ORIGINAL

ARTICLES

Risk of Obstructive Sleep Apnea in

Patients With Type 2 Diabetes Mellitus
Alvah R. Cass, MD, SM; W. Jerome Alonso, MD, CCFP, DABFM; Jamal Islam, MD, MS;
Susan C. Weller, PhD

BACKGROUND AND OBJECTIVES: Type 2 diabetes mellitus (DM) approximately 11.3% of the US pop-
and obstructive sleep apnea (OSA) share several clinical findings: ulation over the age of 20 and 26.9%
obesity, hypertension, and impaired glucose tolerance. OSA may be of the population age 65 years and
an under-recognized comorbidity of DM. The purpose of this study older.1 The prevalence of obstructive
is to estimate the proportion of patients with type 2 DM at risk for sleep apnea may also increase with
OSA and describe factors associated with that risk.
age.2,3 In the Sleep Heart Health
METHODS: This cross-sectional study enrolled 297 adults, ages Study, the prevalence of obstruc-
18 years and older, with type 2 DM from a university-based Fam- tive sleep apnea was two to eight
ily Medicine Center. Participants completed a research question- times higher among participants
naire recording sociodemographic information, medical history, and ages 60 years and older compared
clinical data including medications and hemoglobin A1C. OSA risk to those ages 2560 years.4 Ethnic
was determined using the Berlin Questionnaire. Relationships be- minorities may also be at increased
tween risk of OSA and sociodemographic and clinical variables risk for obstructive sleep apnea.5-8
were evaluated using bivariate analyses and covariate adjusted Obesity is common to diabetes and
logistic regression models. obstructive sleep apnea; however, up-
RESULTS: Thirty-seven participants (12.5%) had a prior diagnosis
per body obesity is more character-
of OSA. Based on the Berlin Questionnaire, 124 (48.6%) of the istic of individuals with obstructive
remaining patients were classified as high risk for OSA. Patients sleep apnea.9
less than age 60 years were at increased risk for OSA, adjust- Type 2 diabetes is a major cause
ed OR=2.67 (1.57, 4.54; 95% CI). Non-Hispanic whites, adjusted of morbidity and mortality linked
OR=1.76 (1.01, 3.06; 95% CI), and patients with symptoms of to microvascular and macrovascu-
depression, adjusted OR=2.64 (1.60, 4.52; 95% CI), were also at lar complications and is considered
higher risk. Gender and hemoglobin A1C were not associated with as a coronary artery disease risk
increased risk of OSA. equivalent for myocardial infarc-
tion.10 Recent studies indicate that
CONCLUSIONS: Nearly half of adults with type 2 DM may be at obstructive sleep apnea is associated
high risk for OSA, and many may be undiagnosed. In a primary
with atherosclerosis and initiation
care setting, the Berlin Questionnaire is an easily applied screen-
and progression of cardiovascular
ing instrument that identifies patients at increased risk of OSA who
disease.11-13 Obstructive sleep apnea
may benefit from further diagnostic studies and treatment of OSA.
is also associated with other risks
(Fam Med 2013;45(7):492-500.) for cardiovascular disease, includ-
ing hypertension,14-16 obesity,17 and

I
dentification and management
of modifiable risk factors such as
obesity, hypertension, and dys-
lipidemia are important to reduce
complications associated with type
2 diabetes mellitus. Sleep-relat-
ed breathing disorders, including ob-
structive sleep apnea (OSA), may be
another comorbidity that deserves
attention in the overall approach to
the care of patients with type 2 dia-
betes mellitus.
Type 2 diabetes mellitus and
obstructive sleep apnea share a
number of common risk factors, in-
cluding advancing age, race/eth-
nicity, and obesity. Based on 2010
estimates, diabetes mellitus affected
dyslipidemia18 and may interact syn-
ergistically with these risk factors
From the University of Texas Medical Branch,
Galveston, TX (Drs Cass and Weller); Canadian
Sleep Consultants (Dr Alonso); and Texas Tech
University Health Sciences Center (Dr Islam).

