Week 6 - Article Review A research paper describes original research that has been conducted in the study (i.e., the researchers are publishing their findings for the first time).For this project, you

Each program component is not available at all VA facilities as implementation and structure of MOVE! is dependent on the staffing and resources available at local VA sites. MOVE! was implemented at VA facilities nationally in 2006. The Miami VA Healthcare System was a program pilot site and started MOVE! Weight Management for Veterans: Examining Change in Weight Before and a fter MOVE!

Jason R. Dahn 1, Stephanie L. Fitzpatrick 1,2, Maria M. Llabre 2, Greta S. Apterbach 1, Rebecca L. Helms 1, Marilyn L. Cugnetto 1, Johanna Klaus 1, Hermes Florez 3 and Tim Lawler 1 In the year 2000, 31% of women and 40% of men receiving outpatient care \ at Veteran Affairs (VA) medical facilities were overweight (BMI ≥ 25 and <30 kg/m 2); 37.4% of women and 32.9% of men were obese (BMI ≥ 30 kg/m 2).

The purpose of the present study was to assess treatment effects of MOVE\ ! Weight Management Program for Veterans by comparing the trajectory of change in weight postinterventio\ n (3, 6, and 12 months postenrollment) to a preintervention period (1, 3, and 5 years before enrollment). The sa\ mple consisted of 862 veterans participating in MOVE! at the Miami VA. All veterans participated in a 2-h Self-Managemen\ t Support (SMS) session, which involved completion of a self-assessment questionnaire and a nutrition education \ group session. After completing SMS, veterans had the option of continuing with Supportive Group Sessions (S\ GS), which included 10-weekly group sessions led by a multidisciplinary team. Veterans served as their own c\ ontrols in the analyses. Veterans gained 2 kg/ year before enrolling in MOVE!. There were similar increases in weight a\ cross sex, racial/ethnic groups, and treatment condition. Weight for participants in SMS stabilized after enrollment wh\ ereas participants in SGS had an average weight loss of 1.6 kg/year. The preintervention slope for weight was significantly differen\ t from the postintervention slope, suggesting treatment effect. Findings from this study support the\ need for a lifestyle modification program such as MOVE! in primary care settings to assist overweight and obese patient\ s in managing their weight. Obesity (2011) 19, 977–981. doi: 10.1038/oby.2010.273 1Miami Veterans Affairs Healthcare System, Mental Health and Behavioral Sciences Service, Miami, Florida,\ USA; 2University of Miami, Department of Psychology, Coral Gables, Florida, USA; 3Divisions of Geriatric Medicine and Endocrinology, University of Miami Miller School of Medicine, and Miami Veterans Affairs Healthcare System, GRECC, Miami, Florida, USA. Correspondence: Jason R. Dahn ( [email protected] ) Received 13 April 2010; accepted 1 August 2010; published online 2 December 2010. doi:

10.1038/oby.2010.273 978 VOLUME 19 NUMBER 5 | May 2011 | www.obesityjournal.org articles intervention and Prevention in January 2005. Currently, the Miami VA Healthcare System offers SMS, SGS, and weight-control medications (see below for description of SMS and SGS as offered in Miami).

To our knowledge, there is no literature on weight trends for individuals before they enroll in weight management programs.

Preintervention weight trends could serve as a control condition when examining postintervention change in weight in a nonran- domized sample. Therefore, the purpose of this present study was to model the trajectory of change in weight postintervention and compare it to that in the preintervention period to examine treat- ment effects. Differences in change in weight pre- and postinter - vention between treatment conditions (SMS and SGS) were also assessed. Finally, race/ethnicity and sex differences for change in weight postintervention were explored to examine the generaliz- ability of MOVE! program results.

Methods and Procedures Participants All aspects of this study were approved by the Miami VA Institutional Review Board. Participants included 1,000 veterans who enrolled in MOVE! at the Miami VA Healthcare System between 1 January 2005 and 24 April 2007. Participants were recruited into MOVE! by several means including being referred by their primary care provider if they were overweight/obese (BMI ≥25) or met other salient inclusion criteria (6) such as being normal weight but hypertensive, and through adver - tisements in patient education materials. A total of 138 veterans were excluded from analyses as a result of missing data on the MOVE!23 questionnaire. The final sample consisted of 862 veterans.

