A Comparative Approach to Using the Diabetes Prevention Program to Reduce Diabetes Risk in a Worksite Setting

Abstract

The rapidly increasing rate of non–insulin-dependent diabetes (NIDDM) among both market economy and developing countries is a worldwide health phenomenon. The number of diabetics worldwide has been projected to increase from 135 million in 1995 to 300 million in 2025. The purpose of this study was to examine the relative effectiveness of three different approaches to the implementation of the Diabetes Prevention Program, a standardized diabetes prevention curriculum, in various worksite organizations within a single community. The methods of implementation selected included an intensive one-on-one counseling approach, a support group meeting approach, and a passive transfer of information approach. The intervention was successful in creating significant mean improvements overall in the participants who completed the 26-week program as follows: (a) reduction in overall mean body weight and mean body mass index, (b) reduction in overall average mean arterial blood pressure, (c) reduction in overall mean diabetes risk score, and (d) increase in overall mean physical activity level. Although the largest proportion of these changes occurred in the one-on-one intervention group, significant changes in some factors were found in all groups. This illustrates the utility of an on-site and incentive-driven approach to diabetes risk factor modification in the workplace.

Keywords

diabetes prevention diabetes prevention program worksite health promotion non–insulin-dependent diabetes

Background/Literature Review

The rapidly increasing rate of Type 2 diabetes among both market economy and developing countries is a worldwide health phenomenon (World Health Organization, 2011). The rate of this disease tripled worldwide between 1985 and 2005 (Halimi, 2006). The number of diabetics worldwide has been projected to increase from 135 million in 1995 to 300 million in 2025 (Seidel, Powell, Zgibor, Siminerio, & Piatt, 2008). The World Health Organization (2011) estimates that diabetes deaths will double between 2005 and 2030. In the United States, the occurrence of Type 2 diabetes is projected to double by the year 2050 (Narayan, Boyle, Geiss, Saaddine, & Thompson, 2006). Researchers studying the problem have concluded that emerging lifestyle patterns of behavior, such as reduced physical activity and increased caloric consumption, have resulted in an increasing rate of both obesity and diabetes (Seidel et al., 2008). There is firm evidence that intense and well-implemented lifestyle interventions can effectively reduce the risk of type 2 diabetes (Hamman et al., 2006; Knowler et al., 2002; Pan, Yang, Li, & Liu, 1997; Parikh et al., 2010). Consequently, an important hypothesis is that a public health focus on changing behaviors, including both an increase in physical activity levels and an improved dietary pattern, will result in a reduction in the occurrence, morbidity, and mortality of Type 2 diabetes on a broad level in a worksite setting.

Increased body weight and elevated body mass index (BMI) have both been identified in the literature as important risk factors in the development of Type 2 diabetes. (Balluz, Okoro, & Mokdad, 2008; Cicero et al., 2008; Yang et al., 2008; Zhang et al., 2010). Additional risk factors include abnormal waist circumference and waist-to-hip ratio (Lear, James, Ko, & Kumanyika, 2010; Qiao & Nyamdor, 2010), high resting blood pressure and heart rate (Owen, Retegan, Rockell, Jennings, & Reid, 2009), and low levels of physical activity (Gill & Cooper, 2008; Hu et al., 2004; Manson et al., 1991).

Specific lifestyle interventions, including the Diabetes Prevention Program (DPP), have been determined to be effective in modifying risk factors in those who exhibit impaired glucose tolerance in clinical settings in comparison with drug therapies (Brink, 2009; Diabetes Prevention Program Research Group, 2002; Knowler et al., 2002). Follow-up studies confirm the long-term effectiveness of such an approach (Knowler et al., 2009; Kramer et al., 2009). In addition, the DPP has been shown to be effective when applied in a variety of community settings (Amundson et al., 2009; Jackson, 2009), including the YMCA (Ackermann, Finch, Brizendine, Zhou, & Marrero, 2008), faith-based settings (Boltri et al., 2008; Frank & Grubbs, 2008), and urban medically underserved communities (Seidel et al., 2008). A review of such studies suggests that the use of incentives is important to successful behavioral change (Otto, Garcia, & Jakicic, 2008). Furthermore, theoretical analysis suggests the possibility of a net financial gain for employers who are able to successfully implement such a program (Ackerman et al., 2008; Musich, Burton, & Edington, 1999). Herman et al. (2005) found that if lifestyle changes were sustained continuously through the remainder of a subject’s life, the costs of treatment with intensive lifestyle intervention are only $1,036 more than that for the placebo group of the DPP study ($51,607 vs. $50,571 for placebo). However, differences in the method of presentation of the DPP curriculum in community settings need further study (Diabetes Prevention Program Research Group, 2002). More detailed information about the DPP curriculum can be found at the following website: http://content.nejm.org/cgi/content/abstract/346/6/393.

