Awareness of Prediabetes and

Engagement in Diabetes
RiskReducing Behaviors
Anjali Gopalan, MD,1,2 Ilona S. Lorincz, MD,3 Christopher Wirtalla, BA,5 Steven C. Marcus, PhD,6,7
Judith A. Long, MD1,4,7
Introduction: Studies have demonstrated the benet of weight loss and physical activity for diabetes
prevention among those with prediabetes. Despite this evidence, only about half of people with
prediabetes report engaging in these behaviors. One presumed barrier is low patient awareness of
prediabetes. The purpose of this study is to examine the impact of prediabetes awareness on the odds
of engagement in diabetes riskreduction behaviors.
Methods: A pooled cross-sectional analysis of adults from two cycles (20072008, 20092010) of
the National Health and Nutrition Examination Survey was conducted. Those with prediabetes were
identied by excluding people with self-reported diabetes and then screening for hemoglobin A1c
values between 5.7% and 6.4%. This group was then divided based on self-reported prediabetes.
Multivariate logistic regression was used to estimate the effect of prediabetes awareness on the odds
of engagement in physical activity, weight management, and the combination of physical activity
and weight management.
Results: Of those meeting the dened criteria for prediabetes (n2,694), only 11.8% (n288) were
aware of their status. Prediabetes-aware individuals had higher odds of engagement in the
combination of moderate physical activity plus BMI-appropriate weight management (AOR1.5,
95% CI1.1, 2.0), and the combination of at least 150 minutes/week of moderate activity and 7%
weight loss in the past year (AOR2.4, 95% CI1.1, 5.6).

Conclusions: Prediabetes-aware adults have increased odds of engagement in physical activity and
weight management. Increasing patients awareness of prediabetes could result in increased
performance of exercise and weight management behaviors and, most importantly, decreased risk
of future diabetes.
(Am J Prev Med 2015;49(4):512519) Published by Elsevier Inc. on behalf of American Journal of Preventive
Medicine

Introduction

From the 1Philadelphia Veterans Affairs Medical Center, Philadelphia,

Pennsylvania; 2Robert Wood Johnson Clinical Scholars Program, University of Pennsylvania, Philadelphia, Pennsylvania; 3Division of Endocrinology, Diabetes, and Metabolism, University of Pennsylvania,
Philadelphia, Pennsylvania; 4Division of General Internal Medicine, University of Pennsylvania, Philadelphia, Pennsylvania; 5Perelman School of
Medicine, University of Pennsylvania, Philadelphia, Pennsylvania; 6School
of Social Policy and Practice, University of Pennsylvania, Philadelphia,
Pennsylvania; and the 7Philadelphia Veterans Affairs Center for Health
Equity Research and Promotion, Philadelphia, Pennsylvania
Address correspondence to: Anjali Gopalan, MD, Blockley Hall, 423
Guardian Drive, Philadelphia PA 19104. E-mail: agopalan@upenn.edu.
0749-3797/$36.00
http://dx.doi.org/10.1016/j.amepre.2015.03.007

512 Am J Prev Med 2015;49(4):512519

pproximately 86 million adults, or 37% of the

U.S. adult population, meet the diagnostic criteria for prediabetes.1 Every year, 11% of people
with prediabetes go on to develop diabetes.2 Multiple
studies25 have demonstrated the benet of lifestyle
modication, specically modest weight loss and
increased physical activity, in reducing the incidence of
diabetes. The largest study was the Diabetes Prevention
Program (DPP), an RCT that randomized overweight
individuals into one of three arms: placebo, metformin,
or lifestyle intervention. The lifestyle arm, which had an
intervention goal of at least 150 minutes/week of
moderate physical activity and a goal weight loss of

