Date of Award

Fall 1-8-2016

Degree Type


Degree Name

Master of Public Health (MPH)


Public Health

First Advisor

Richard Rothenberg, MD, MPH

Second Advisor

Lawrence S. Phillips, MD


INTRODUCTION: The prevalence of obesity is increasing in the United States and globally, and impacts many aspects of health. To understand the contribution of body weight to chronic diseases such as diabetes, it is necessary to characterize secular and longitudinal weight trends prior to evaluating the weight effects that may result from medical interventions. The cross-sectional National Health and Nutrition Examination Survey (NHANES) indicates that mean body weight in the adult population increased from 152 lb (69 kg) to 181 lb (82 kg) between 1959 and 2008. However, there are no previously published studies on secular or longitudinal weight trends in a veteran population.

OBJECTIVES: The purpose of this study is to describe secular and longitudinal trends in body weight for a large population of male and female individuals with and without diabetes in the Veterans Administration (VA) healthcare system, the largest integrated healthcare system in the United States.

METHODS: Retrospective observational analysis of data from VA facilities throughout the United States, in patients who had at least 4 outpatient visits within any consecutive 4-year interval during 2000–2014. The dataset included men and women with and without type 2 diabetes. The primary outcomes were longitudinal trends in body weight stratified by birth cohort, sex, and diabetes status.

RESULTS: A total of 4,680,735 unique patients, 1,666,346 with diabetes, were included in the analysis. Regressions were performed on the patient-level data and segmented by birth cohort. A total of 176,034,543 weight observations were included in the analysis, with a median of 15 to 36 weight observations per patient in individuals without diabetes, and a median of 22 to 49 weight observations in individuals with diabetes across birth cohorts. In the year 2000, the y-intercept for the regression equations indicated a mean body weight for men without diabetes of 188 lb (85 kg), for women without diabetes of 166 lb (75 kg), for men with diabetes of 213 lb (97 kg), and for women with diabetes of 195 lb (88 kg). Secular trends in body weight during the study period had median linear increases of 0.53 lb/y (0.24 kg/y) in men with diabetes, 0.50 lb/y (0.23 kg/y) in women with diabetes, 0.53 lb/y (0.24 kg/y) in men without diabetes, and 0.86 lb/y (0.39 kg/y) among women without diabetes, respectively. In cohorts born before 1940, body weight decreased. In the cohorts born between 1940-1949, body weight was stable. In all cohorts born after 1950, body weight increased. Across birth cohorts, the rate of weight increase accelerated from older to younger groups, with higher rates in the groups with diabetes than in the groups without diabetes: β2 = 0.0260 lb2/y (0.01179 kg2/y) in men without diabetes, 0.0398 lb2/y (0.01805 kg2/y) in men with diabetes, 0.0127 lb2/y (0.00576 kg2/y) in women without diabetes, and 0.0895 lb2/y (0.04060 kg2/y) in women with diabetes.

CONCLUSIONS: This is the first report of secular and longitudinal weight trends in a large, contemporary veteran population that includes both men and women. Consistent with findings from the Normative Aging Study, a longitudinal study of male veterans from the northeastern United States, weight changes varied from decreases among the oldest birth cohorts to increases in the youngest birth cohorts. Secular changes in body weight by birth cohort were consistent with the patterns reported in the Global Burden of Disease Study. The rate of weight change is accelerated in all younger birth cohorts relative to all older birth cohorts, with the highest rates in women with diabetes. Further analyses of this dataset are recommended to elucidate clinical characteristics associated with longitudinal weight change among individuals with and without diabetes in the veteran population.