ProceduresOnly SMS and SGS were examined in this present study. Participants were not randomized, but instead self-selected into these treatment conditions. All veterans participated in SMS, which consisted of a 2-h nutrition education session. Veterans then had the option of enroll- ing and continuing in SGS (i.e., 10-weekly group sessions). In this present study, ~45% of veterans opted into SGS after completing SMS. Participation in SGS was open-ended in that veterans were given the option to repeat the group sessions.

MOVE!23 questionnaire. During the SMS session with the regis- tered dietitian veterans completed the MOVE!23 questionnaire (www.

move.va.gov/Move23. asp), which facilitated enrollment in the pro- gram. The MOVE!23 questionnaire is a multifactorial self-assessment that consists of 23 items covering the following areas: demographic information (e.g., race/ethnicity); medical and psychiatric history; weight management history; perceptions of body size; physical activ- ity, and eating habits; self-efficacy and readiness to change lifestyle habits; social support; and barriers to making lifestyle changes (6).

This instrument is evidence-based and provides tailored feedback to patients based on their responses. The MOVE!23 questionnaire can be completed online or via hardcopy and takes ~20–25 min to complete.

All participants included in this present study completed the paper and pencil version.

MOVE! intervention. In SMS and SGS, MOVE! participants were encouraged to set realistic and attainable goals and were instructed that a sustainable rate of weight loss was about 0.5–2 pounds/week. The 2-h nutrition education session in SMS was conducted in a group format led by a registered dietitian and focused on healthy eating and life- style change. During this meeting, patients completed the MOVE!23 questionnaire and received the MOVE! standard handouts (see http:// vaww.move.med.va.gov/handouts.asp?standard). Veterans interested in further participation were enrolled in SGS, comprised of a 10-week (90-min per session) multidisciplinary group intervention, addressing nutrition, physical activity, and behavioral modifications. In January 2007, our SGS were expanded from 90 min to 120 min, adding 30 min of low impact activity to seven sessions and 30 min of recreational ther - apy to two sessions. Each group session focused on a particular theme, but was divided to include nutrition, physical activity, and behavioral health perspectives. Veterans received MOVE! handouts (www.move.

va.gov/handouts.asp) each week to provide additional information on the topics discussed. Ta b l e 1 includes a list of topics presented during SGS.

Weight measurements. Preintervention and postintervention weight in pounds was obtained from veterans’ electronic medical records.

Weight was measured and entered by the medical staff when the vet- eran attended his or her regularly scheduled medical appointment.

Preintervention weights consisted of measurements from 5 years, 3 years, and 1 year before enrollment in the program. Postintervention weight consisted of measurements obtained at 3-, 6-, and 12-months postenrollment in MOVE!. A large window (±3 months) was used for preintervention weights and at 12-month follow-up whereas a window of ±1 month was used at 3- and 6-month follow-up to increase the likelihood of obtaining a valid weight value from the medical record.

Statistical analyses. To evaluate the change resulting from the intervention, our design relied on preintervention weight (up to 5 years prior) as the expected comparison. This quasiexperimental design has the features of an interrupted time series design (7), where t able 1 M o V e ! s upportive Group s essions c ondition ( s G s ): program outline NutritionPsychology Session 1 Obesity-related health risks and benefits of small behavior changes Maintaining motivation: identifying and overcoming barriers Session 2 Nutrition basics; using a food log Costs and benefits of behavior change Session 3 Physical therapy: facts on fitness, developing an exercise program, and using a pedometer Session 4 Food guide pyramid and nutrition facts Using goal setting to make and measure change Session 5 Modifying meal plans Planning ahead and managing impulses Session 6 Healthy shopping and food label reading Identifying and changing irrational ideas about food and eating Session 7 Basics of a low-fat diet Understanding emotions and behavior Session 8 Coping with cravings Coping with stress, anxiety, and depression Session 9 Eating out: restaurant options Eating with others and focusing on food Session 10 Special occasion tips and maintaining success Relapse prevention: tools for maintaining successful behavior change obesity | VOLUME 19 NUMBER 5 | May 2011 979 articles intervention and Prevention participants serve as their own controls, and analyses compared the change in weight postintervention, to the expected change in weight based on the trajectory before the intervention.