Published by Elsevier Inc. on behalf of American Journal of Preventive Medicine

Gopalan et al / Am J Prev Med 2015;49(4):512519

Z7% body weight, demonstrated a 58% decrease in the

incidence of diabetes over 3 years and a 34% reduction at
10-year follow-up.2,6 Importantly, these results have also
been replicated in diabetes prevention programs in
healthcare and community settings.712
Despite this evidence, only half of those with prediabetes report engaging in risk-reduction behaviors.13
Additionally, it is unknown how many of these individuals achieve the targets shown to be of benet by the
DPP. One potential barrier to higher engagement rates is
that awareness of prediabetes, although increasing,
remains low, with only 11% of adults aware of their
diagnosis.14 Willingness to perform a health behavior
depends on the perceived need for action; thus, patients
who are unaware of their diagnosis may lack the impetus
to engage in risk-reduction behaviors.15 It is currently
unknown whether awareness of prediabetes is associated
with the increased performance of healthy behaviors.
In this study, nationally representative data are used to
(1) assess differences in demographics, health care, and
intermediate health outcomes between individuals who
are aware and unaware of having prediabetes and (2) to
examine the association between awareness of prediabetes and the performance of diabetes riskreducing
behaviors, specically physical activity and weight
management.

Methods
A pooled cross-sectional analysis from two consecutive cycles
(20072008 and 20092010) of the National Health and Nutrition
Examination Survey (NHANES) was conducted to investigate
whether adults with prediabetes aware of their diagnosis were
more likely to report engaging in diabetes riskreducing behaviors
than adults who were unaware of their diagnosis.

Survey Design and Population

NHANES is conducted by the National Center for Health Statistics
and CDC.16 NHANES is a biannual, stratied, multistage probability sample of the U.S. civilian non-institutionalized population.
Survey participants are interviewed at home and invited to attend a
mobile examination center to undergo a medical examination and
laboratory measurements.

Measures
Between 2007 and 2010, 20,686 individuals participated in the
NHANES survey. Subjects aged o20 years (n8,533) and
pregnant women (n125) were excluded (Figure 1). Nonphysiologic hemoglobin A1c values (HbA1c o3.5%) were recoded
to missing (n1). Participants missing the HbA1c value (n1,036)
or responses to questions regarding weight history (n452) were
excluded.
The remaining participants were then classied by glycemic
status. Participants were asked if they had been told by a doctor or
other health professional that they had diabetes (other than during
October 2015

513

Figure 1. Applying exclusion criteria and identifying those in

an analytic sample.
NHANES, National Health and Nutrition Examination Survey.

pregnancy). Based on this question, 1,258 respondents were

classifed as having been diagnosed with diabetes. HbA1c diagnostic criteria were then used to identify those in the prediabetes
range (HbA1c5.7%, 6.4%, n2,694). The remaining individuals
were classied as undiagnosed diabetes (HbA1c Z6.5%, n359)
or were normoglycemic by HbA1c value (n6,229). Only those
meeting the HbA1c criteria for prediabetes were included in the
analytic sample.
The prediabetes group (n2,694) was further divided based on
their responses to two questions: (1) Other than during pregnancy,
have you ever been told by a doctor or health professional that you
have diabetes or sugar diabetes? (2) Have you ever been told by a
doctor or other health professional that you have any of the
following: prediabetes, impaired fasting glucose, impaired glucose
tolerance, borderline diabetes, or that blood sugar is higher than
normal but not high enough to be called diabetes or sugar
diabetes? Individuals were classied as being prediabetes aware if
they responded to the rst question with a self-initiated answer of
borderline diabetes (n98) or if they answered yes to the second
question (n190), for a total prediabetes-aware population of 288.
Outcomes for three major activity groups included physical
activity, weight-related activity, and the combination of weightrelated and physical activity. Within each of these major groups,
three binary outcome variables were created, which progressed
from a basic level of engagement in the activity to more intense
and specic engagement.
For the physical activity outcomes, we created three outcome
variables: (1) any physical activity; (2) at least moderate (comprised of moderate or vigorous) weekly physical activity; and (3)
Z150 minutes/week of at least moderate physical activity. Any
physical activity was dened as stating participation in any of the
following: vigorous or moderate work activity, vigorous or
moderate recreational activity, or any type of transit-related
activity (e.g., light, moderate, or vigorous walking or biking to

