. It is unclear
whether specific subgroups of
patients are more prone to harms
or benefits as the result of ag-
gressive control. The most logical
conclusion from these studies is
that moderate levels of control
(HbA1c level between 7% and
8.5%, which may vary depending
on diabetes duration) will proba-
bly provide the most benefit for
most patients. However, because
patients with a long life expec-
tancy (≥20 years) may eventually
realize benefit from more inten-
sive control (e.g., HbA1c level
<7%), glycemic targets should be
adjusted depending on life ex-
pectancy and comorbid condi-
tions. Evidence from modeling
studies suggests that a long time
horizon for benefit may cause the
burden of therapy—particularly
injectable agents—to outweigh
the benefits for many patients
with type 2 diabetes (24).
When should treatment
include drugs?
Once an HbA1c goal has been
established, pharmacologic man-
agement should be instituted if
3 March 2015 Annals of Internal Medicine
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017
diet and exercise do not achieve
the goal. In general, aside from
patients with only mild HbA1c
elevations, if diet and exercise do
not accomplish the targeted re-
duction in glycemic values within
approximately 6 weeks, pharma-
cologic therapy should be initi-
ated (4). Patients with severe hy-
perglycemia or symptoms may
require pharmacologic interven-
tion immediately, sometimes with
insulin.
How should physicians select
therapies from among the
many oral drug options?
Table 2 provides an overview of
the classes of noninsulin agents
available to treat type 2 diabetes.
Data on the relative efficacy of
the oral therapies at improving
clinical end points, such as re-
duction in diabetes complica-
tions, are insufficient.
The UKPDS found that, in pa-
tients who exceeded ideal body
weight by 20%, metformin was
superior to sulfonylureas and in-
sulin in reducing mortality, de-
spite identical levels of glycemic
control (25). Metformin was also
associated with lower rates of
hypoglycemia and weight gain
than insulin or sulfonylureas. Met-
formin should not be used in per-
sons with severe renal insuffi-
ciency (glomerular filtration rate
<30 mL/min/1.73 m2), symptom-
atic heart failure, or severe liver
disease. It must be stopped be-
fore radiologic procedures that
require intravenous contrast be-
cause of risk for lactic acidosis.
If metformin is contraindicated or
not tolerated, the choice of oral
agents should be dictated by pa-
tient preferences regarding po-
tential side effects, efficacy, and
cost. Although most drugs
achieve similar glycemic control,
differences in mechanism, tolera-
bility, and timing of administra-
tion may help to individualize
care. For example, nonsulfony-
In the Clinic
21. Gerstein HC, Miller ME,
Byington RP, Goff DC Jr,
Bigger JT, Buse JB, et al;
Action to Control Cardio-
vascular Risk in Diabetes
Study Group. Effects of
intensive glucose lower-
ing in type 2 diabetes. N
Engl J Med. 2008;358:
2545-59. [PMID:
18539917]
22. Abraira C, Colwell JA,
Nuttall FQ, Sawin CT,
Nagel NJ, Comstock JP,
et al. Veterans Affairs
Cooperative Study on
glycemic control and
complications in type II
diabetes (VA CSDM).
Results of the feasibility
trial. Veterans Affairs
Cooperative Study in
Type II Diabetes. Diabe-
tes Care. 1995;18:1113-
23. [PMID: 7587846]
23. Patel A, MacMahon S,
Chalmers J, Neal B, Billot
L, Woodward M, et al;
ADVANCE Collaborative
Group. Intensive blood
glucose control and
vascular outcomes in
patients with type 2
diabetes. N Engl J Med.
2008;358:2560-72.
[PMID: 18539916]
24. Vijan S, Sussman JB,
Yudkin JS, Hayward RA.
Effect of patients' risks
and preferences on
health gains with plasma
glucose level lowering in
type 2 diabetes mellitus.
JAMA Intern Med. 2014;
174:1227-34. [PMID:
24979148]
25. Effect of intensive blood-
glucose control with
metformin on complica-
tions in overweight pa-
tients with type 2 diabe-
tes (UKPDS 34). UK
Prospective Diabetes
Study (UKPDS) Group.
Lancet. 1998;352:854-
65. [PMID: 9742977]
ITC7 姝 2015 American College of Physicians
 Table 2. Noninsulin Medications for Type 2 Diabetes
Drug Class Name Initial Dose Maximum Dose Usual Dose
Biguanides Metformin 500 mg twice daily or 2550 mg/d 500–1000 mg twice
850 mg/d daily
Metformin XR 500 mg/d 2000 mg/d 1500–2000 mg/d
Sulfonylureas Glimepiride 1–2 mg/d 8 mg/d 4 mg/d
Glipizide 2.5–5 mg/d 40 mg/d 10–20 mg/d (or twice
daily)
Glipizide SR 5 mg/d 20 mg/d 5–20 mg/d (or twice
daily)
Glyburide 2.5–5 mg/d 20 mg/d 5–20 mg/d (or twice
daily)
Glyburide 0.75–3 mg/d 12 mg/d 3–12 mg/d (or twice
micronized daily)
Thiozolidinediones Pioglitazone 15–30 mg/d 45 mg/d 15–45 mg/d
Rosiglitazone 4 mg/d (or twice daily) 8 mg/d 4–8 mg/d (or twice
daily)
Alpha-glucosidase Acarbose 25 mg/d (w/meal) 100 mg three times 50–100 three times
inhibitors daily daily
Miglitol 25 mg/d (w/meal 100 mg three times 25–100 mg three
daily times daily
Nonsulfonylurea insulin Repaglinide 0.5 mg before meals 4 mg before meals 0.5–4 mg with meals
secretagogues (16 mg/d); wait 1
wk between does
increases
Nateglinide 120 mg three times daily 120 mg three times 60–120 mg three
before meals (60 mg daily before times daily before
three times daily if near meals meals
glycemic goals)
Dipeptidyl peptidase-IV Sitagliptin 100 mg/d 100 mg/d 100 mg/d
inhibitors Saxagliptin 2.5 mg/d 5 mg/d 5 mg/d
Linagliptin 5 mg/d 5 mg/d 5 mg/d
Alogliptin 25 mg/d 25 mg/d 25 mg/d
Sodium glucose-linked Canaglifozin 100 mg/d 300 mg/d 100–300 mg/d
transporter-2 inhibitors Empaglifozin 10 mg/d 25 mg/d 10–25 mg/d
Dapagliflozin 5 mg/d 10 mg/d 5–10 mg/d
Glucagon-like peptide-1 Exenatide 5 mcg twice daily (within 10 mcg twice daily 5–10 mcg/d
agonists (injectable) 60 min before meals)
Exenatide extended 2 mg once/wk 2 mg once/wk 2 mg once/wk
release
Liraglutide 0.6 mg/d 1.8 mg/d 1.2 mg/d
Abliglutide 30 mg once/wk 50 mg once/wk 30–50 mg/wk
Dulaglutide 0.75 mg/wk 1.5 mg/wk 0.75–1.5 mg/wk