Using Hierarchical Linear Modeling software version 6.03 (8), a mixed model approach was applied with a piecewise linear function being speci-fied for each participant with an intercept at the start of the MOVE! program and two slopes: one before the intervention and one after the intervention. The piecewise approach allows the estimation of the two slopes separately but simultaneously. The mixed model estimates the aver - age within person change, as well as individual differences in intercept and slope across persons. Time was measured in years. Covariates (i.e., race/ethnicity, sex, and treatment condition) were included as predictors of the intercept, slope before enrollment, and slope after enrollment. The two slopes were compared using linear contrasts. In addition, a mixed design, repeated measures ANOVA was conducted to explore the effect of the intervention on weight status at each follow-up time point (i.e., baseline to 3-months, 3–6-months, and 6–12-months). Difference between SMS and SGS for change in weight trajectories at pre- and postintervention was assessed in a separate analysis. To examine the generalizability of the program across racial/ethnic groups and sex, racial/ethnic groups as well as men and women were compared within treatment condition on change in weight at postintervention. Missing data were handled by applying full maximum likelihood.

results Mean age for the 862 veterans at time of enrollment was 54.3 years (s.d. = 11.4). Ta b l e 2 displays sample characteristics endorsed on the MOVE!23 questionnaire. The majority of the veterans met criteria for obesity and over half of the sample endorsed having high-blood pressure. Ta b l e 3 presents treat- ment condition comparisons on age, BMI, and total number of medical and psychiatric conditions endorsed on the MOVE!23.

Participants in the SMS condition were younger (P = 0.001), had significantly lower baseline BMI (P < 0.001), and had fewer number of medical conditions (P < 0.001) and psychiatric con- ditions (P < 0.001) than SGS participants. Despite treatment condition differences, there were no significant racial/ethnic group differences in BMI at time of enrollment (P = 0.22).

Veterans gained an average of 2 kg (β = 4.32 (pounds), s.e. = 0.32, P < 0.001) per year before enrolling in MOVE!. More specifically, there were no sex differences (β = 0.93, s.e. = 0.66, P = 0.16) and no difference between white non-Hispanics com- pared to Hispanics (β = 0.51, s.e. = 0.59, P = 0.39) for trajec- tory change in weight preintervention. There was a marginally significant difference in preintervention slope between white non-Hispanics compared to African-Americans (β = 0.98, s.e. = 0.53, P = 0.07). After enrolling in the program, veterans on average lost about 1 kg/year (β = −2.07 (pounds), s.e. = 0.48, P < 0.001) postintervention. A test of linear contrast (control- ling for sex, race/ethnicity, and treatment condition) suggested that the slope before the intervention was significantly different from the slope after the intervention (χ 2(1) = 7.85, P < 0.01), indicating treatment effect.

The graph in Figure 1 displays the pattern of change in mean weight using the linear regression equations for SMS and SGS before and after enrolling in MOVE! There was no significant difference in slope before enrolling in the intervention between the two groups (β = 0.41, s.e. = 0.44, P > 0.05), which suggests that participants in SMS and SGS were gaining weight at the same rate per year before enrolling. There was a significant treatment condition difference in body size at the start of the intervention ( β = 15.57 (pounds), s.e. = 3.44, P < 0.001). Veterans who opted to continue with SGS weighed on average 7 kg more at the start of the intervention than veterans who only participated in SMS.

Weight for veterans that only participated in SMS stabilized after enrollment in MOVE! (β = 0.44 (pounds), s.e. = 0.78, P = 0.58).

However, participants in SGS lost, on average, 1.6 kg/year postintervention (β = −3.58 (pounds), s.e. = 0.78, P < 0.001).

Based on post-hoc analyses from the repeated measures ANOVA, veterans in SGS demonstrated significant weight loss between the start of MOVE! and 3-month follow-up (M = −2.91 (pounds), s.e.

= 0.47, P < 0.001) as well as 3–6-month follow-up (M = −1.06 (pounds), s.e. = 0.38, P < 0.01), but had a nonsig- nificant decline in weight from 6-month to 12-month follow-up (M = −0.44 (pounds), s.e. = 0.49, P = 0.37).