The purpose of this study was to implement a diabetes risk factor reduction service program while simultaneously examining the relative effectiveness of three different approaches to the implementation of a standardized diabetes prevention curriculum, the DPP, in various large organizations within a single community. The study provides a model for the implementation of the DPP curriculum using a self-selected multidimensional approach in an incentive-driven environment with the intent of deriving maximum diabetes risk factor reduction benefit with the most efficient use of resources. The study used the goal-based theoretical approach inherent in the DPP. The goal-based approach provides credible evidence regarding the effectiveness of the treatment method and the method of implementation. In this study, the participants chose their own method of implementation in a manner essential to the maintenance of free will within the workplace. The methods of implementation selected included an intensive one-on-one counseling approach, a support group meeting approach, or a passive transfer-of-information approach. Qualitative assessment was also made of the financial costs and gains of implementing such a program. The study promotes the expanded use of the DPP curriculum, which supports a workplace environment that promotes health. This study is consistent with the Journal of Health Promotion Practice’s mission to enhance research to practice links.

Methods/Strategies/Intervention Applications

Participants

Study participants included 264 staff members from four worksite organizations in a midsized urban area of the western United States, including a newspaper publisher, two departments within a public hospital, the city/county health department, and the city/county police department. Worksites were selected based on the following characteristics: (a) management support for participation, (b) ability to provide space for on-site application of the program, and (c) willingness to provide incentives for participation. The project was approved by the University Internal Review Board.

Research Design

The study used a three-group quasi-experimental design because of the ethical decision not to randomly assign subjects to groups in the context of a workplace application of the model. Five primary quantitative variables were compared pre and post as well as between groups: body weight, mean arterial blood pressure, self-report physical activity level, BMI, and a diabetes risk score. The variables were measured before and after the intervention period in successive group health screenings conducted by program staff trained in health promotion methods. Body weight was measured using a calibrated balance beam scale and height, using a standard stadiometer. BMI was calculated from those variables. Daily physical activity kilocaloric consumption was estimated based on the participants’ self-reported activity levels, obtained using the Baecke Questionnaire of Habitual Physical Activity (Baecke, Burema, & Frijters, 1982; Hertogh, Monninkhof, Schouten, Peeters, & Schuit, 2008). Blood pressure was measured by auscultation while seated at rest, using a standard sphygmomanometer and stethoscope. Every effort was made to conduct the screenings in the same settings both pre and post. The diabetes risk score was calculated from a standardized survey based on a validated diabetes risk score system (Lindstrom & Tuomilehto, 2003).

Individual participants were incentivized to increase participation in the screenings and in the program. Incentives were determined based on prior survey of the priority population, and a lottery approach was used to minimize costs. The most typical incentive used was the provision of a gift certificate to a retail vendor offering a wide assortment of products. Group assignment in the program was by self-selection because of the in-parallel service nature of the program. Financial costs and gains were assessed by tracking program costs and interviewing key personnel in the sole provider who was externally insured.

Treatment

The intervention was based on the DPP curriculum and was implemented over 26 weeks using the following three methods: (a) an intensive one-on-one education cluster, (b) a support group intervention cluster, and (c) a passive intervention cluster serving as the control. Each method of implementation was used at each site. Participants self-selected their method of participation following a brief description of each at the initial screening. One-on-one group members met weekly with a lifestyle counselor, who held them accountable for program goals and delivered the DPP curriculum in a sequential manner. Support group members met once weekly with each other and a lifestyle health educator, who presented the curriculum in a class format. Accountability for the program meeting behavioral goals was not individualized or formalized, although individuals assisted each other in a typical support group format. Passive intervention members simply received the information via e-mails, flyers, and periodic educational presentations available to all employees. The health screenings, lifestyle counseling, and educational services were offered on site by grant employees trained in health promotion methods. Health-related presentations available to all employees were conducted by both grant staff and selected professionals in the community.