SR = sustained release; XR = extended release.

lurea insulin secretagogues (nat- agents is generally the first step,

eglinine, replaglinide) and the
alpha-glucosidase inhibitors (ac-
arbose, miglitol) are generally
administered before meals and
may be useful in persons with
inconsistent mealtimes (e.g.,
truck drivers).
Most patients with diabetes have
worsening glycemic control over
time and require more than one
agent to maintain control. In-
creasing the dose of existing oral
but the response from dose esca-
lation, particularly with metformin
and sulfonylureas, is limited. Pa-
tients therefore often require the
addition of a second oral agent.
Although data showing the effect
of various drug combinations on
glycemic control are available,
few studies have assessed clinical
end points. Several combination
formulations are available and
may provide advantages in con-

姝 2015 American College of Physicians ITC8 In the Clinic Annals of Internal Medicine 3 March 2015

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017

 Table 3. Onset and Mechanisms of Action of Various Types of Insulin
Class Onset of Peak of Duration of
Action Action Action
Rapid acting (insulin analogues ≤ 30 min 0.5–3 h 3–5 h
lispro, aspart, glulisine)
Short-acting (human regular) 0.5–1 h 2–5 h Up to 12 h
Concentrated insulin (U500) 0.5–1 h 6–8 h Up to 24 h
Intermediate-acting (human NPH) 1.5–4 h 4–12 h Up to 24 h
Long-acting (insulin analogues 0.8–4 h Relatively Up to 24 h
glargine, detemir) peakless 26. Nissen SE, Wolski K.
Rosiglitazone revisited:
Human insulin mixtures an updated meta-
70% NPH/30% regular 0.5–2 h 2–12 h Up to 24 h analysis of risk for myo-
cardial infarction and
50% NPH/50% regular 0.5–2 h 2–5 h Up to 24 h cardiovascular mortality.
Analogue mixtures Arch Intern Med. 2010;
170:1191-1201. [PMID:
75% lispro protamine/25% lispro < 15 min 1–2 h Up to 24 h 20656674]
27. Mahaffey KW, Hafley G,
50% lispro protamine/50% lispro <15 min 1–2 h Up to 24 h Dickerson S, Burns S,
70% aspart protamine/30% 10–20 min 1–4 h Up to 24 h Tourt-Uhlig S, White J,
aspart et al. Results of a reeval-
uation of cardiovascular
Inhaled insulin (Afrezza) 12–15 min 1h 2.5–3 h outcomes in the RECORD
trial. Am Heart J. 2013;
166:240-249.e1. [PMID:
23895806]
28. Home PD, Pocock SJ,
venience or cost for some pa- basal insulin, although rates of Beck-Nielsen H, Curtis
PS, Gomis R, Hanefeld
tients. Sulfonylureas and thiazoli- hypoglycemia were lowest in the M, et al; RECORD Study
dinediones in particular can basal insulin group (30). Most Team. Rosiglitazone
evaluated for cardiovas-
cause hypoglycemia and weight patients have a 1%–2% decrease cular outcomes in oral
agent combination ther-
gain, and there is mixed evidence in HbA1c levels after starting insu- apy for type 2 diabetes
on the cardiovascular safety of lin (31, 32). When intensive glyce- (RECORD): a multicentre,
randomised, open-label
rosiglitazone (26 –28). Patients mic control is planned, a fasting trial. Lancet. 2009;373:
should be warned about these glucose level <6.7 mmol/L (<120 2125-35. [PMID:
19501900]
possibilities and educated to rec- mg/dL) is a reasonable goal. The 29. Meneghini LF. Early
insulin treatment in type
ognize and treat hypoglycemia. primary risks of insulin therapy 2 diabetes: what are the
are hypoglycemia and weight pros? Diabetes Care.
When should physicians 2009;32 Suppl
consider insulin therapy? gain. Patients must be warned 2:S266-9. [PMID:
19875562]
about these possibilities and ed-
Patients who are unable to 30. Holman RR, Farmer AJ,