492 JULY-AUGUST 2013 VOL. 45, NO. 7 FAMILY MEDICINE


to increase morbidity and mortality
related to cardiovascular events.13,19
Investigators20-24 also have ob-
served an association between
obstructive sleep apnea, insulin re-
sistance, and impaired glucose toler-
ance. The association persists after
controlling for other risk factors such
as obesity and duration of disease.25
Further support for a link between
obstructive sleep apnea and diabetes
is the observation that insulin resis-
tance decreases in individuals with
obstructive sleep apnea treated with
continuous positive airway pressure
(CPAP).26-28
Postulating that obstructive sleep
apnea is more prevalent in patients
with diabetes than in the general
population seems a logical conclu-
sion. A limited number of studies
have supported such a relation-
ship;29-33 however, these studies are
not representative of primary care
patients. The studies by Shim,32
Lam,31 and Katsumata33 are reports
of Korean, Chinese, and Japanese
patients enrolled predominantly
through specialty endocrine clinics,
and the Katsumata study includ-
ed only men. Laaban30 reported on
French patients hospitalized with
uncontrolled diabetes. The findings
described by Einhorn34 were from
patients referred to a diabetic clinic
in California. Only the United King-
dom study by West and colleagues29
included patients from general prac-
tices; however, the majority were re-
cruited from hospital registries, all
were male, and most were Cauca-
sian. Studies from primary care are
notably lacking.
We planned this study to address
the paucity of data on the risk of
obstructive sleep apnea in primary
care patients with type 2 diabetes
and designed the study to estimate
the proportion of patients with type
2 diabetes mellitus at risk for ob-
structive sleep apnea in a primary
care setting. We also addressed other
limitations in previously published
reports by enrolling a diverse popu-
lation of men and women from three
major racial and ethnic groups and
across a wide age range. Further, we
FAMILY MEDICINE VOL. 45, NO. 7 JULY-AUGUST 2013
report on a broad range of modifiable
and non-modifiable factors potential-
ly associated with risk for obstruc-
tive sleep apnea in patients with
type 2 diabetes mellitus.
Methods
Study Population
This is a cross-sectional study of pa-
tients with type 2 diabetes mellitus
attending a university-based Fam-
ily Medicine Center serving an eth-
nically diverse population. Patients,
ages 25 years and older, with type 2
diabetes mellitus were enrolled from
February 2007 to September 2008.
The Institutional Review Board for
Research Involving Human Subjects
approved this study. All participants
provided written informed consent
prior to participation.
Procedures
Interviewers enrolled patients at the
time of a regular visit. After obtain-
ing informed consent, participants
were asked about demographic data,
medical history, including duration
of diabetes, current treatment, pri-
or diagnosis of hypertension and
medications, and associated comor-
bidities and diabetic complications.
Interviewers administered the Berlin
Questionnaire35 to assess risk for ob-
structive sleep apnea and the Center
for Epidemiologic Studies Depression
Scale (CES-D).36 The most recent
HbA1C value and blood pressure
reading were obtained from medi-
cal records. Patients height, weight,
and neck and waist circumferences
were recorded.
Instruments
The Berlin Questionnaire,35 an out-
come of the Conference on Sleep in
Primary Care held in Berlin, Ger-
many in 1996, was used to estimate
the risk of obstructive sleep apnea,
the primary outcome for this study.
Eleven items cover three domains
related to risk of obstructive sleep
apnea: (1) snoring and sleep-related
symptoms, (2) awake-time sleepiness
and drowsiness, and (3) hyperten-
sion and/or a BMI greater than
30. Responses are coded present or
ORIGINAL ARTICLES
absent for each category, and the fi-
nal score ranges from 0 to 3. Subjects
who score 2 are classified as high
risk, whereas subjects who score <2
are classified as low risk. The scale
has good reliability and validity, a
sensitivity of 0.86, a specificity of
0.77, and a positive predictive val-
ue of 0.89 for detecting a Respira-
tory Disturbance Index >5,35 which
is consistent with obstructive sleep
apnea.37,38
The Center for Epidemiologic
Studies Depression Scale is a self-
reported scale designed to measure
depressive symptomatology in the
general population.36 The scale con-
tains 20 items addressing various
aspects of depression including emo-
tional and somatic symptoms. The
CES-D demonstrated good reliabili-
ty (Cronbachs a=0.85) and has been
validated in a variety of settings.39
Radloffs36 original work showed that
a score 16 (range 060) discriminat-
ed psychiatric inpatients from the
general population and that higher
CES-D scores were associated with
more clinically severe depression
among inpatients.
Sample Size
Sample size was estimated based on
a conservative estimate of the pro-
portion of persons at high risk for
obstructive sleep apnea of 0.50. A
sample of 300 patients with type 2
diabetes mellitus would provide an
estimate of the proportion of patients
at high risk for sleep apnea, defined
by the Berlin Questionnaire, with a
95% confidence interval of 0.057
(6%).
Data Analysis
The Berlin Questionnaire35 provided
an estimate of the proportion of type
2 diabetics at high risk of obstructive
sleep apnea (composite score 2).
Reliability was assessed using the
reliability coefficient, Cronbachs a.40
Modifiable (control of diabetes,
hypertension, and medications used
to treat diabetes and hypertension)
and non-modifiable (age, gender,
race/ethnicity, duration of diabetes,
and associated comorbid diseases)
493
 ORIGINAL ARTICLES
factors associated with risk of ob-
structive sleep apnea were evalu-