Data Analysis

Means and standard deviations were calculated pre and post for the main effect on each dependent variable as well as for the individual group effects on each dependent variable. Mean differences pre and post and interaction effects between groups for each dependent variable were analyzed using separate 3 × 2 mixed-factorial analysis of variance (ANOVA) procedures. Post hoc assessment of individual group differences was performed using Tukey’s test of honestly significant differences.

Results

A total of 264 participants officially began the study, with 151 completing the program, resulting in a 57.20% completion rate. A total of 45 individuals completed the group intervention subset, 49 individuals completed the one-on-one intervention subset, and 57 individuals completed the passive intervention subset. Results of the 3 × 2 ANOVAs for each of the five dependent variables studied across the three intervention methods (groups) pre and post (trials) are as follows.

Mean Arterial Blood Pressure (MAP)

The main effect of trials was significant (F = 10.17, p < .002, observed power = 0.887). The overall study group showed a significant reduction in resting mean arterial blood pressure from 95.09 ± 0.76 to 92.64 ± 0.89 mmHg following the treatment period. There were no significant differences between groups and no significant interaction between groups and trials.

Body Mass Index (BMI)

The main effect of groups was significant (F = 6.00, p < .003, observed power = 0.867). The passive and group intervention subset means were not significantly different. The one-on-one intervention subset had a significantly higher mean BMI (m = 27.76 ± 0.90) than the passive intervention subset (m = 27.13 ± 0.80). The main effect of trials was significant (F = 8.996, p < .003, observed power = 0.846). The overall study group showed a significant reduction in BMI from 28.79 ± 0.49 to 28.47 ± 0.50 following the treatment period. The main effect of interaction (trials × group) was significant (F = 6.552, p < .002, observed power = 0.904). The one-on-one subset showed a significant reduction in BMI from 31.43 ± 6.27 to 30.59 ± 6.29, whereas the group and passive intervention subsets were unchanged following the treatment period, as illustrated in Table 1.

Table 1

Pre Versus Post Body Mass Index (BMI)

	One-on-One Intervention	Group Intervention	Passive Intervention
Pre-BMI	31.43	27.77	27.17
Post-BMI	30.56	27.74	27.10

Body Weight (Weight)

The main effect of group was significant (F = 3.800, p < .025, observed power = 0.684). The passive and group intervention subset means were not significantly different. The one-on-one intervention subset had a significantly higher mean body weight (m = 188.63 ± 5.99 lb) than the group intervention (m = 165.40 ± 6.25 lb). The main effect of trials was significant (F = 15.767, p < .001, observed power = 0.976). The overall study group showed a significant reduction in mean body weight from 176.98 ± 3.43 lb to 174.40 ± 3.46 lb following the treatment period. The main effect of interaction (trials × group) was significant (F = 4.47, p < .014, observed power = 0.752). The one-on-one subset showed a significantly greater reduction in mean body weight from 191.26 ± 46.58 lb to 185.99 ± 46.59 lb than the group and passive intervention subsets, which were reduced by a similar amount following the treatment period (from 166.16 ± 36.19 lb to 164.64 ± 35.41 lb and from 173.51 ± 41.90 lb to 172.55 ± 43.48 lb, respectively), as illustrated in Table 2.

Table 2

Pre Versus Post Body Weight

	One-on-One Intervention	Group Intervention	Passive Intervention
Pre-weight	191.26	166.16	173.51
Post-weight	185.99	164.64	172.55

Physical Activity Levels (Pascoe)

The main effect of group was significant (F = 9.357, p < .00, observed power = 0.977). The one-on-one and group intervention subset means were not significantly different. The passive intervention subset had a significantly higher reported mean physical activity score (m = 8.07 ± 0.16) than the group intervention and one-on-one intervention subsets (m = 7.41 ± 0.18 and m = 7.06 ± 0.17, respectively). The main effect of trials was significant (F = 28.6, p < .0001, observed power = 1.00). The overall study group showed a significant increase in reported physical activity level from 7.25 ± 0.11 to 7.77 ± 0.12 following the treatment period. The main effect of interaction (trials × group) was significant (F = 3.326, p < .039, observed power = 0.622). The one-on-one subset showed a significantly greater mean increase in reported physical activity level (from 6.62 ± 1.30 to 7.49 ± 1.55) than the group and passive intervention subsets (from 7.22 ± 1.35 to 7.59 ± 1.35 and from 7.91 ± 1.30 to 8.22 ± 1.30, respectively) following the treatment period, as illustrated in Table 3.