Because only the SGS condition resulted in significant weight loss, race/ethnicity, and sex differences at postintervention were assessed for those that participated in SGS. There were t able 2 s ample characteristics (N = 862) Percent Race/ethnicity White non-Hispanic 29.8 African American 36.4 Hispanic 26.3 Other 7.6 Gender Men 85.8 Women 14.2 Normal weight (BMI <25) 1.1 Overweight (BMI ≥25 and ≤29.9) 18.4 Obese (BMI ≥30) 80.5 Smoker (yes) 15.5 Hypertension 55.7 High blood cholesterol 46.8 Diabetes 29.2 Heart disease 18.6 Health status Excellent, very good, or good 55.1 Fair 32.7 Poor 12.2 table 3 t reatment condition comparisons: mean (s.d.) Self-management support condition (N = 470) Supportive group sessions condition (N = 392) Age (years) 53.03 (12.3)55.55 (10.4) BMI (kg/m 2) 34.19 (6.1) 36.56 (6.9) Total medical conditions 3.47 (2.5)4.14 (2.6) Total psychiatric conditions 1.89 (2.1)2.60 (2.4) All mean tests are significant at P ≤ 0.001.

980 VOLUME 19 NUMBER 5 | May 2011 | www.obesityjournal.org articles intervention and Prevention significant racial/ethnic group differences at postintervention.

Figure 2 presents the pattern of change in mean weight before and after enrolling in MOVE! for each racial/ethnic group.

The postintervention slope for white non-Hispanics was sig- nificantly and marginally different from the postintervention slopes for African Americans ( β = 3.93, s.e. = 1.96, P < 0.05) and Hispanics (β = 3.93, s.e. = 2.21, P = 0.08), respectively.

More specifically, white non-Hispanics participating in SGS lost on average 2.7 kg/year (β = −6.10 (pounds), s.e. = 1.38, P < 0.001). African Americans had a marginally significant weight loss of 1 kg, on average, per year (β = −2.21 (pounds), s.e. = 1.25, P = 0.08). In contrast, Hispanics, on average, did not have significant weight loss after participating in SGS (β = −2.22 (pounds), s.e. = 1.49, P = 0.14). The nonsignifi- cant weight loss among Hispanics is due to low power given that only 25% of participants in SGS were of Hispanic back- ground. Men participating in SGS lost, on average, 1.8 kg/year (β = −3.99 (pounds), s.e. =0.92, P < 0.001). On the other hand, women participating in SGS had a nonsignificant decline in weight (β = −1.6 (pounds), s.e. = 2.30, P = 0.49). Interestingly, there was no significant difference in postintervention slope (β = 2.39, s.e. = 2.48, P = 0.34) between men and women. Perhaps the lack of significant difference in slope postinterven- tion between men and women, despite an obvious difference in amount of weight loss, is also due to low power given that only 15% of participants in SGS were women.

dIscuss I on The purpose of the present study was to model the trajectory of change in weight preintervention and compare it to change in weight postintervention in order to assess treatment effects of the Miami VA MOVE! program in a large sample of overweight and obese veterans. To our knowledge, this is the first weight management intervention study to assess the trajectory of weight change preintervention, allowing for assessment of change after the intervention with respect to preintervention weight trends.

Racial/ethnic, sex, and treatment condition differences for these trajectories were also examined. Results indicated that veter - ans gained 2 kg/year before enrolling in MOVE!. Trajectory of change in weight postintervention suggested that veterans, on average, lost weight (i.e., ~1 kg/year). The effectiveness of the MOVE! intervention was supported given the significant differ - ence between the preintervention slope and the postinterven- tion slope. Introduction of MOVE! appeared to prevent further weight gain. Veterans who attended group sessions conducted by a multidisciplinary team had more weight loss than those only attending session designed to promote self-management.