ated using bivariate analyses and
logistic regression. We evaluated the
relationship between the Berlin risk
categories (high versus low) and cat-
egorical variables (gender, ethnici-
ty, and medical and disease-related
variables) with the x2 test or Fish-
ers exact test and interval scaled
variables (age, duration of diabetes,
and CES-D) with Students t test.
We used a logistic regression model
to predict high risk for obstructive
sleep apnea adjusted for the com-
bined effects of modifiable and non-
modifiable covariates and to assess
the effect of risk for obstructive sleep
apnea on glycemic control controlling
for age, gender, race/ethnicity, dura-
tion of disease, treatment, and BMI.
All data were analyzed using PASW
Statistics 18 (Version 18.0.0, 2009,
IBM SPSS, Inc, Chicago).
Results
We enrolled 297 patients in this
study. One patient may have had
type 1 diabetes and was excluded.
The mean age of the participants
(n=296) was 59.7 11.6 years. Men
comprised 39.9% of the study popu-
lation. Non-Hispanic whites made
up 37.5% of the sample, 34.8% were
African-American, and 26.0% were
Hispanic. The average duration of
diabetes was 10.1 8.5 years. Over
half of patients (57.4%) were treat-
ed with two or more hypoglycemic
agents and 21.6% of participants
used insulin, either alone or in com-
bination with oral agents. The mean
hemoglobin A1C was 8.1 2.1 (range
5.216.5). Thirty-seven participants
(12.5%) indicated a prior diagnosis
of obstructive sleep apnea, and 22
of these individuals (59.5%) reported
use of a CPAP machine. Table 1 sum-
marizes the sociodemographics and
medical profile of the primary study
population and compares those with
and without a prior diagnosis of ob-
structive sleep apnea.
Excluding patients with a prior
diagnosis of obstructive sleep ap-
nea (n=37) and patients with miss-
ing data on the Berlin Questionnaire
494 JULY-AUGUST 2013 VOL. 45, NO. 7 FAMILY MEDICINE
(n=4), the Berlin Questionnaire clas-
sified 48.6% (42.3, 54.9, 95% CI;
n=124) as high risk for obstructive
sleep apnea. (See Figure 1.) The re-
liability for the 11 items contained
in the Berlin Questionnaire was ac-
ceptable, yielding a Cronbachs of
0.79.39 The reliability for each of the
Berlin Questionnaire component
scales 1 and 2 was similar to previ-
ously reported values.35 Cronbachs
for Category 1 (five items), the
snoring component, was 0.89 and
for Category 2 (four items), or the
sleepiness component, Cronbachs
was 0.67.
Further analysis of the high-risk
patients found that 68.5% (n=85)
described excessive snoring, 64.5%
(n=80) noted excessive daytime
sleepiness, and 96.8% (n=120) re-
ported a prior diagnosis of hyper-
tension or had a BMI greater than
30. Only 3.2% (n=4) of patients were
classified as high risk for obstruc-
tive sleep apnea solely based on the
combination of snoring and sleepi-
ness. Combining snoring or sleep-
iness with hypertension or BMI
accounted for 66.9% (n=83) of the
high-risk group. All three risk fac-
tors were present in 29.8% (n=37) of
high-risk patients. Only 6.7% (n=17)
of the patients without a prior diag-
nosis of obstructive sleep apnea were
negative for all three of the Berlin
Questionnaire risk factors.
In bivariate analysis, patients at
increased risk for obstructive sleep
apnea tended to be younger, 57.2
11.3 years versus 62.9 11.4 years
(P value <.001) compared to the low-
risk group. Non-Hispanic whites
and more highly educated individu-
als tended to be in the Berlin high-
risk group; however, neither achieved
statistical significance. There were
no significant differences in mean
systolic or diastolic blood pressures
between the high-risk and low-risk
groups. With regard to diabetes mel-
litus, we found no significant differ-
ences in duration of disease, mean
hemoglobin A1C levels, use of oral
hypoglycemic agents, use of insulin,
or the number of medications used
to lower blood sugar between the
Berlin Questionnaire risk groups.
Table 2 summarizes the results of
the bivariate analyses comparing
high- and low-risk patients.
Three variables from the logistic
regression analysis, which is sum-
marized in Table 3, predicted risk
for obstructive sleep. Age less than
60 years was associated with an in-
creased risk for obstructive sleep ap-
nea (OR=2.67; 95% CI=1.57, 4.54).
Non-Hispanic whites were classified
as high risk (OR=1.76; 95% CI =1.01,
3.06) compared to other racial/ethnic
groups and patients with symptoms
of depression, modeled as a CES-D
score of 16 or more, were more like-
ly classified as high risk (OR=2.64;
95% CI=1.60, 4.52). Gender, hemo-
globin A1C, duration of diabetes, and
intensity of treatment of diabetes,
including use of insulin, were not
significant predictors of risk for ob-
structive sleep apnea.
In a second logistic regression
model, predicting glycemic control
as a function of risk for obstructive
sleep apnea, patients at increased
risk of obstructive sleep apnea were
more likely to have a hemoglobin
A1C 7.0 compared to low-risk pa-
tients; however, this difference did
not achieve statistical significance
(OR=1.37, 95% CI=0.728, 2.58; ad-
justed for age, gender, race/ethnicity,
duration of disease, BMI, and inten-
sity of treatment).
Comment
Using the Berlin Questionnaire35
to categorize the risk of obstructive
sleep apnea, we found a higher risk
in patients with type 2 diabetes mel-
litus than in the general population.
Three findings from this study of pa-
tients with type 2 diabetes differed
from previous reports of the general
population. In contrast to other stud-
ies, we found a decrease in risk with
advancing age, no difference in risk
based on gender, and an increased
risk in non-Hispanic whites com-
pared to other racial ethnic groups.
We also noted an association of more
severe depressive symptomatology
and risk of obstructive sleep apnea.
Finally, we found no significant
 ORIGINAL ARTICLES