Table 3

Pre Versus Post Physical Activity Level

	One-on-One Intervention	Group Intervention	Passive Intervention
Pre-physical activity	6.62	7.22	7.91
Post-physical activity	7.49	7.59	8.22

Diabetes Risk Score (Rescore)

The main effect of groups was significant (F = 12.20, p < .001, observed power = 0.995). The passive and group intervention subset means were not significantly different. The one-on-one intervention subset had a significantly higher mean diabetes risk score (m = 9.32 ± 0.57) than the group intervention and the passive intervention subsets (m = 6.70 ± 0.59 and m = 5.55 ± 0.53, respectively). The main effect of trials was significant (F = 23.70, p = .00, observed power = 0.998). The overall study group showed a significant reduction in diabetes risk score from 7.91 ± 0.35 to 6.47 ± 0.36 following the treatment period. The main effect of interaction (trials × group) was significant (F = 5.38, p < .006, observed power = 0.837). The one-on-one subset showed a significantly greater reduction in mean diabetes risk score (from 10.71 ± 3.70 to 7.91 ± 4.35) than the group and passive intervention subsets (from 7.11 ± 4.93 to 6.29 ± 4.96 and from 5.89 ± 4.29 to 5.21 ± 3.94, respectively) following the treatment period, as illustrated in Table 4.

Table 4

Pre Versus Post Diabetes Risk Score

	One-on-One Intervention	Group Intervention	Passive Intervention
Pre-diabetes risk	10.71	7.11	5.89
Post-diabetes risk	7.92	6.29	5.21

Discussion

The intervention was successful in creating significant mean improvements overall in the 151 participants who completed the 26-week program, as follows: (a) reduction in overall mean body weight and mean BMI, (b) reduction in overall average mean arterial blood pressure, (c) reduction in overall mean diabetes risk score, and (d) increase in overall mean physical activity level. The highest risk individuals self-selected participation in the one-on-one intervention group, inferring self-awareness of their need for greater assistance in the health promotion process. Although the largest proportion of these changes occurred in the one-on-one intervention group, significant changes in some factors were found in all groups when analyzed individually. This illustrates the utility of an on-site and incentive-driven approach to diabetes risk factor modification in the workplace in general.

Although all methods of intervention were successful to some degree with regard to change in the majority of the risk factors measured, the one-on-one approach produced significantly greater improvements than the other interventions in four of the five risk factors measured, in spite of the fact that the self-selected one-on-one participants began the treatment with significantly worse mean risk factor scores than the group and the passive intervention groups in every case. Various other studies have also shown significant improvements with DPP in high-risk groups (Amundson et al., 2009; Jackson, 2009), so our finding is not unexpected. This study provides evidence that high-risk individuals will self-identify when a health-screening process is initiated with participation incentivized. The finding that higher risk individuals in an organization will both self-select and prosper to a greater degree in a one-on-one approach to application of the DPP at the worksite has important implications for the worksite-based implementation of such curriculums in general. It strongly supports the utility of hiring a health promotion professional to work in a one-on-one manner with high-risk individuals.

In the first of the four organizations used to implement this study, the management team was able to renegotiate their insurance premiums to a lower level after illustrating the broad reduction in chronic disease risk factors created by implementing this program. Although a specific cost analysis cannot be reported in this context because of the confidential nature of the information, interviews conducted with key management individuals were clear in referring to the financial utility of the program. This is further illustrated by the organization’s decision to hire a health promotion specialist to continue the program once the grant support was removed. The second major site, a medical provider, was self-insured and so was unable to test this concept, and the last sites used were unable to provide any kind of insightful analysis as to the impact of the program financially as the grant intervention ended.

Combined with our findings regarding self-selection of the intervention method, the financial outcomes illustrate both a cost-effective and a benefit-intensive strategy for implementing a diabetes risk reduction program on the worksite. Organizations that target the highest risk members of their population with a one-on-one goal-based intervention lifestyle modification approach can aspire to reduce diabetes risk and lower health care costs.