achieve glycemic control goals ucated to recognize and treat Davies MJ, Levy JC,
Darbyshire JL, Keenan
through oral medications, hypoglycemia. JF, et al; 4-T Study
Group. Three-year effi-
whether alone or in combination, At the start of insulin therapy, cacy of complex insulin
regimens in type 2 dia-
are candidates for insulin ther- most patients can be treated with betes. N Engl J Med.
apy. Other indications include a a once-daily injection. Those
2009;361:1736-47.
[PMID: 19850703]
desire for rapid reduction of without hypoglycemia can often 31. Buse JB, Wolffenbuttel
BH, Herman WH, Shem-
blood glucose in those with se- be treated with a single bedtime onsky NK, Jiang HH,
vere symptoms; some recom- dose of NPH insulin combined
Fahrbach JL, et al. DURA-
bility of basal versus
mend early initiation for persons lispro mix 75/25 insulin
with an oral agent, such as met-
with markedly elevated HbA1c efficacy (DURABLE) trial
formin. In patients with normal 24-week results: safety
levels at diagnosis due to the and efficacy of insulin
fasting glucose levels or those lispro mix 75/25 versus
possibility of prolonging beta-
who are at high risk for hypoglyce- insulin glargine added to
cell function (29). oral antihyperglycemic
mia, a basal analogue (glargine or drugs in patients with
type 2 diabetes. Diabetes
Many formulations of insulin are detemir) may be a preferred first Care. 2009;32:1007-13.
available, separated primarily by choice, although they are consid- [PMID: 19336625]
32. Hayward RA, Manning
their onset of action and duration erably more expensive than NPH. WG, Kaplan SH, Wagner
EH, Greenfield S. Start-
(Table 3). It is not clear that any Typical starting doses of insulin are ing insulin therapy in
particular regimen is superior; in 0.1– 0.2 U/kg. If HbA1c level re- patients with type 2
diabetes: effectiveness,
1 randomized trial, median mains elevated despite normal complications, and re-
HbA1c levels were similar be- fasting glucose levels, prandial source utilization. JAMA.
1997;278:1663-9.
tween biphasic, prandial, and insulin may be considered. [PMID: 9388085]

