Cross-Sectional and Longitudinal Associations of BMI with Socioeconomic Characteristics


  • The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Department of Epidemiology, 1214 South University, Ann Arbor MI 48103. E-mail:


Objective: To examine cross sectional and longitudinal associations of socioeconomic position and neighborhood environments with BMI in a middle-aged and bi-ethnic cohort.

Research Methods and Procedures: Analyses were based on 13, 167 subjects (45 to 64 years) who participated in the Atherosclerosis Risk in Communities Study, a population-based study. Census block groups were used as proxies for neighborhoods and were characterized using a summary socioeconomic score. BMI was measured at baseline and at three follow-up visits over a 9-year period.

Results: Individual and neighborhood socioeconomic characteristics were independently and inversely associated with BMI at baseline in women [mean difference in kilograms per meter squared per unit increase in socioeconomic category (SE) for white and black women respectively; −1.56 (0.14), −1.59 (0.19) for education; −1.07 (0.10), −1.18 (0.18) for income; and −1.04 (0.09), −0.77 (0.18) for neighborhood characteristics]. Results for men were not as consistent. Baseline BMI was negatively associated with income in white men but was positively associated with education, income, and neighborhood characteristics in black men. BMI increased over time regardless of gender or race and in most age groups. In whites, there were no consistently patterned differences in longitudinal trends in BMI by individual or neighborhood socioeconomic characteristics. However, in blacks, there was some evidence of greater increases in the higher socioeconomic status groups.

Discussion: Socioeconomic factors are inversely associated with BMI in middle-aged women, possibly reflecting socially patterned exposures occurring in childhood and adolescence. However, recent increases over time in BMI are either not clearly patterned by socioeconomic factors or are greater in the higher socioeconomic status groups.