Conclusion

In conclusion, this study illustrates the effectiveness of implementing a diabetes risk factor reduction program, which is based on an established curriculum (the DPP) and delivered by health promotion professionals on the worksite and which uses an external incentive program, on improvement of the most significant measurable diabetes risk factors, including mean body weight, mean BMI, mean physical activity level, mean arterial blood pressure, and mean diabetes risk score. The study further illustrates the effectiveness of a one-on-one delivery approach with high-risk individuals and the lesser effectiveness of both passive and group format approaches in addressing the risk factor status of the lower risk members of the organization. The findings indicate the importance of identifying high-risk individuals in a worksite setting and the utility of an on-site and incentive-driven approach to diabetes risk factor modification and health care cost reduction in the workplace.

Footnotes

Authors’ Note:

This research was supported by a grant from the Colorado Trust.

References

Ackermann

R. T.

Finch

E. A.

Brizendine

Zhou

Marrero

D. G.

(2008). Translating the diabetes prevention program into the community. The DEPLOY pilot study. American Journal of Preventive Medicine, 35, 357-363.

Amundson

H. A.

Butcher

M. K.

Gods

Hall

T. O.

Harwell

T. S.

Halverson

S. D.

Vanderwood

K. K.

(2009). Translating the diabetes prevention program into practice in the general community: Findings from the Montana cardiovascular disease and diabetes prevention program. The Diabetes Educator, 35, 209-223.

Baecke

J. A.

Burema

Frijters

J. E.

(1982). A short questionnaire for the measurement of habitual physical activity in epidemiological studies. American Journal of Clinical Nutrition, 36, 936-942.

Balluz

L. S.

Okoro

C. A.

Mokdad

(2008). Association between selected unhealthy lifestyle factors, body mass index, and chronic health conditions among individuals 50 years of age or older, by race/ethnicity. Ethnicity & Disease, 18, 450-457.

Boltri

J. M.

Davis-Smith

Y. M.

Seale

J. P.

Shellengerger

Okosun

I. S.

Cornelius

M. E.

(2008). Diabetes prevention in a faith-based setting; results of translational research. Journal of Public Health Management and Practice, 14, 29-32.

Brink

(2009). The diabetes prevention program: How the participants did it. Health Affairs, 28, 57-62.

Cicero

A. F.

Dormi

Nascetti

Panourgia

M. P.

Grandi

D’Addato

Gaddi

(2008). The prevention of type 2 diabetes. Nature Clinical Practice Endocrinology & Metabolism, 4, 382-393.

Diabetes Prevention Program Research Group. (2002). Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. New England Journal of Medicine, 346, 391-403.

Frank

Grubbs

(2008). A faith-based screening/education program for diabetes, CVD and stroke in rural African Americans. The Association of Black Nursing Faculty Journal, 19, 96-101.

10.

Gill

J. M. R.

Cooper

A. R.

(2008). Physical activity and prevention of type 2 diabetes mellitus. Sports Medicine, 38, 807-824.

11.

Halimi

(2006). Diabetes in 2005. Nephrologie & Therapeutique, 2(Suppl. 1), S2-S7.

12.

Hamman

R. F.

Wing

R. R.

Edelstein

S. L.

Lachin

J. M.

Bray

G. A.

Delahanty

. . . Wylie-Rosett

(2006). Effect of weight loss with lifestyle intervention on risk of diabetes. Diabetes Care, 29, 2102-2107.

13.

Herman

W. H.

Hoerger

T. J.

Brandle

Hicks

Sorensen

Zhang

. . . Ratner

R. E.

(2005). The cost-effectiveness of lifestyle modification or metformin in preventing type 2 diabetes in adults with impaired glucose tolerance. Annals of Internal Medicine, 142, 323-332.

14.

Hertogh

E. M.

Monninkhof

E. M.

Schouten

E. G.

Peeters

P. H.

Schuit

A. J.

(2008). Validity of the modified Baecke questionnaire: Comparison with energy expenditure according to the doubly labeled water method. International Journal of Behavioral Nutrition and Physical Activity, 5, 30.

15.

Lindström

Valle

Eriksson

Jousilahti

Silventoinen

. . . Tuomilehto

(2004). Physical activity, body mass index, and risk of type 2 diabetes in patients with normal or impaired glucose regulation. Archives of Internal Medicine, 164, 892-896.