The weight change trajectory for participants in SMS (i.e., 2-h group session with a dietitian) flattened after starting the MOVE! program. These findings suggest that participat- ing in a brief nutrition education session helped veterans to maintain their baseline weight (i.e., weight at time of MOVE!

enrollment) over a period of 1 year. Weight maintenance is an important accomplishment in that one must stop gaining weight in order to lose weight. At the start of the program, par - ticipants in SGS (i.e., 10-week interdisciplinary program) had significantly higher weight and thus higher BMI, were older, and endorsed more medical and psychiatric conditions com- pared to participants in SMS. Given that veterans self-select into additional program components, perhaps veterans with greater weight and comorbid medical/mental illness were more motivated or encouraged to seek additional help and therefore opted into SGS. Participants in SGS, on average, lost 1.6 kg/ year, with majority of weight loss occurring in the 3 months after enrollment. Previous literature suggests that even moder - ate weight loss has health benefits (6) including a reduction in incident diabetes (9). Also, findings are generally consistent with those reported from a randomized trial comparing popu- lar diets over a comparable time period (10).

Results of this study suggest that level of participation and other demographic factors may be important determinants of change following weight management intervention. Among those veterans participating in the SGS intervention, white non-Hispanics lost, on average, 2.7 kg/year postintervention.

African Americans had a marginally significant weight loss of 1 kg/year and although Hispanics had the same amount of weight loss as African Americans, it was not significant. Men lost, on average, 1.8 kg/year and women participating in SGS 95 100 105 110 115 531 Enrolled in MOVE! 0.25 Years 0.50 1 Weight (kg) Self-management support (SMS) Supportive group sessions (SGS) Figure 1 Trajectory of pre- and postintervention weights for MOVE!

Self-Management Support (SMS) and Supportive Group Sessions (SGS) using linear regression equations. Preintervention is defined as 1, 3, and 5 years before enrollment in MOVE!; postintervention is 3 months (0.25 years), 6 months (0.5 years), and 12 months (1 year) after enrollment.

100 103 106 109 112 115 118 531 Enrolled in MOVE! 0.25 Years 0.50 1 Weight (kg) Hispanics African Americans White non-Hispanics Figure 2 Trajectory of pre- and postintervention weights for white non- Hispanics, African Americans, and Hispanics using the respective linear regression equations. Preintervention is defined as 1, 3, and 5 years before enrollment in MOVE!; postintervention is 3 months (0.25 years), 6 months (0.5 years), and 12 months (1 year) after enrollment.

obesity | VOLUME 19 NUMBER 5 | May 2011 981 articles intervention and Prevention had a nonsignificant weight loss of 0.73 kg postintervention.

The differences in weight outcomes between men and women and among the racial/ethnic groups may be related to an issue of power and/or a need for more culturally sensitive and rel- evant program components. MOVE! program materials have been translated and are available in Spanish though pro- grams may be limited in their ability to deliver the program in Spanish. Greater sensitivity to cultural issues within the inter - vention program such as addressing the role of food within the culture and having bilingual and bicultural group facilitators may lead to better outcomes (11,12). Given that the sample consisted of veterans, there may be some unique barriers or motivators to weight management that should be further addressed. For example, veterans who receive services at VA facilities tend to have more medical comorbidity, greater disability, and are less likely to have private health insurance (13).

Although MOVE! is provided to veterans free of charge, veterans do not receive compensation for their participation in the pro- gram as in some randomized control trials. Future studies may involve examining the impact of monetary rewards on program participation, attrition, and weight loss maintenance. Weight loss research has tended to evaluate the efficacy of spe- cific behavioral treatments (14), and treatment effects are opti- mized by participant selection, controlled conditions, incentives, and resource intensive interventions (15). The present study evaluated the effectiveness of MOVE! as a large-scale, hospital- based program targeting overweight and obese veterans. While the results might underestimate intervention effects, it provides a more realistic estimate of change given the actual contingencies affecting both patients and treatment providers. Furthermore, the findings from this study support the implementation of a prevention oriented health program that has been called for by VHA policy and clinical practice guidelines. With regard to limitations, the main threat to internal valid- ity in our design is history, such as there being other factors in the veterans’ environments that changed coincidentally with their start in the MOVE! program that could account for the change in weight. However, this is an unlikely threat because the coincidental change would have to be robust enough to affect a diverse sample of veterans in the same way at different times above and beyond enrolling in a weight management program.

MOVE! enrollment is predominantly triggered by primary care staff responding to an annual clinical reminder such that it is offered to veterans on a yearly basis without regard to health status and/or recent diagnosis of weight-related condition.Obesity is a biopsychosocial problem that impacts health out- comes, quality of life, and health care costs. As a result, weight management involves a multifactorial approach including lifestyle changes in nutrition, physical activity, and behavioral modifications. Findings from this study suggest the need for a multidisciplinary lifestyle modification program in primary care settings to screen for overweight/obesity, enhance patient motivation for change, and assist patients in setting nutrition and physical activity goals to promote weight management.