Table 1: Comparison of Patients With Prior Diagnosis of Obstructive Sleep

Apnea (OSA) to the Patients at Risk for the Total Sample

Prior Diagnosis No Prior Diagnosis

Total Sample
OSA OSA P value
(n=296)
(n=37) (n=259)
Age in years (mean SD) 59.7 11.6 56.6 10.1 60.1 11.7 .083
Gender (% male) 39.9 45.9 39.0 .474
Race (%)
Non-Hispanic White 37.5 51.4 35.5
African-American 34.8 27.0 35.9 .270
Hispanic 26.0 21.6 26.6
Education (% high school graduate) 76.7 81.1 76.1 .678
Marital status (% living with partner) 46.6 51.4 45.9 .599
BMI (mean SD) 34.3 8.4 43.0 10.5 33.1 7.3 <.001
Male 32.9 7.2 39.1 8.8 31.8 6.3 .004
Female 35.3 9.0 46.4 11.0 33.9 7.7 <.001
Neck circumference (mean SD in 15.9 1.8 17.2 1.8 15.7 1.7 <.001
inches)
Male 17.0 1.6 17.9 1.7 16.8 1.5 .014
Female 15.2 1.5 16.6 1.7 15.0 1.4 <.001
Waist circumference (mean SD in 43.0 6.9 49.3 7.6 42.0 6.3 <.001
inches)
Male 42.4 6.4 47.7 7.6 41.6 5.7 <.001
Female 43.3 7.2 50.8 7.4 42.3 6.6 <.001
History of hypertension (%) 76.7 89.2 74.9 .062
Systolic BP (mean SD) 131 17 128 16.0 131 18.0 .228
Diastolic BP (mean SD) 75 11 74 10 75 11 .529
Antihypertensive therapy (% 2 meds) 47.3 56.8 45.9 .224
Duration of diabetes (mean SD) 10.1 8.5 7.9 6.4 10.4 8.7 .095
Glycemic control
Hemoglobin A1C (mean SD) 8.0 2.1 7.4 1.5 8.1 2.1 .101
Hemoglobin A1C (% < 7.0) 38.5 51.6 38.5 .175
Diabetic therapy (% 2 meds) 42.4 37.8 43.1 .597
Comorbidity
Coronary artery disease 19.9 37.8 17.4 .007
Restless leg syndrome 7.1 27.0 4.2 <.001
Depression 29.7 43.2 27.8 .082
Retinopathy 7.1 8.1 6.9 .735
CES-D (mean SD) 18.6 7.2 22.6 9.1 18.0 6.8 <.001
Distressed (% CES-D 16) 60.0 83.3 56.8 .002