16.

Jackson

(2009). Translating the diabetes prevention program into practice: A review of community interventions. The Diabetes Educator, 35, 309-320.

17.

Knowler

W. C.

Barrett-Connor

Fowler

S. E.

Hamman

R. F.

Lachin

J. M.

Walker

E. A.

Nathan

D. M.

(2002). Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. New England Journal of Medicine, 346, 393-403.

18.

Knowler

W. C.

Fowler

S. E.

Hamman

R. F.

Christophi

C. A.

Hoffman

H. J.

Brenneman

A. T.

. . . Nathan

D. M.

(2009). 10-year follow-up of diabetes incidence and weight loss in the diabetes prevention program outcomes study. Lancet, 374, 1677-1686.

19.

Kramer

M. K.

Kriska

A. M.

Venditti

E. M.

Miller

R. G.

Brooks

M. M.

Burke

L. E.

. . . Orchard

T. J.

(2009). Translating the diabetes prevention program: A comprehensive model for prevention training and program delivery. American Journal of Preventive Medicine, 37, 505-511.

20.

Lear

S. A.

James

P. T.

G. T.

Kumanyika

(2010). Appropriateness of waist circumference and waist-to-hip ratio cutoffs for different ethnic groups. European Journal of Clinical Nutrition, 64, 42-61.

21.

Lindstrom

Tuomilehto

(2003). The diabetes risk score: A practical tool to predict type 2 diabetes risk. Diabetes Care, 26, 725-731.

22.

Manson

J. E.

Rimm

E. B.

Stampfer

M. J.

Colditz

G. A.

Willett

W. C.

Krolewski

A. S.

. . . Speizer

F. E.

(1991). Physical activity and incidence of non-insulin-dependent diabetes mellitus in women. Lancet, 338, 774-778.

23.

Musich

S. A.

Burton

W. N.

Edington

D. W.

(1999). Costs and benefits of prevention and disease management. Disease Management and Health Outcomes, 5, 153-166.

24.

Narayan

K. M. V.

Boyle

J. P.

Geiss

L. S.

Saaddine

J. B.

Thompson

T. J.

(2006). Impact of recent increase in incidence on future diabetes burden: U.S., 2005-2050. Diabetes Care, 29, 2005-2050.

25.

Otto

A. D.

Garcia

D. O.

Jakicic

J. M.

(2008). Lifestyle intervention strategies to prevent and control type 2 diabetes. Current Diabetes Reports, 8, 407-412.

26.

Owen

A. J.

Retegan

Rockell

Jennings

Reid

C. M.

(2009). Inertia or inaction? Blood pressure management and cardiovascular risk in diabetes. Clinical and Experimental Pharmacology & Physiology, 36, 643-647.

27.

Pan

X. R.

Yang

W. Y.

G. W.

Liu

(1997). Prevalence of diabetes and its risk factors in China. Diabetes Care, 20, 1664-1669.

28.

Parikh

Simon

E. P.

Kezhen

Looker

Goytia

Horowitz

C. R.

(2010). Results of a pilot diabetes prevention intervention in East Harlem, New York City: Project HEED. American Journal of Public Health, 100(Suppl. 1), S232-S239.

29.

Qiao

Nyamdor

(2010). Is the association of type II diabetes with waist circumference or waist-to-hip ratio stronger than that with body mass index? European Journal of Clinical Nutrition, 64, 30-34.

30.

Seidel

M. C.

Powell

R. O.

Zgibor

J. C.

Siminerio

L. M.

Piatt

G. A.

(2008). Translating the diabetes prevention program into an urban medically underserved community. Diabetes Care, 31, 684-689.

31.

World Health Organization. (2011). Diabetes fact sheet. Retrieved from http://www.who.int/mediacentre/factsheets

32.

Yang

Telama

Hirvensalo

Mattsson

Viikari

J. S.

Raitakari

O. T.

(2008). The longitudinal effects of physical activity history on metabolic syndrome. Medicine and Science in Sports and Exercise, 40, 1424-1431.

33.

Zhang

Luo

Gregg

E. W.

Mukhtar

Rivera

Barker

Albright

(2010). Obesity prevention and diabetes screening at local health departments. American Journal of Public Health, 100, 1434-1441.