Gopalan et al / Am J Prev Med 2015;49(4):512519

514

Table 1. Population Demographic and Clinical Characteristics by Prediabetes Awareness

Prediabetes aware (n288)

Prediabetes unaware (n2,406)

M (SD) or %

M (SD) or %

p-value

Age, years (n; M [SD])

288

57.6 (1.0)

2,406

55.3 (0.5)

0.02

Gender (% female)

149

54.2

1,193

51.9

0.6

White

153

72.8

1,079

65.8

0.1

Black

54

11.0

547

14.6

Hispanic

67

8.3

665

12.8

Other

14

7.9

115

6.9

Less than high school

79

17.2

794

23.9

High school or GED

72

28.3

613

27.7

Some college/associates

73

24.0

599

27.2

At least college graduate

63

30.4

396

21.0

49

12.6

409

11.8

67

15.7

656

21.6

4200

149

65.2

1,096

57.7

Missing

23

6.5

245

8.8

Private insurance

89

41.9

846

45.4

Uninsured

39

12.4

538

18.7

Medicaid

17

4.5

126

3.8

Medicare/governmental

143

41.3

893

32.1

Place of regular care (% yes)

274

95.0

2,084

87.6

0.02

None

18

5.9

349

13.6

o0.001

13

107

38.5

1,099

46.7

49

110

39.2

642

27.4

410

53

16.4

315

12.3

Family history of diabetes

145

46.3

932

38.3

0.05

Activity limitation

131

39.7

687

25.8

o0.001

PHQ-9 score (M [SD])

264

3.7 (0.5)

2,157

3.0 (0.1)

0.1

BMI (M [SD])

288

30.9 (0.5)

2,402

29.7 (0.1)

0.01

History of CV disease (% yes)

238

84.1

1,428

70.5

No. of CV conditions, max 6 (M [SD])

288

1.4 (0.1)

2,406

Systolic BP, mmHg (M [SD])

277

126 (1.2)

2,291

Characteristics

Race/ethnicity (%)

Education
0.04

Income, % of povertyincome ratio

o100
100200

0.2

Insurance status
0.03

Healthcare visits within last year

o0.01

0.9 (0.03)
126 (0.7)

o0.001
1.0

(continued on next page)

www.ajpmonline.org

Gopalan et al / Am J Prev Med 2015;49(4):512519

515

Table 1. Population Demographic and Clinical Characteristics by Prediabetes Awareness (continued)

Prediabetes aware (n288)

Prediabetes unaware (n2,406)

M (SD) or %

M (SD) or %

p-value

Diastolic BP, mmHg (M [SD])

277

70.4 (0.7)

2,291

71.3 (0.5)

0.2

LDL cholesterol, mg/dL (M [SD])

148

115 (3.4)

1,126

124 (1.1)

0.01

HbA1c, % (M [SD])

288

6.0 (0.01)

2,406

5.9 (0.01)

o0.001

Characteristics

Note: Boldface indicates signicant difference between prediabetes aware and unaware at po0.05.
BP, blood pressure; CV, cardiovascular disease; GED, general educational development; HbA1c, hemoglobin A1c; LDL, low-density lipoprotein; PHQ,
Patient Health Questionnaire.