FAMILY MEDICINE VOL. 45, NO. 7 JULY-AUGUST 2013 495

 ORIGINAL ARTICLES

Table 2: Comparison of Berlin Questionnaire Risk Classification for Sleep Apnea

in the Study Population Without a Prior Diagnosis of Obstructive Sleep Apnea

Berlin Berlin
High Risk Low Risk Odds Ratio
Variable n=124 n=131 (95% CI) P Value
Age (mean SD) 57.4 11.1 62.8 11.6 < .001
Gender (% male) 36.3 40.5 0.84 (0.51, 1.39) .521
Race/ethnicity (%)
Non-Hispanic white 42.9 29.8 Ref
African-American 29.8 41.2 0.52 (0.29, 0.95) .087
Mexican-American 25.0 29.0 0.62 (0.33, 1.17)
Education (% high school graduate) 79.8 72.5 1.50 (0.84, 2.69) .188
Marital status (% living with 50.0 42.7 1.34 (0.82, 2.19) .260
partner)
BMI (mean SD) 35.8 7.2 30.4 6.0 NR NR
Male (mean SD) 33.3 6.4 30.8 6.2 NR NR
Female (mean SD) 37.2 7.4 30.2 5.9 NR NR
Neck circumference 16.2 1.7 15.3 1.6 NR NR
Male (mean SD in inches) 17.3 1.5 16.5 1.5 NR NR
Female (mean SD in inches) 15.5 1.4 14.5 1.2 NR NR
Waist circumference 43.8 6.7 40.3 5.3 NR NR
Male (mean SD in inches) 42.3 6.0 40.8 5.5 NR NR
Female (mean SD in inches) 44.6 7.0 40.0 5.2 NR NR
Systolic BP (mean SD) 132 17 131 18 NR NR
Diastolic BP (mean SD) 75 11 76 11 NR NR
Duration of diabetes mellitus (mean 10.1 8.4 10.7 8.9 --- .589
SD)
Hemoglobin A1C (mean SD) 8.1 2.1 7.9 2.2 --- .492
Hemoglobin A1C (% <7.0) 35.0 41.7 0.75 (0.43, 1.30) .331
Level of control (%)
<7.0 35.0 41.7 Ref
7.07.9 21.4 23.5 1.09 (0.53, 2.21) .395
8.0 43.7 34.8 1.50 (0.82, 2.75)
Treatment of diabetes (%)
No medications 8.1 11.5 0.68 (0.29, 1.57) .405
Metformin 65.3 57.3 1.41 (0.85, 2.33) .200
Sulfonylurea 41.9 42.0 1.00 (0.61, 1.64) >.999
Thiazolidinediones 16.1 9.2 1.91 (0.89, 4.09) .129
Other 4.8 5.3 0.90 (0.29, 2.76) >.999
Insulin 24.2 19.1 1.35 (0.74, 2.46) .362
Two or more medications 46.8 40.5 1.29 (0.79, 2.12) .316

(continued on next page)

496 JULY-AUGUST 2013 VOL. 45, NO. 7 FAMILY MEDICINE

 ORIGINAL ARTICLES

Table 2: Continued

Berlin Berlin
High Risk Low Risk Odds Ratio
Variable n=124 n=131 (95% CI) P Value
Treatment of hypertension (%)
No medications 15.3 31.3 0.40 (0.22, 0.73) .003
ACE inhibitor 59.7 50.4 1.46 (0.89, 2.39) .166
b blocker 30.6 21.4 1.62 (0.92, 2.86) .115
Diuretic 29.8 30.5 0.97 (0.57, 1.65) >.999
Calcium channel blocker 12.9 13.0 0.99 (0.48, 2.06) >.999
ARB 11.3 9.2 1.26 (0.56, 2.85) .680
Potassium sparing diuretic 4.0 2.3 1.79 (0.42, 7.66) .490
Peripheral a blocker 1.6 0.8 2.13 (0.19, 23.80) .613
Central a blocker 0.8 0.8 1.06 (0.06, 17.08) >.999
Other antihypertensive 3.2 1.5 2.15 (0.39, 12.00) .436
Two or more medications 49.2 43.5 1.26 (0.77, 2.06) .381
Comorbidity (%)
Hypertension 83.1 66.4 NR NR
Depression 34.7 21.4 1.95 (1.12, 3.41) .025
GERD 20.2 16.8 1.25 (0.66, 2.36) .521
Coronary artery disease 15.3 19.1 0.77 (0.40, 1.48) .508
Asthma 10.5 5.3 2.08 (0.80, 5.38) .163
Insomnia 9.7 5.3 1.90 (0.72, 4.99) .235
Retinopathy 8.1 6.1 1.35 (0.51, 3.54) .628
Restless leg 6.5 2.3 2.94 (0.76, 11.4) .128
Exercise 3 or more days a week (%) 37.1 43.5 0.77 (0.46, 1.26) .310
CES-D (mean SD) 19.8 7.1 16.4 6.1 <.001
Distressed (% CES-D 16) 71.0 45.0 2.98 (1.78, 5.01) <.001