3 March 2015 Annals of Internal Medicine In the Clinic ITC9 姝 2015 American College of Physicians

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017


33. Yki-Järvinen H, Ryysy L,
Nikkilä K, Tulokas T,
Vanamo R, Heikkilä M.
Comparison of bedtime
insulin regimens in
patients with type 2
diabetes mellitus. A
randomized, controlled
trial. Ann Intern Med.
1999;130:389-96.
[PMID: 10068412]
34. Sheffield CA, Kane MP,
Busch RS, Bakst G, Abels-
eth JM, Hamilton RA.
Safety and efficacy of
exenatide in combina-
tion with insulin in pa-
tients with type 2 diabe-
tes mellitus. Endocr
Pract. 2008;14:285-92.
[PMID: 18463034]
35. Yoon NM, Cavaghan MK,
Brunelle RL, Roach P.
Exenatide added to insu-
lin therapy: a retrospec-
tive review of clinical
practice over two years in
an academic endocrinol-
ogy outpatient setting.
Clin Ther. 2009;31:
1511-23. [PMID:
19695400]
36. Scirica BM, Bhatt DL,
Braunwald E, Steg PG,
Davidson J, Hirshberg B,
et al; SAVOR-TIMI 53
Steering Committee and
Investigators. Saxagliptin
and cardiovascular out-
comes in patients with
type 2 diabetes mellitus.
N Engl J Med. 2013;
369:1317-26. [PMID:
23992601]
姝 2015 American College of Physicians
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017
A combination therapy of insulin
and an oral agent (typically NPH,
lantus, or glargine with met-
formin) can be effective and can
reduce insulin dosing to once
daily at bedtime, which is often
more acceptable to patients (33).
Some patients need twice-daily
insulin to achieve glycemic tar-
gets, but more frequent injec-
tions (such as preprandial injec-
tions) are usually not necessary
for most patients with type 2 dia-
betes. For patients receiving high
doses of insulin, the U-500
(highly concentrated insulin) for-
mulation can be used. Inhaled
insulin (Afrezza) has also been
recently approved by the U.S.
Food and Drug Administration
(FDA).
What other options are
available if control is
inadequate on traditional oral
drugs or insulin?
Newer classes of agents include
glucagon-like peptide-1 (GLP-1)
agonists, dipeptidyl peptidase-IV
(DPP-IV) inhibitors, sodium
glucose-linked transporter-2
(SGLT2) inhibitors, and synthetic
forms of pancreatic hormones
(Table 2).
Exenatide, liraglutide, abligutide,
and dulaglutide are injectable
incretin mimetics, which act
though GLP-1, a naturally occur-
ring hormone involved in glu-
cose homeostasis. They have a
variety of effects, but the most
prominent are enhanced
glucose-dependent insulin secre-
tion and in many patients, de-
layed gastric emptying, suppres-
sion of glucagon, and decreased
appetite and weight loss. These
drugs should be considered for
patients receiving oral agents
who have not achieved glycemic
goals. They can be used in com-
bination with insulin (34, 35). The
primary risks are gastrointestinal
effects, particularly nausea and
vomiting. An extended-release
form of exenatide can be used
ITC10 In the Clinic
for once-weekly injection, and
abligutide and dulaglutide have
both been FDA approved specifi-
cally (and only) for once-weekly
injection. Postmarketing surveil-
lance for longer-term safety as-
sessment of these drugs is
ongoing.
Several DPP-IV inhibitors are now
available, including sitagliptin,
saxagliptin, linagliptin, and
alogliptin. These drugs are orally
administered and work through
the incretin and GLP-1 pathway.
They can be used alone or com-
bined with other oral drugs, par-
ticularly metformin. They rarely
cause hypoglycemia. However,
one clinical trial has raised the
possibility that they increase risk
for hospitalization for heart fail-
ure, although this has only been
investigated in saxagliptin thus
far (36). The drugs do not seem
to have adverse effects on other
cardiovascular outcomes. The
main side effects of sitagliptin are
headache, upper respiratory
symptoms, and less commonly
nausea and diarrhea.
SGLT-2 inhibitors block glucose
transport in the kidney. Canagli-
flozin, dapagliflozin, and empa-
gliflozin are the approved SGLT-2
inhibitors in the United States,
although several others are in
development. These drugs re-
duce HbA1c by about 1% as
monotherapy and effect mild
weight loss. Major adverse
events include increased risk for
genital mycotic infection, eleva-
tion of low-density lipoprotein
(LDL) cholesterol levels, in-
creased urination, and hypoten-
sion (37). Trials are ongoing to
assess cardiovascular safety—
early trials suggested a possible
increase in major cardiovascular
end points (38, 39).
Pramlintide is a synthetic form of
the pancreatic hormone amylin. It
requires preprandial dosing,
which makes it somewhat less
Annals of Internal Medicine 3 March 2015
 convenient than other agents for
many patients with diabetes. It is
typically started at a dose of 60
mg subcutaneously before meals
and can be increased to 120 mg
if tolerated. The insulin dose
should be decreased by 50% in
patients receiving short-acting
insulin to minimize risk for hypo-
glycemia. Frequent monitoring