work). At least moderate physical activity included vigorous or

moderate work activity and vigorous or moderate recreational
activity (i.e., standard exercise) but excluded transit-related activity
because the intensity of this type of physical activity was not
assessed. The nal physical activity category, Z150 minutes/week
of at least moderate physical activity, was created to determine the
percentage of people achieving the DPP intervention target
quantity of exercise. To create this variable, the amount of time
for moderate and vigorous work activity and moderate or vigorous
recreational activity were summed to create a binary variable.
Values for weekly exercise that were considered unrealistic, 47
days/week of activity or 48 hours/day of any individual activity
type, were recoded to maximum values of 7 days/week and 8
hours/day (n46).
The weight-related outcomes were similarly structured and
included (1) any weight management behavior; (2) BMIappropriate weight management; and (3) successful weight loss of
Z7% of body weight in the past year. Any weight-related behavior
was a based on responses to questions about intention to lose
weight or maintain weight during the past year. BMI-appropriate
weight management behavior was dened by a combination of
measured BMI and stated intention to lose weight or maintain
weight during the past year. Individuals were dened as engaging in
BMI-appropriate weight management if they had a BMI o25 and
reported weight maintenance efforts or a BMI Z25 and efforts to
lose weight. The nal weight management variable, successful
intentional weight loss of Z7% of their body weight during the past
year, referenced the intervention goal percentage of weight loss in
the DPP. This variable was created from the combination of three
other variables: intention to lose weight, measured current weight,
and self-reported weight from 1 year before.
The last category of outcome variables was created based on
combining the outcome variables described above for physical
activity and weight-related behaviors. The three combination
variables were as follows: (1) any physical activity plus any
weight-related behavior; (2) weekly moderate or vigorous physical
activity plus BMI-appropriate weight behavior; and (3) Z150
minutes/week of moderate or vigorous activity plus Z7% weight
loss in past year. The more-specic categories were nested within
the more broad outcome categories. For example, if an individual
had a mismatch in levels of performance of the two behaviors
(e.g., moderate physical activity, but only any weight-related behavior), they would be included in the broadest combined outcome
category, any physical activity, plus any weight-related behavior.
Covariates examined in the analysis included demographic
factors (gender, age, race/ethnicity, and education level); insurance
October 2015

status; place of regular care; number of healthcare visits within the

past year; family history of diabetes; and a count of other
cardiovascular comorbidities (hypertension, hyperlipidemia,
stroke, heart attack, coronary artery disease, and congestive heart
failure). Limited functional status was a binary variable based on
responding yes to one of several questions regarding limitations on
working, needing special equipment for ambulation, memory
problems, and emotional problems. The Patient Health
Questionnaire-9 (PHQ-9) was used to calculate a depression
severity score for all participants.17 BMI; seated blood pressure
(averaged over one to four readings); and low-density lipoprotein
cholesterol were obtained from examination and laboratory data.
Details on the collection of NHANES data are available online.16

Statistical Analyses
A 4-year survey weight (exam weight) for the merged samples
was created per instructions on the NHANES website.16 All
analyses accounted for the complex survey design, including
weights, strata, and primary sampling units. Chi-square tests and
t tests were performed to determine whether prediabetes-aware
and -unaware groups differed with respect to their demographic
and clinical characteristics. Multivariate logistic regressions were
run to model the association of prediabetes awareness with the
three outcome variables in each of the three outcome categories
(nine total outcomes: the three physical activity, three weight
management, and three combined activity and weight management outcomes). The regression models were adjusted for the
following characteristics: gender, age, race/ethnicity, educational
attainment, number of healthcare visits within the past year, a
family history of diabetes, BMI, number of cardiovascular
conditions, PHQ-9 score, and functional limitations. These
covariates were chosen a priori, and their inclusion was based
on their hypothesized and documented relationships to both
prediabetes awareness and engagement in risk-reducing behaviors.13 Stata, version 12.1, for Mac was used for data management
and analysis.

Results
A total of 2,694 participants met criteria for prediabetes.
Of this group, 11.8% (n288) were aware of this
diagnosis (Figure 1). Those who were aware of prediabetes differed signicantly from those who were unaware
by age (57.6 vs 55.3 years, p0.02) and educational

516

Gopalan et al / Am J Prev Med 2015;49(4):512519

attainment (p0.04), but did not differ based on gender,

income, or ethnicity (Table 1). Participants aware of
prediabetes were less likely to have had no healthcare
visits in the past year (5.9% vs 13.6%, po0.001) and had
a higher mean number of cardiovascular conditions (1.4
[SE0.1] vs 0.9 [SE0.03], po0.001). Those aware of
prediabetes had a slightly higher mean BMI (30.9