ARBangiotension receptor blocker

GERDgastroesophageal reflux disease
CES-D Center for Epidemiologic Studies Depression Scale
SDstandard deviation
NR = Results included for informational purposes. Inferential statistics are not reported for these variables because the variable is either used in
the Berlin Questionnaire to define risk (hypertension and body mass index) or is directly associated with variables used to define risk.

association between risk of obstruc-
tive sleep apnea and glycemic con-
trol. These findings are important
because they suggest that diabetic
patients may not conform to the risk
profile for obstructive sleep apnea
seen in the general population.
In our study, 48.6% of patients
(n=124) without a prior diagnosis of
obstructive sleep apnea were classi-
fied as high risk for obstructive sleep
apnea. Netzer et al,41 using the Ber-
lin Questionnaire in a multinational
study to quantify risk of obstructive
sleep apnea in primary care, report-
ed that 35.8% of the US sample was
at high risk for obstructive sleep ap-
nea. Grover and colleagues42 found
a similar risk of obstructive sleep
apnea (33%) using the Berlin Ques-
tionnaire in a consecutive sample of
249 primary care patients undergo-
ing preventive examinations.
Approximately 89% of the high-
risk patients would be expected to
meet diagnostic criteria for obstruc-
tive sleep apnea.35 Combining this
proportion of the patients at high
risk (n=110) with the 37 patients
who had a prior diagnosis of sleep
apnea, we estimated a prevalence
of obstructive sleep apnea in our
study population of approximately
50%. This estimate is notably greater
than the prevalence estimates in the
landmark study of a random sam-
ple of employed men and women,
ages 3060 years, by Young and col-
leagues.43 They estimated a preva-
lence of sleep-disordered breathing
of 24% for middle-aged men and 9%
for middle-aged women, defined as
5 apnea-hypopnea events per hour.
Requiring excessive daytime sleepi-
ness and 5 apnea-hypopnea events
per hour, they estimated that 4% of
men and 2% of women met the mini-
mal criteria to diagnose obstructive

FAMILY MEDICINE VOL. 45, NO. 7 JULY-AUGUST 2013 497

 ORIGINAL ARTICLES
FigureI:EnrollmentFlowChart
Figure 1: Enrollment Flow Chart
Type2diabtesmellitus
n=296
PriordiagnosisofOSA
n=37
CPAP NoCPAP StudyPopula;on
n=22 n=15
HighriskforOSA
n=124
OSAobstructive sleep apnea
CPAPcontinuous positive airway pressure
sleep apnea. The increased risk for
obstructive sleep apnea in the cur-
rent study may be partially attrib-
utable to the prevalence of obesity
in the study population. Vgontzas
et al44 reported as many as 50% of
obese patients had obstructive sleep
apnea. More recently, Young and col-
leagues45 affirmed an increased prev-
alence of obstructive sleep apnea
RunningTitle:RiskofObstructiveSleepApnea
(30%60%) in obese patients.
Our findings, from a primary
care population, are consistent with
some of the findings from studies
of adults with type 2 diabetes from
a highly selected population. West
and colleagues,29 in a study of mainly
Caucasian men with type 2 diabe-
tes from Great Britain, found that
57% of the men scored in the high-
risk range using the Berlin Ques-
tionnaire.35 Similarly, Einhorn et al34
reported characteristics of 330 con-
secutive adults with type 2 diabe-
tes, referred to a diabetes clinic, who
were enrolled in a prevalence study
of sleep apnea. A total of 279 enroll-
ees completed the study. Using an
apnea-hypopnea index of 10 events
per hour during a home-based sleep
498 JULY-AUGUST 2013 VOL. 45, NO. 7 FAMILY MEDICINE
Enrolled
n=297
Excluded
Possibletype1
diabetesmellitus
n=1
NopriordiagnosisOSA
n=259
Excluded
Missingdata(Berlin)
n=255
n=4
LowriskforOSA
n=131
study, 48% of the participants met
criteria and were diagnosed with
sleep apnea.
We did not find an increased risk
of obstructive sleep apnea with ad-
vancing age. Several epidemiologic
studies found an increase in sleep-
disordered breathing and obstruc-
tive sleep apnea with advancing
age.2,4,34,46,47 Others report a peak in
Page 14
the prevalence of obstructive sleep
apnea around age 60 years and a
decline thereafter,48,49 which is sim-
ilar to the distribution of risk we
observed. A direct comparison is dif-
ficult, because we only assessed risk
of obstructive sleep apnea, where-
as the other studies reported prev-
alence of sleep apnea. Risk, more
so than prevalence, may be influ-
enced by an age-related decline in
the number of sleep partners avail-
able to report on snoring and other
characteristics of sleep-disordered
breathing; however, we found no as-
sociation between marital status and
risk of obstructive sleep apnea.
In contrast to other studies, we
found no difference in risk of obstruc-
tive sleep apnea related to gender. In
other studies, men are two to three
times as likely as women to have ob-
structive sleep apnea.43,47 Several of
the prevalence studies that included
women are older studies and may
reflect a less obese population. The
BMI for women in our study was 6%
higher than for men. As a result, a
greater proportion of women (63.7%)
met the BMI criteria for risk than
did men (57.1%). The greater ten-
dency toward obesity may explain
partially the lack of a gender differ-
ence in risk observed in this study.
A menopause effect may also explain
the lack of a gender difference seen
in our study. Bixler50 reported a male
to female ratio of 1:1.4 in menopaus-
al age groups compared to 1:3.3 in
pre-menopausal age groups. Tishler51
also noted a 1:1 male to female ratio
in older individuals.
Race/ethnicity, in a multivariable
logistic regression model, was pre-
dictive of risk for obstructive sleep
apnea. Non-Hispanic whites were
at greater risk than other racial
ethnic groups, which differs from
other epidemiologic studies where
African-Americans were reported at
increased risk. In our study, this dif-
ference could not be explained by a
skewed distribution of gender, age,
or weight-based factors.
The difference in diabetic control
observed between risk groups in this
study is consistent with the meta-
bolic evidence that obstructive sleep
apnea is associated with insulin re-
sistance and impaired glucose toler-
ance.20,21,24,28 Individuals at increased
risk of obstructive sleep apnea had
higher hemoblobin A1C levels com-
pared to individuals at low risk (8.1
versus 7.9), were less likely to have
a hemoglobin A1C level less than 7.0
(35% versus 42%), and more likely to
have a hemoglobin A1C level greater
than 8.0 (44% versus 35%); howev-
er, these differences were not statis-
tically significant. Our sample size
may have been insufficient to detect
a small but perhaps meaningful dif-
ference in glycosylated hemoglobin
levels.
Our study had several strengths.
The study population represented
 ORIGINAL ARTICLES