of blood sugar should be done
before and after meals and at
bedtime. The most common side
effects are nausea and hypogly-
cemia.
What novel therapeutic options
are on the horizon?
Several additional DPP-IV inhibi-
tors are being developed, includ-
ing one (vildagliptin) that has
been approved for use in the Eu-
ropean Union and two (anaglip-
tin and teneligliptin) that have
been approved in Japan. Several
SGLT-2 inhibitors are also in
development.
Aside from glycemic control,
what other clinical
interventions reduce
complications?
Hypertension is a major risk fac-
tor for diabetes complications,
and some evidence suggests that
control of blood pressure is actu-
ally the most important treatment
for patients with diabetes (40).
However, recent evidence shows
that aggressive treatment of
blood pressure to a target of
120/80 mm Hg does not lead to
improved outcomes compared
with a target of 140/90 mm Hg
(10). Most guidelines now reflect
this less aggressive blood pres-
sure goal (4).
Current evidence is not entirely
clear on the optimal choice of
drugs for blood pressure control.
The weight of the evidence pri-
marily supports thiazide diuretics
and one of either angiotensin-
converting enzyme (ACE) inhibi-
tors or angiotensin-receptor
blockers (ARBs) as the initial and
3 March 2015 Annals of Internal Medicine
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017
second drug in patients with dia-
betes and hypertension. Other
agents should be added as
needed to achieve blood pres-
sure goals (40).
Use of lipid-lowering agents is
also a priority in patients with dia-
betes. For primary prevention,
recent guidelines suggest using
a risk-based approach to select-
ing patients for lipid-lowering
therapy; nearly all patients with
diabetes older than 40 years
meet treatment thresholds and
are therefore likely to benefit
from statin therapy, regardless of
initial LDL cholesterol level (11).
Optimal LDL cholesterol targets,
if any, have not been established
in trials; instead, the benefits
seem to be driven primarily by
use of statins. Thus, moderate
doses are recommended from
most patients with type 2 diabe-
tes (41, 42). For secondary pre-
vention, statin use should be en-
couraged in essentially all
patients. Again, ideal LDL choles-
terol targets have not been firmly
established in the literature;
there is some evidence that
higher-dose statins (e.g., simvasta-
tin, 80 mg, or atorvastatin, 80 mg)
may be more effective than lower-
dose statins in patients with exist-
ing coronary artery disease (43,
44). Combination therapy with st-
atins and other agents, particularly
fibrates, does not seem to improve
cardiovascular outcomes in pa-
tients with diabetes (45).
Aspirin therapy is generally rec-
ommended in patients with type
2 diabetes (4), although its bene-
fit in preventing progression of
cardiovascular disease in patients
with type 2 diabetes is unclear. A
recent randomized, controlled
study of aspirin in patients with
type 1 or 2 diabetes found no
evidence to support use in pri-
mary prevention of cardiovascu-
lar events (46). Patients with a
history of heart disease should
In the Clinic
37. Yang XP, Lai D, Zhong
XY, Shen HP, Huang YL.
Efficacy and safety of
canagliflozin in subjects
with type 2 diabetes:
systematic review and
meta-analysis. Eur J Clin
Pharmacol. 2014;70:
1149-58. [PMID:
25124541]
38. Neal B, Perkovic V, de
Zeeuw D, Mahaffey KW,
Fulcher G, Stein P, et al.
Rationale, design, and
baseline characteristics of
the Canagliflozin Cardio-
vascular Assessment
Study (CANVAS)---a ran-
domized placebo-
controlled trial. Am Heart
J. 2013;166:217-
223.e11. [PMID:
23895803]
39. Zinman B, Inzucchi SE,
Lachin J, Wanner C,
Ferrari R, Bluhmki E,
et al. Design of the Em-
pagliflozin Cardiovascu-
lar (CV) Outcome Event
Trial in Type 2 Diabetes
(T2D). Can J Diabetes.
2013;37:S29-S30.
40. Vijan S, Hayward RA.
Treatment of hyperten-
sion in type 2 diabetes
mellitus: blood pressure
goals, choice of agents,
and setting priorities in
diabetes care. Ann Intern
Med. 2003;138:593-
602. [PMID: 12667032]
41. Vijan S, Hayward RA;
American College of
Physicians. Pharmaco-
logic lipid-lowering ther-
apy in type 2 diabetes
mellitus: background
paper for the American
College of Physicians.
Ann Intern Med. 2004;
140:650-8. [PMID:
15096337]
42. Snow V, Aronson MD,
Hornbake ER, Mottur-
Pilson C, Weiss KB; Clini-
cal Efficacy Assessment
Subcommittee of the
American College of
Physicians. Lipid control
in the management of
type 2 diabetes mellitus:
a clinical practice guide-
line from the American
College of Physicians.
Ann Intern Med. 2004;
140:644-9. [PMID:
15096336]
43. LaRosa JC, Grundy SM,
Waters DD, Shear C,
Barter P, Fruchart JC,
et al; Treating to New
Targets (TNT) Investiga-
tors. Intensive lipid low-
ering with atorvastatin in
patients with stable coro-
nary disease. N Engl J
Med. 2005;352:1425-
35. [PMID: 15755765]
ITC11 姝 2015 American College of Physicians