[SE0.5] vs 29.7 [SE0.1], p0.01) as well as a slightly

higher mean HbA1c value (6.0% [SE0.01%] vs 5.9%
[SE0.01%], po0.001).
There were no differences between the prediabetesaware and -unaware groups in any of the three physical
activity outcomes (as dened previously) in the unadjusted chi-square analyses (Figure 2A). The vast majority
of individuals in both groups reported engaging in some
physical activity weekly (74.2% of those who were aware
of prediabetes and 72.5% of those who were unaware),
whereas fewer reported meeting the goal of Z150
minutes/week of at least moderate activity (54.6% and
51.2%, respectively). There were signicant differences in
engagement in the weight management behaviors and
the combined activity weight management outcomes.
Those aware of having prediabetes were more likely to
engage in any weight management (65.9% vs 49.3%,
po0.001) and BMI-appropriate weight management
(54.9% vs 38.8%, po0.001) (Figure 2B). There was no
difference in the prevalence of weight loss of Z7%
between the groups (22.4% vs 15.6%, p0.08). Those
who were aware of prediabetes were more likely to report
any physical activity plus any weight management
(49.7% vs 39.0%, p0.01), and at least moderate activity
plus BMI-appropriate weight management (37.2% vs
27.5%, p0.001). The prediabetes-aware and -unaware
groups also differed in achievement of DPP intervention
targets for physical activity and weight loss (9.1% vs 4.6%,
p0.02) (Figure 2C).
In the multivariate logistic regression models, no
signicant differences were seen in the odds of engagement in any of the three physical activity outcomes
(Table 2). Those aware of having prediabetes had higher
odds of engaging in any weight-related behavior
(AOR1.8, 95% CI1.2, 2.7) and in BMI-appropriate
weight management (AOR1.7, 95% CI1.1, 2.5). Prediabetes awareness was also associated with increased
odds of all three of the combination outcomes: any
physical activity plus any weight management (AOR1.5,
95% CI1.0, 2.1); at least moderate activity plus BMIappropriate weight behavior (AOR1.5, 95% CI1.1,
2.0); and Z150 minutes/week of physical activity plus
Z7% weight loss (AOR2.4, 95% CI1.1, 5.6).

Discussion
Figure 2. Physical activity, weight-related behavior, and the
combination of physical activity and weight-related behavior
by prediabetes awareness. 2a. Physical activity by prediabetes awareness. 2b. Weight management by prediabetes
awareness. 2c. Physical activity and weight management by
prediabetes awareness.
Note: Asterisk indicates signicant difference between prediabetes
aware and unaware (*po0.05).

In this analysis of a large nationally representative sample

performed between 2007 and 2010, only 11.8% of
patients with HbA1c-diagnosed prediabetes were aware
of their diagnosis. Regardless of this knowledge, engagement in both moderate or vigorous physical activity and
BMI-appropriate weight management was uncommon in
our study. Awareness of prediabetes was associated with
www.ajpmonline.org