Table 3: Multivariable Logistic Regresion Models Predicting Risk of Obstructive Sleep Apnea Based on the Berlin Score

Coefficient Standard Wald

Variable (b) Error (2) P Value Odds Ratio 95% CI
Initial logistic model (n=255)
Intercept -1.58 0.41 14.54 <.001 0.21
Age (<60 years) 0.89 0.29 9.01 .003 2.42 1.364.31
Gender (male) -0.22 0.30 0.531 .466 0.81 0.451.44
Race (Non-Hispanic white) 0.573 0.29 3.82 .051 1.76 1.003.11
Education (high school graduate) -0.34 0.34 0.962 .327 0.72 0.371.40
Married/living with partner 0.34 0.29 1.42 .233 1.41 0.802.49
Duration of DM (years) 0.01 0.02 0.13 .721 1.01 0.971.04
Hemoglobin A1C (7.0) 0.30 0.31 0.97 .324 1.35 0.742.46
Oral hypoglycemics (2 meds) 0.08 0.30 0.07 .786 1.08 0.601.95
Insulin 0.01 0.36 0.00 .984 1.01 0.502.04
CES-D Score (16 points) 1.05 0.29 13.30 <.001 2.86 1.625.02
Final logistic model (n=255)
Intercept -1.28 0.27 23.05 <.001 0.28
Age (<60 years) 0.98 0.27 13.07 <.001 2.67 1.574.54
Race (non-Hispanic white) 0.56 0.28 3.98 .046 1.76 1.013.06
CES-D Score (16 points) 0.97 0.27 12.59 <.001 2.64 1.544.52

CIconfidence interval
DMdiabetes mellitus
CES-DCenter for Epidemiologic Studies Depression Scale

three major racial/ethnic groups
across a wide age range, and more
than half of the participants were
women. We were able to collect clin-
ically meaningful data on duration
of diabetes, glycemic control, treat-
ment of diabetes and hypertension,
and symptoms of depression.
There were also limitations. As
with all cross-sectional studies,
cause and effect relationships must
be interpreted with caution. With-
in the scope of this project, we were
not able to confirm obstructive sleep
apnea with a formal sleep study and
therefore could not assess the oper-
ating characteristics of the Berlin
Questionnaire.35 However, 26 of 37
patients with a prior diagnosis of
obstructive sleep apnea scored in
the high-risk range on the Berlin
Questionnaire, which yields a posi-
tive predictive value of 72.2%. This
value is somewhat lower than previ-
ously reported;35 however, more than
to improve symptoms in obstructive
sleep apnea. As a result one would
expect lower scores on the Berlin
Questionnaire.
This study adds to the growing
body of evidence that type 2 diabe-
tes mellitus is associated with an
increased risk for obstructive sleep
apnea; however, diabetic patients
may not fit the typical risk profile.
Our study also lends support for
more aggressive screening for ob-
structive sleep apnea in patients
with type 2 diabetes. The Berlin
Questionnaire is a simple screening
instrument that easily can be used
and quickly scored in a busy clinical
practice. A lack of awareness among
primary care practitioners regarding
risk and prevalence of obstructive
sleep apnea in patients with diabe-
tes may translate into missed op-
portunities to improve patient care
and outcomes through diagnosis and
treatment of obstructive sleep apnea.
with type 2 diabetes may be an im-
portant step in improving outcomes.
ACKNOWLEDGMENTS: Funding in support
of this research was provided by an American
Academy of Family Physicians Resident Re-
search Award (G0606PF: Alonso/Islam/Weller)
and by a Texas Academy of Family Physicians
Research Award (Islam/Alonso/Cass).
The results of the research described in this
manuscript were presented in part at the 2010
North American Primary Care Research Group
Annual Meeting in Seattle, WA.
CORRESPONDING AUTHOR: Address cor-
respondence to Dr Cass, University of Texas
Medical Branch, 301 University Blvd., Galves-
ton, TX 77555-1123. 409-772-0620. acass@
utmb.edu.
References
1. Centers for Disease Control and Prevention.
National diabetes fact sheet, 2011: National
estimates and general information on diabetes
and pre-diabetes in the United States. Atlanta:
US Department of Health and Human Servic-
es, Centers for Disease Control and Prevention,
2011.
2. Phillips B, Cook Y, Schmitt F, Berry D. Sleep
apnea: prevalence of risk factors in a general
population. South Med 1989;82(9):1090-2.