44. Cannon CP, Braunwald
E, McCabe CH, Rader DJ,
Rouleau JL, Belder R,
et al; Pravastatin or Ator-
vastatin Evaluation and
Infection Therapy-
Thrombolysis in Myocar-
dial Infarction 22 Investi-
gators. Intensive versus
moderate lipid lowering
with statins after acute
coronary syndromes. N
Engl J Med. 2004;350:
1495-504. [PMID:
15007110]
45. Ginsberg HN, Elam MB,
Lovato LC, Crouse JR 3rd,
Leiter LA, Linz P, et al;
ACCORD Study Group.
Effects of combination
lipid therapy in type 2
diabetes mellitus. N Engl
J Med. 2010;362:1563-
74. [PMID: 20228404]
46. Belch J, MacCuish A,
Campbell I, Cobbe S,
Taylor R, Prescott R, et al;
Prevention of Progres-
sion of Arterial Disease
and Diabetes Study
Group. The prevention of
progression of arterial
disease and diabetes
(POPADAD) trial: factorial
randomised placebo
controlled trial of aspirin
and antioxidants in pa-
tients with diabetes and
asymptomatic peripheral
arterial disease. BMJ.
2008;337:a1840.
[PMID: 18927173]
47. Vijan S, Hofer TP, Hay-
ward RA. Cost-utility
analysis of screening
intervals for diabetic
retinopathy in patients
with type 2 diabetes
mellitus. JAMA. 2000;
283:889-96. [PMID:
10685713]
48. Brenner BM, Cooper ME,
de Zeeuw D, Keane WF,
Mitch WE, Parving HH,
et al; RENAAL Study
Investigators. Effects of
losartan on renal and
cardiovascular outcomes
in patients with type 2
diabetes and nephropa-
thy. N Engl J Med. 2001;
345:861-9. [PMID:
11565518]
49. Hansson L, Lindholm LH,
Niskanen L, Lanke J,
Hedner T, Niklason A,
et al. Effect of
angiotensin-converting-
enzyme inhibition com-
pared with conventional
therapy on cardiovascular
morbidity and mortality
in hypertension: the
Captopril Prevention
Project (CAPPP) ran-
domised trial. Lancet.
1999;353:611-6. [PMID:
10030325]
姝 2015 American College of Physicians
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017
Table 4. Therapies to Reduce Neuropathy Symptoms
Agent
Tricyclic antidepressants
Duloxetine
Capsaicin cream
Antiepileptic agents
Notes
RCT = randomized, controlled trial.
take between 75 and 325 mg of
aspirin per day.
Retinal examination reduces the
incidence of vision loss in pa-
tients with type 2 diabetes. The
frequency of examination for pa-
tients without high-risk retinal
lesions can range from 1–3 years
depending on underlying risk
(47). Measurement of the urine
microalbumin– creatinine ratio
allows detection of early diabetic
nephropathy; clinical trials have
clearly shown that treatment of
these patients with ACE inhibi-
tors or ARBs clearly reduces risk
for progression to end-stage re-
nal disease (48 –50). Neuropathy
screening and foot care are es-
sential in reducing risk for ampu-
tation. Painful neuropathy is un-
common in type 2 diabetes but
can be treated with a variety of
agents (Table 4).
How frequently should
physician see patients with
type 2 diabetes, and what
should be included in
follow-up visits?
No direct evidence examines the
ideal frequency of visits for pa-
tients with type 2 diabetes. Ex-
pert opinion and the recom-
ITC12 In the Clinic
Notes
RCT evidence shows efficacy
Start at small bedtime dose and titrate to efficacy
Anticholinergic side effects common, use with
particular caution in the elderly
Approved for diabetic neuropathy by the U.S. Food
and Drug Administration
Not appropriate with liver disease or substantial
alcohol use
RCT evidence shows efficacy
Causes burning sensation that often decreases over
time
Carbamazepine, gabapentin, and pregabalin have
RCT evidence of efficacy
There are no data on the relative efficacy of the
medications listed above
Use patient preference for dosing and administration,
along with comorbid conditions and side effect
profiles, to determine initial choice of agent
mended frequency of monitoring
HbA1c levels suggest that quar-
terly visits are reasonable; for pa-
tients with stable disease, this
can be reduced to every 6
months (4).
When should specialists be
consulted?
Meta-analyses show that diabe-
tes education by a certified edu-
cator is effective in improving
many key domains in diabetes
care, including glycemic control,
although the durability of these
effects is not clear.
Endocrinology consultation is
helpful for addressing questions
about diagnosis or when glucose
management has become diffi-
cult (e.g., those with highly labile
blood glucose levels). Patients
who are pregnant or are contem-
plating pregnancy should be re-
ferred for assistance with glucose
control, as poor glucose control
is associated with adverse fetal
outcomes.
Ophthalmologic examination,
whether by ophthalmology, op-
tometry, or through retinal pho-
tography, should be done every
1–3 years depending on prior
Annals of Internal Medicine 3 March 2015
 examination results and level of
glucose control. Other condi-
tions (e.g., known retinopathy,
glaucoma, cataracts) may
necessitate more frequent
examination.
Nephrology evaluation is required
for patients in whom the glomeru-
lar filtration rate has decreased to
<30 ml/min/1.73 m2; earlier refer-
ral can be considered, although
interventions to reduce progres-
sion center around risk factor con-
trol and use of ACE inhibitors or
ARBs as outlined above. Referral
should also be considered if the
origin of renal insufficiency is un-
clear. Patients with hyperkalemia,
acidemia, or difficulty with control-
ling blood pressure may also
benefit.
Podiatry evaluation is helpful for
management of lesions, such as
calluses or deformities, which
require intervention to reduce
risk for foot ulcers and
amputation.
When should patients with
type 2 diabetes be hospitalized?
Some patients with severe, symp-
tomatic hyperglycemia may re-
quire hospitalization, particularly
at the time of diagnosis. Diabetic
ketoacidosis or hyperosmolar
coma requires hospitalization
for management. Diabetes
complications may require hos-
pitalization; for example, celluli-
tis or osteomyelitis may require
intravenous antibiotics or
surgery.