Gopalan et al / Am J Prev Med 2015;49(4):512519

517

behaviors: changing diet, exercise, or attempting to lose

weight. Our ndings demonOutcomes
AOR (95% CI)
p-value
strate that although more than
half of subjects reported
Physical activity
engaging in some level
Any physical activity
1.3 (1.0, 1.8)
0.06
of physical activity and
At least moderate activity
1.3 (1.0, 1.7)
0.07
weight-loss efforts, in both
the prediabetes-aware and
Z150 minutes/week of at least moderate activity
1.3 (0.9, 1.9)
0.2
-unaware groups, only about
Weight management
one third of respondents
Any weight-related behavior
1.8 (1.2, 2.7)
0.01
reported engaging in the combination of at least moderate
BMI-appropriate weight-related behavior
1.7 (1.1, 2.5)
0.01
physical activity and BMILost Z7% body weight in the past year
1.8 (1.1, 3.2)
0.03
appropriate weight managePhysical activity weight management
ment. Even fewer reported
achieving the goals of at least
Any physical any weight-related behavior
1.5 (1.0, 2.1)
0.04
150 minutes/week of moderAt least moderate activity BMI-appropriate weight behavior
1.5 (1.1, 2.0)
0.02
ate exercise along with at least
ExerciseZ150 minutes/week Z7% weight loss
2.4 (1.1, 5.6)
0.04
7% weight loss. The low rates
of adequate engagement highNote: Boldface indicates signicant difference between prediabetes aware and unaware (po0.05).
All of the above models adjusted for gender, age, race/ethnicity, educational attainment, number of
light the inadequacy of physihealthcare visits within the past year, a family history of diabetes, BMI, number of cardiovascular
cian advice alone to support
conditions, Patient Health Questionnaire (PHQ)-9 score, and functional limitations.
these
difcult
behavior
changes and the need for greater patient participation
an increased likelihood of engaging in the combination of
in structured programs. Numerous interventions have
physical activity and weight management efforts, but
demonstrated that the DPP can be translated effectively
very few individuals reported achieving the targets
into real-world healthcare and community settings. A
advocated by the DPP.
2011 systematic review of 16 studies in distinct settings
The impact of prediabetes awareness on the odds of
showed that the percentage of patients achieving at least
engagement in physical activity and weight management
7% weight loss ranged from 18% to 49% at a follow-up of
highlights the special inuence physician advice and
610 months.8 The most successful programs offered a
input can have on patient behaviors. This nding is
built-in structure with case management support, freconsistent with prior research. In an RCT of adults in a
quent contact, supervised activity sessions, and, imporprimary care setting, individuals randomized to receiving
tantly, individualized plans. In healthcare settings, inphysician advice on quitting smoking, reducing fat
person and online health coaches, group lifestyle proconsumption, and increasing exercise were more likely
grams, and telehealth programs have been used to deliver
to believe these topics were relevant to them and more
the DPP intervention.812 Highlighting the importance of
likely to report attempting to quit smoking and making
healthcare professionals in promoting diabetes risk
some dietary changes.18 Based on the low levels of
reduction behavior, the lowest attrition and greatest
prediabetes awareness, it seems that many patients with
weight loss were seen in DPP programs conducted in
prediabetes are not receiving, or not recalling, this
hospital outpatient clinics and primary care settings.8
information. To maximize the impact of physician
Efforts in the primary care setting are especially critical,
discussions about prediabetes on reducing the incidence
as they have the potential to reach more diverse
of diabetes, physicians must not only communicate this
populations of patients than hospital clinics, the patients
diagnosis to patients but must do so in a meaningful way.
are already established in the practice, and other medical
Physicians should counsel patients that having prediaproblems can be managed concurrently as healthy
betes greatly increases the chances of developing diabetes
behaviors are instituted and supported.
and has been linked to increased risk of cardiovascular
disease.1921 The benet of lifestyle modication in
Limitations
reducing the incidence of diabetes must also be stressed.
12,22
In previous studies,
more than 50% of adults with
This analysis has several limitations. First, the crossprediabetes reported performing some risk-reduction
sectional nature of the data limits our ability to accurately
Table 2. Prediabetes Awareness and the Odds of Engagement in Risk-Reducing
Behaviors

October 2015

518

Gopalan et al / Am J Prev Med 2015;49(4):512519

capture weight loss or assess actual changes in behavior

over time. Additionally, we are only capturing stated
intentions and reported engagement; however, intention
and attempts to engage are arguably key initial steps
toward behavior change. Recall bias among survey
respondents, both in regard to past weight and exercise
amounts, may affect the validity of the ndings. We are
also unable to fully capture and examine the impact of
light physical activity in this population. We are unable
to capture the healthiness of participants diets, a key
element of weight management. There is also potential for
social desirability bias: those aware of prediabetes may be
more likely to report making the changes they are
supposed to make. We are unable to identify prediabetes
among survey non-participants and among those without
an HbA1c value. It is also possible that, given the use of the
HbA1c to identify those with prediabetes, we may have
incorrectly categorized individuals with undiagnosed diabetes as having prediabetes. However, there is no reason to
believe that an undiagnosed condition alone should
impact perceived risk or diabetes riskreduction behaviors. Also, space considerations prohibited examining and
discussing potentially important demographic and clinical
predictors of engagement in diabetes riskreducing behaviors. Finally, there are some factors unavailable in the
NHANES, such as activation, intrinsic motivation, and
provider differences, which might have been useful to
distinguish between those aware and unaware of
prediabetes.