60% of these patients were using a Improved recognition and treatment

CPAP apparatus, which is known of obstructive sleep apnea in patients

FAMILY MEDICINE VOL. 45, NO. 7 JULY-AUGUST 2013 499

 ORIGINAL ARTICLES
3. Dealberto MJ, Ferber C, Garma L, Lemoine
P, Alperovitch A. Factors related to sleep ap-
nea syndrome in sleep clinic patients. Chest
1994;105(6):1753-8.
4. Young T, Shahar E, Nieto FJ, et al. Predictors
of sleep-disordered breathing in community-
dwelling adults: The Sleep Heart Health Study.
Arch Intern Med 2002;162(8):893-900.
5. Grunstein RR, Lawrence S, Spies JM, et al.
Snoring in paradise: The Western Samoa Sleep
Survey. Eur Respir J 1989;2(Suppl 5):401-5.
6. Schmidt-Nowara WW, Coultas DB, Wiggins C,
Skipper BE, Samet JM. Snoring in a Hispanic-
American population. Risk factors and associa-
tion with hypertension and other morbidity.
Arch Intern Med 1990;150(3):597-601.
7. Kripke DF, Ancoli-Israel S, Klauber MR,
Wingard DL, Mason WJ, Mullaney DJ. Preva-
lence of sleep-disordered breathing in ages
4064 years: a population-based survey. Sleep
1997;20(1):65-76.
8. Redline S, Neubauer J. Beyond the fat boy: the
intertwining pathophysiologies of sleep apnea.
J Appl Physiol 2005;99(4):1243-4.
9. Millman RP, Carlisle CC, McGarvey ST, Evel-
off SE, Levinson PD. Body fat distribution
and sleep apnea severity in women. Chest
1995;107(2):362-6.
10. Haffner SM, Lehto S, Rnnemaa T, Pyrl
K, Laakso M. Mortality from coronary heart
disease in subjects with type 2 diabetes
and non-diabetic subjects with and without
prior myocardial infarction. N Engl J Med
1998;339(4):229-34.
11. Suzuki T, Nakano H, Maekawa J, et al. Ob-
structive sleep apnea and carotid-artery inti-
ma-media thickness. Sleep 2004;27(1):129-33.
12. Drager LF, Bortolotto LA, Lorenzi MC,
Figueiredo AC, Krieger EM, Lorenzi-Filho
G. Early signs of atherosclerosis in obstruc-
tive sleep apnea. Am J Respir Crit Care Med
2005;172(5):613-8.
13. Young T, Finn L, Peppard PE, et al. Sleep dis-
ordered breathing and mortality: eighteen-year
follow-up of the Wisconsin Sleep Cohort. Sleep
2008;31(8):1071-8.
14. Hedner J, Bengtsson-Bostrom K, Peker Y,
Grote L, Rastam L, Lindblad U. Hypertension
prevalence in obstructive sleep apnoea and
sex: a population-based case-control study. Eur
Respir J 2006;27(3):564-70.
15. Peppard PE, Young T, Palta M, Skatrud J.
Prospective study of the association between
sleep-disordered breathing and hypertension.
N Engl J Med 2000;342(19):1378-84.
16. Gottlieb DJ, Redline S, Nieto FJ, et al. Asso-
ciation of usual sleep duration with hyper-
tension: the Sleep Heart Health Study. Sleep
2006;29(8):1009-14.
17. Schafer H, Pauleit D, Sudhop T, Gouni-Ber-
thold I, Ewig S, Berthold HK. Body fat distri-
bution, serum leptin, and cardiovascular risk
factors in men with obstructive sleep apnea.
Chest 2002;122(3):829-39.
18. Borgel J, Sanner BM, Bittlinsky A, et al. Ob-
structive sleep apnoea and its therapy influ-
ence high-density lipoprotein cholesterol serum
levels. Eur Respir J 2006;27(1):121-7.
500 JULY-AUGUST 2013 VOL. 45, NO. 7 FAMILY MEDICINE
19. Vgontzas AN, Bixler EO, Chrousos GP. Sleep
apnea is a manifestation of the metabolic syn-
drome. Sleep Med Rev 2005;9(3):211-24.
20. Ip MSM, Lam B, Ng MMT, Lam WK, Tsang
KWT, Lam KSL. Obstructive sleep apnea is in-
dependently associated with insulin resistance.
Am J Respir Crit Care Med 2002;165(5):670-6.
21. Punjabi NM, Shahar E, Redline S. Sleep-dis-
ordered breathing, glucose intolerance, and
insulin resistance (The Sleep Heart Health
Study). Am J Epidemiol 2004;160(6):521-30.
22. Punjabi NM, Polotsky VY. Disorders of glucose
metabolism in sleep apnea. J Appl Physiol
2005;99(5):1998-2007.
23. Reichmuth K, Austin D, Skatrud J, Young T.
Association of sleep apnea and type II diabetes:
a population-based study. Am J Respir Crit
Care Med 2005;172(12):1590-5.
24. Yaggi HK, Araujo AB, McKinlay JB. Sleep
duration as a risk factor for the development of
type 2 diabetes. Diabetes Care 2006;29(3):657-
61.
25. Vgontzas AN, Bixler EO, Chrousos GP. Meta-
bolic disturbances in obesity versus sleep ap-
noea: the importance of visceral obesity and in-
sulin resistance. J Intern Med 2003;254(1):32-
44.
26. Harsch IA, Schahin SP, Bruckner K, et al. The
effect of continuous positive airway pressure
treatment on insulin sensitivity in patients
with obstructive sleep apnoea syndrome and
type 2 diabetes. Respiration 2004;71(3):252-9.
27. Harsch IA, Schahin SP, Radespiel-Troger M,
et al. Continuous positive airway pressure
treatment rapidly improves insulin sensi-
tivity in patients with obstructive sleep ap-
nea syndrome. Am J Respir Crit Care Med
2004;169(2):156-62.
28. Babu AR, Herdegen J, Fogelfeld L, Shott S,
Mazzone T. Type 2 diabetes, glycemic control,
and continuous positive airway pressure in
obstructive sleep apnea. Arch Intern Med
2005;165(4):447-52.
29. West SD, Nicoll DJ, Stradling JR. Prevalence
of obstructive sleep apnea in men with type 2
diabetes. Thorax 2006;61(11):945-50.
30. Laaban JP, Daenen S, Lger D, et al. Preva-
lence and predicitive factors of sleep apnoea
syndrome in type 2 diabetic patients. Diabetes
Metab 2009;35:372-7.
31. Lam DC, Lui MMS, Lam JCM, Ong LHY,
Lam KSL, Ip MSM. Prevalence and recog-
nition of obstructive sleep apnea in Chinese
patients with type 2 diabetes mellitus. Chest
2010;138(5):1101-7.
32. Shim U, Lee H, Oh J-Y, Sung Y-A. Sleep dis-
order and cardiovascular risk factors among
patients with type 2 diabetes mellitus. Korean
J Intern Med 2011;26:277-84.
33. Katsumata K, Tamatsu O, Miyao M, Katsuma-
ta Y. High incidence of sleep apnea syndrome
in a male diabetic population. Diabetes Res
Clin Pract 1991;13:45-52.
34. Einhorn D, Stewart D, Erman M, Gordon N,
Philis-Tsimikas A, Casal E. Prevalence of sleep
apnea in a population of adults with type 2
diabetes mellitus. Endocr Prac 2007;13(4):355-
62.
35. Netzer N, Stoohls R, Netzer C, Clark K, Strohl
K. Using the Berlin Questionnaire to identify
patients at risk for the sleep apnea syndrome.
Ann Intern Med 1999;131(7):485-91.
36. Radloff L. The CES-D scale: a self-report de-
pression scale for research in general popula-
tion. Appl Psychol Meas 1997;1:385-401.
37. American Academy of Sleep Medicine Task
Force. Sleep-related breathing disorders in
adults: recommendations for syndrome defin-
tion and measurement techniques in clincal
sleep research. Sleep 1999;22:667-90.
38. American Academy of Sleep Medicine. Inter-
national Classification of Sleep Disorders, Re-
vised: diagnostic and coding manual. Chicago:
American Academy of Sleep Medicine, 2001:1-
401.
39. Nunnally JC. Psychometric theory, second edi-
tion. New York: McGraw-Hill, 1978.
40. Cronbach LJ. Coefficient alpha and the in-
ternal structure of tests. Psychometrika
1951;16(3):297-334.
41. Netzer NC, Hoegel JJ, Loube D, et al. Preva-
lence of symptoms and risk of sleep apnea in
primary care. Chest 2003;124(4):1406-14.
42. Grover M, Mookadam M, Armas D, et al. Iden-
tifying patients at risk for obstructive sleep
apnea in a primary care practice. J Am Board
Fam Med 2011;24:152-60.
43. Young T, Palta M, Dempsey J, Skatrud J, Weber
S, Badr S. The occurrence of sleep-disordered
breathing among middle-aged adults. N Engl
J Med 1993;328(17):1230-5.
44. Vgontzas A, Tan T, Bixler E, Martin L, Shu-
bert D, Kales A. Sleep apnea and sleep dis-
ruption in obese patients. Arch Intern Med
1994;154:1705-11.
45. Young T, Peppard PE, Taheri S. Excess weight
and sleep-disordered breathing. J Appl Physiol
2005;99(4):1592-9.
46. Duran J, Esnalo S, Rubio R, Izutueta A. Ob-
structive sleep apnea-hypopnea and related
clinical features in a population-based sample
of subjects aged 30 to 70 years. Am J Respir
Crit Care Med 2001;163:685-9.
47. Jennum P, Riha RL. Epidemiology of sleep ap-
noea/hyponoea syndrome and sleep-disordered
breathing. Eur Respir J 2009;33(4):646-55.
48. Gislason T, Almqvist M, Eriksson G, Taube
A, Boman G. Prevalence of sleep apnea syn-
drome among Swedish menan epidemiologi-
cal study. J Clin Epidemiol 1988;41(6):571-6.
49. Bixler EO, Vgontzas AN, Have TT, Tyson K,
Kales A. Effects of age on sleep apnea in men:
prevalence and severity. Am J Respir Crit Care
Med 1998;157:144-8.
50. Bixler EO, Vgontzas AN, Lin H-M, et al. Preva-
lence of sleep-disordered breathing in women:
effects of gender. Am J Respir Crit Care Med
2011;163:608-13.
51. Tishler PV, Larkin EK, Schluchter MD, Red-
line S. Incidence of sleep-disordered breath-
ing in an urban adult population: the relative
importance of risk factors in the develop-
ment of sleep-disordered breathing. JAMA
2003;289(17):2230-7.