Treatment... The goal of treating type 2 diabetes is to achieve glycemic

targets on an individual basis according to life expectancy and patient
preference. Patients should reach at least moderate levels of control
(HbA1c level <8.0%– 8.5%) to minimize hyperglycemia and because mi-
crovascular risk increases exponentially above this level. More aggres-
sive targets (e.g., <7.0%) should be reserved for patients with a long life
expectancy because reductions in advanced diabetes complications
take 15–20 years to accrue.

CLINICAL BOTTOM LINE

What measures do U.S.
stakeholders use to evaluate
the quality of care for patients
with type 2 diabetes?
The Ambulatory Care Quality Al-
liance recommends several mea-
sures of diabetes care (see the
Box: Quality Measures for Diabe-
tes). Note that these recommen-
dations do not perfectly align
with clinical targets.
What do professional
organizations recommend
regarding the care of patients
with type 2 diabetes?
Several profession associations
publish guidelines for diabetes
Practice Improvement
care. These guidelines do not
always agree on all aspects and
the nature of the organization
inevitably influences its recom-
mendations. Many guidelines for
diabetes can be found at the Na-
tional Guideline Clearinghouse
(www.guidelines.gov). The fol-
lowing organizations are four of
the most commonly cited
50. Parving HH, Lehnert H,
sources.
Bröchner-Mortensen J,
Gomis R, Andersen S,
The American College of Physi- Arner P; Irbesartan in
cians (ACP) conducted system-
Patients with Type 2
Diabetes and Microalbu-
atic reviews of the evidence to minuria Study Group.
The effect of irbesartan
construct guidelines on the man- on the development of
diabetic nephropathy in
agement of hypertension and patients with type 2
lipids in type 2 diabetes (40 – 42, diabetes. N Engl J Med.
2001;345:870-8. [PMID:
51); however, these guidelines 11565519]

3 March 2015 Annals of Internal Medicine In the Clinic ITC13 姝 2015 American College of Physicians

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017

 have not been updated to reflect
the most recent clinical trials. The Quality Measures for Diabetes
ACP also reviewed and rates ex- Eye examination
isting guidelines for glycemic • Percentage of patients who
control and developed an assess- received a retinal or dilated
eye examination by an
ment of the best existing guide- eye care professional
lines on the topic (52). (optometrist or
ophthalmologist) during the
The American Diabetes Associa- reporting year or during the
tion releases diabetes standards prior year if patient is at low
of care yearly. The standards are risk for retinopathy. Patients
are considered low risk if
broad and encompass most rele- they meet all of the following
vant areas of diabetes screening, criteria: not taking insulin;
prevention, and management (4). have an HbA1c measurement
<8.0%; and no evidence of
The American Association of retinopathy in the prior year.
Clinical Endocrinologists last up- HbA1c management
dated their guidelines in 2013. • Percentage of patients who
have had one or more HbA1c
The USPSTF has a draft recom- tests in the measurement
year.
mendation that recommends
HbA1c management control
screening for diabetes every 3
• Percentage of patients
years in adults who are at in- whose most recent HbA1c
creased diabetes risk as outlined level was <8.0% (good
in the Box, Risk Factors for Type 2 control)
Diabetes. These recommenda- • Percentage of patients
whose most recent HbA1c
tions are under review and are level was >9.0% (poor
not finalized, but align with rec- control)
ommendations from other Lipid measurement
groups, such as the American • Percentage of patients who
Diabetes Association. More de- have had at least 1 LDL
tail and finalized recommenda- cholesterol test (or all
component tests)
tions can be accessed at www • LDL cholesterol level
.uspreventiveservicestaskforce • Percentage of patients
.org/uspstf/uspsdiab.htm. whose most recent LDL
cholesterol level was <2.59
mmol/L (100 mg/dL) or <3.7
mmol/L (130 mg/dL)
Blood pressure management
• Percentage of patients with
51. Snow V, Weiss KB, blood pressure documented
Mottur-Pilson C; Clinical
Efficacy Assessment
in the past year as <140/90
Subcommittee of the mm Hg.
American College of Nephropathy screening test
Physicians. The evidence
base for tight blood HB = hemoglobin; LDL =
pressure control in the low-density lipoprotein.
management of type 2
diabetes mellitus. Ann
Intern Med. 2003;138:
587-92. [PMID:
12667031]
52. Qaseem A, Vijan S, Snow
V, Cross JT, Weiss KB,
Owens DK; Clinical Effi-
cacy Assessment Sub-
committee of the Ameri-
can College of
Physicians. Glycemic
control and type 2 diabe-
tes mellitus: the optimal
hemoglobin A1c targets.
A guidance statement
from the American Col-
lege of Physicians. Ann
Intern Med. 2007;147:
417-22. [PMID:
17876024]

姝 2015 American College of Physicians ITC14 In the Clinic Annals of Internal Medicine 3 March 2015

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017

 In the Clinic Smart Medicine Module
http://smartmedicine.acponline.org/index.aspx

Tool Kit
Access the Smart Medicine module on type 2 diabetes.