Conclusions
Our ndings show that awareness of prediabetes is
associated with efforts to engage in diabetes riskreduction behaviors. Prediabetes awareness may provide
patients the motivation for behavior change; however,
most individuals with prediabetes are unaware of their
diagnosis. Additionally, more than 90% of individuals
with prediabetes, including both those who are aware and
those who are unaware of their diagnosis, are not meeting
the DPP intervention target levels of exercise and weight
loss found to be effective in preventing and delaying the
onset of diabetes. The U.S. Preventive Services Task
Force is considering changes to the recommendations for
diabetes screening that will likely lead to higher diagnosis
rates of prediabetes.23 These recommendations should be
approved and quickly adopted by healthcare providers in
order to increase levels of prediabetes awareness among
our patients. Our study also demonstrates that although
it is critical to diagnose prediabetes and counsel patients
about how to reduce their diabetes risk, this alone may be
insufcient for most people. Healthcare providers must
build strong ties with healthcare systems, communities,

and payers to increase the availability of evidence-based

structured lifestyle programs. The prediabetes period
presents an opportunity to intervene. Primary care
physicians, health systems, and insurers all have a vested
interest in making sure this opportunity is not missed.
All authors listed on the title page met the criterion for authorship
based on the Committee on Publication Ethics guidelines. AG, ISL,
and SCM contributed to study design, AG performed the data
analysis, CW assisted with the programming used for the data
analysis, SCM oversaw the statistical analysis, ISL and JAL
provided content expertise and contributed to discussion, AG
wrote the manuscript, ISL, CW, SCM, and JAL reviewed/edited the
manuscript, and JAL is the guarantor of this study and manuscript.
This work was presented as an oral abstract at the 2014 Society
for General Internal Medicine Annual Research Meeting.
No nancial disclosures were reported by the authors of
this paper.

References
1. CDC. National diabetes statistics report: estimates of diabetes and its
burden in the United States, 2014. Atlanta, GA: USDHHS, 2014.
2. Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the
incidence of type 2 diabetes with lifestyle intervention or metformin.
N Engl J Med. 2002;346(6):393403. http://dx.doi.org/10.1056/NEJ
Moa012512.
3. Eriksson KF, Lindgard F. Prevention of type 2 (non-insulin-dependent) diabetes mellitus by diet and physical exercise. The 6-year Malmo
feasibility study. Diabetologia. 1991;34(12):891898. http://dx.doi.org/10.
1007/BF00400196.
4. Pan XR, Li GW, Hu YH, et al. Effects of diet and exercise in preventing
NIDDM in people with impaired glucose tolerance. The Da Qing IGT and
Diabetes Study. Diabetes Care. 1997;20(4):537544. http://dx.doi.org/10.
2337/diacare.20.4.537.
5. Tuomilehto J, Linstrm J, Eriksson JG, et al. Prevention of type 2 diabetes
mellitus by changes in lifestyle among subjects with impaired glucose
tolerance. N Engl J Med. 2001;344(18):13431350. http://dx.doi.org/10.
1056/NEJM200105033441801.
6. Knowler WC, Fowler SE, Hamman RF, et al. 10-year follow-up of diabetes
incidence and weight loss in the Diabetes Prevention Program Outcomes
Study. Lancet. 2009;374(9702):16771686. http://dx.doi.org/10.1016/
S0140-6736(09)61457-4.
7. Ackermann RT, Finch EA, Brizendine E, Zhou H, Marrero DG. Translating the Diabetes Prevention Program into the community. The DEPLOY
Pilot Study. Am J Prev Med. 2008;35(4):357363. http://dx.doi.org/10.
1016/j.amepre.2008.06.035.
8. Whittemore R. A systematic review of the translational research on the
Diabetes Prevention Program. Transl Behav Med. 2011;1(3):480491.
http://dx.doi.org/10.1007/s13142-011-0062-y.
9. Ma J, Yank V, Xiao L, et al. Translating the Diabetes Prevention Program
lifestyle intervention for weight loss into primary care: a randomized trial.
JAMA Intern Med. 2013;173(2):113121. http://dx.doi.org/10.1001/
2013.jamainternmed.987.
10. McTigue KM, Conroy MB, Bigi L, Murphy C, McNeil M. Weight loss
through living well: translating an effective lifestyle intervention into
clinical practice. Diabetes Educ. 2009;35(2):199204. http://dx.doi.org/10.
1177/0145721709332815 208.
11. Kramer MK, McWilliams JR, Chen HY, Siminerio LM. A communitybased diabetes prevention program: evaluation of the group lifestyle