The impact of the program should be further addressed by examining the implications of weight maintenance and weight reduction on health outcomes (including number of newly diagnosed diabetes or cardiovascular disease cases as well as medication use) and health-care costs.

Furthermore, future studies should examine if the changes in weight are related to changes in lifestyle behaviors such as dietary intake and physical activity after enrolling in MOVE!

Process analyses should also be conducted to examine which components of the program (i.e., group facilitators, session topics, and session materials) were the most or least effective for men, women, and racial/ethnic groups. This type of infor - mation could aide in developing materials or modules that are more culturally sensitive and tailored to the unique needs of specific high-risk groups (e.g., women and ethnic minorities).

a cknowled GM entsWe would like to acknowledge the support and dedication of our MOVE!

Team at the Miami Veterans a ffairs Healthcare System as well as our VISN8 MOVE! Program partners. We owe a very special thanks to Kenneth R. Jones, PhD, National Program Director for Weight Management/MOVE!, and his staff at the National Center for Health Promotion and Disease Prevention, VH a Office of Patient Care Services, for their ongoing support and helpful comments.

dI sclosureThe authors declared no conflict of interest.

© 2010 The Obesity Society re F erences1. Flegal KM, Carroll MD, Ogden CL, Curtin LR. Prevalence and trends in obesity among US adults, 1999-2008. JAMA 2010;303:235–241.

2.

Das SR, Kinsinger LS, Yancy WS Jr et al. Obesity prevalence among veterans at Veterans Affairs medical facilities. Am J Prev Med 2005;28:291–294.

3.

Nelson KM. The burden of obesity among a national probability sample of veterans. J Gen Intern Med 2006;21:915–919.

4.

Young BA, Pugh JA, Maynard C, Reiber G. Diabetes and renal disease in veterans. Diabetes Care 2004;27 Suppl 2:B45–B49.

5.

Powell LH, Calvin JE 3rd, Calvin JE Jr. Effective obesity treatments. Am Psychol 2007;62:234–246.

6.

Kinsinger LS, Jones KR, Kahwati L et al. Design and dissemination of the MOVE! Weight-Management Program for Veterans. Prev Chronic Dis 2009;6:A98.

7.

Shadish WR, Cook TD, Campbell DT. Experimental and Quasi-experimental Designs for Generalized Causal Inference. Houghton Mifflin Company:

Boston, 2002.

8.

Raudenbush SW, Bryk AS, Congdon RT. HLM Version 6.03. Scientific Software International: Chicago, 2006.

9.

Tuomilehto J, Lindström J, Eriksson JG et al.; Finnish Diabetes Prevention Study Group. Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med 2001;344:1343–1350.

10.

Dansinger ML, Gleason JA, Griffith JL, Selker HP, Schaefer EJ. Comparison of the Atkins, Ornish, Weight Watchers, and Zone diets for weight loss and heart disease risk reduction: a randomized trial. JAMA 2005;293:43–53.

11.

Mier N, Ory MG, Medina AA. Anatomy of culturally sensitive interventions\ promoting nutrition and exercise in hispanics: a critical examination of existing literature. Health Promot Pract 2010;11:541–554.

12.

Stolley MR, Sharp LK, Oh A, Schiffer L. A weight loss intervention for African American breast cancer survivors, 2006. Prev Chronic Dis 2009;6:A22.

13.

Elhai JD, Grubaugh AL, Richardson JD, Egede LE, Creamer M. Outpatient medical and mental healthcare utilization models among military veterans:

results from the 2001 National Survey of Veterans. J Psychiatr Res 2008;42:858–867.

14.

Jones LR, Wilson CI, Wadden TA. Lifestyle modification in the treatment of obesity: an educational challenge and opportunity. Clin Pharmacol Ther 2007;81:776–779.

15.

Glasgow RE, Lichtenstein E, Marcus AC. Why don’t we see more translation of health promotion research to practice? Rethinking the efficacy-to- effectiveness transition. Am J Public Health 2003;93:1261–1267.