Clinical Guidelines
www.acponline.org/clinical_information/guidelines.
Guidelines from the American College of Physicians.
https://www.aace.com/files/algorithm-07-11-2013.pdf
Guidelines from the American Association of Clinical
Type 2 Diabetes Endocrinologists.
http://care.diabetesjournals.org/content/37
/Supplement_1/S14.extract
Guidelines from the American Diabetes Association;
updated yearly.
https://www.nice.org.uk/guidance/cg87/chapter
/guidance
Guidelines from The National Institute for Health and
Clinical Excellence for the United Kingdom's National
Health Service; updated in 2009.

Patient Information
http://diabetes.niddk.nih.gov/dm/pubs/type2_ES
/index.htm (English) http://diabetes.niddk.nih.gov
/dm/pubs/type2_ES/index.htm#Spanish (Spanish)
Type 2 Diabetes: What You Need to Know. Information

IntheClinic
for patients by the National Institute of Diabetes and
Digestive and Kidney Diseases in both English and
Spanish.
www.nlm.nih.gov/medlineplus/diabetes.html
MEDLINE Plus information about diabetes for patients,
including an interactive tutorial available in both
English and Spanish.
Recent Evidence
http://diabetes.acponline.org/
American College of Physicians' Diabetes Monthly update
from ACP Internist for information on the latest studies
in type 2 diabetes.
Diagnostic Tests and Criteria
http://smartmedicine.acponline.org/content
.aspx?gbosId=203
List of screening and diagnostic tests for diabetes mellitus
from the American College of Physicians.
Quality Measures
www.qualitymeasures.ahrq.gov/index.aspx
Information from the National Quality Measures Clear-
inghouse relating to diabetes.
www.qualityforum.org/Publications/2002/10
/National_Voluntary_Consensus_Standards_for
_Adult_Diabetes_Care.aspx
Quality measures related to diabetes from the National
Quality Forum.

3 March 2015 Annals of Internal Medicine In the Clinic ITC15 姝 2015 American College of Physicians

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017

 WHAT YOU SHOULD In the Clinic
Annals of Internal Medicine
KNOW ABOUT TYPE 2
DIABETES
What Is Type 2 Diabetes?
Diabetes is a condition where there is too much
sugar (glucose) in your blood. Sugar can build
up because your body doesn't make enough of
a hormone called “insulin”. Diabetes can hap-
pen if you don't have enough insulin to turn the
sugar into energy. It also may happen if your
body doesn't respond to the insulin it does have.
Most people with diabetes make at least some insu-
lin, but it doesn't work to keep the blood sugar
under control. This is called type 2 diabetes. When
type 2 diabetes is not controlled, it can cause
sugar to build up. If the sugar stays high, it can
slowly damage the heart, kidneys, nerves, eyes,
and feet. It is very important to keep type 2 dia-
betes under control to prevent complications.
Complications from uncontrolled diabetes can
include:
• Vision loss
• Kidney damage
• Nerve damage
• Foot ulcers
• Heart disease
• Possible amputation from infections
• There are many different types of medicine for
What Are the Warning Signs? type 2 diabetes, including insulin. Not all
• Extreme thirst and/or hunger people with type 2 diabetes will need insulin.
• Fatigue Talk to your doctor about what treatment plan
• Frequent need to urinate is best for you.

• Unusual weight loss
• Blurred vision
• Tingling or numbness in hands or feet
• Frequent infections
• Bruises that are slow to heal
How Is It Diagnosed?
Diabetes is diagnosed by testing the level of glu-
cose, or sugar, in your blood. Two or more tests
might be used to diagnose diabetes. You may
need to fast before some diabetes tests. This
means you will not have any food or drink (ex-
cept water) for several hours before your blood
is tested. Other tests might require you to drink
a special drink before your blood is tested. Talk
to your doctor about how you should prepare
for your diabetes test.
How Is It Treated?
People with diabetes need to improve sugar (glu-
cose) control in their bodies.
• Sometimes, lifestyle changes such as eating
healthy, losing weight, or exercising regularly
can help improve glucose control.
• If lifestyle changes don't improve glucose
control, your doctor may prescribe medicines.
Patient Information
What Can I Expect From Doctor
Visits?
• Your doctor will talk to you about your blood
sugar levels and the results of any blood tests.
• Your doctor will also ask about your diet and
exercise.
• Diabetes can cause nerve damage, which may
feel like tingling or burning pain. Your doctor
will check your feet for injuries and loss of
feeling during visits.
• Your doctor will also check your blood
pressure, cholesterol levels, and kidney
function.
• You will also need an eye examination to
check for any problems.
Questions for My Doctor
• Will I have to use insulin?
• Do I have to check my blood sugar?
• How can I check my feet at home?
• Will I need to lose weight? If so, what is the
best way to do this?
• How often should I make follow up visits?
• Are there any other tests I need?
• Do I need any shots (vaccines)?

For More Information

ACP:
https://www.acponline.org/cgi-bin/cpph
.cgi?CPP3003_Living_With_Diabetes_2014.pdf
www.acponline.org/patients_families/products/health_tips
/diab_en.pdf
American Diabetes Association:
www.diabetes.org

Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/933271/ by a University of California San Diego User on 04/27/2017