www.ajpmonline.org

Gopalan et al / Am J Prev Med 2015;49(4):512519

12.

13.

14.
15.

16.

17.

18.

balance program delivered by diabetes educators. Diabetes Educ.

2011;37(5):659668. http://dx.doi.org/10.1177/0145721711411930.
Weinstock RS, Trief PM, Cibula D, Morin PC, Delahanty LM. Weight
loss success in metabolic syndrome by telephone interventions: results
from the SHINE Study. J Gen Intern Med. 2013;28(12):16201628.
http://dx.doi.org/10.1007/s11606-013-2529-7.
Geiss LS, James C, Gregg EW, Albright A, Williamson DF, Cowie CC.
Diabetes risk reduction behaviors among U.S. adults with prediabetes.
Am J Prev Med. 2010;38(4):403409. http://dx.doi.org/10.1016/
j.amepre.2009.12.029.
CDC. Awareness of prediabetesUnited States, 2005-2010. MMWR
Morb Mortal Wkly Rep. 2012;62(11):209212.
Rosenstock C, Strecher VJ, Becker MH. Social learning theory and the Health
Belief Model. Health Edu Q. 1998;15(2):175183. http://dx.doi.org/10.1177/
109019818801500203.
CDC. National Health and Nutrition Examination Survey, 2005-2008
[codebook, analytic guidelines, datasets]. August 7, 2014. www.cdc.
gov/nchs/nhanes.htm.
Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief
depression severity measure. J Gen Intern Med. 2001;16(9):606613.
http://dx.doi.org/10.1046/j.1525-1497.2001.016009606.x.
Kreuter MW, Chheda SG, Bull FC. How does physician advice
inuence patient behavior? Evidence for a priming effect. Arch

October 2015

19.

20.

21.

22.

23.

519

Fam Med. 2000;9(5):426433. http://dx.doi.org/10.1001/archfami.9.5.

426.
Coutinho M, Gerstein HC, Wang Y, Yusuf S. The relationship between
glucose and incident cardiovascular events. A metaregression analysis
of published data from 20 studies of 95,783 individuals followed for
12.4 years. Diabetes Care. 1999;22(2):233240. http://dx.doi.org/10.
2337/diacare.22.2.233.
Meigs JB, DAgostino Sr, Nathan DM, Rifai N, Wilson PW. Longitudinal
association of glycemia and microalbuminuria: the Framingham Offspring
Study. Diabetes Care. 2002;25(6):977983. http://dx.doi.org/10.2337/
diacare.25.6.977.
Smith NL, Barzilay JI, Shaffer D, et al. Fasting and 2-hour postchallenge
serum glucose measures and risk of incident cardiovascular events in the
elderly: the Cardiovascular Health Study. Arch Intern Med. 2002;162
(2):209216. http://dx.doi.org/10.1001/archinte.162.2.209.
Self-reported prediabetes and risk-reduction activitiesUnited
States, 2006. MMWR Morb Mortal Wkly Rep. 2008;57(44):1203
1205.
U.S. Preventive Services Task Force. Topic Update in Progress,
Abnormal Glucose and Type 2 Diabetes Mellitus in Adults: Screening.
December 20, 2014. www.uspreventiveservicestaskforce.org/Page/
Document/UpdateSummaryDraft/screening-for-abnormal-glucose-andtype-2-diabetes-mellitus.