Differences in magnitude and rates of change in BMI distributions by socioeconomic and geographic factors in Mexico, Colombia, and Peru, 2005-2010.

Abstract

Previous studies about obesity and its associated factors in low- and middle-income countries have been based mostly on women of reproductive age. Furthermore, disproportionally changing BMI distributions have been a challenge for its appropriate modeling. In this context, we assessed the magnitude and rate of change in BMI distribution by socioeconomic and geographic factors in both sexes in Latin American countries, modeling the shape of BMI distributions. We used data from national surveys conducted in Mexico, Colombia, and Peru at two time points between 2005 and 2013 (N = 57,414, 13,5403, and 30,811, respectively). We estimated shapes of BMI distributions for 2005 and 2010, and assessed their changes, using the generalized additive model for location, scale, and shape (GAMLSS), in which BMI was assumed to follow a Box-Cox Power Exponential (BCPE) distribution. In all the three countries, higher education was negatively associated with BMI in women but somewhat positive in men; and household wealth was positively associated in men but not in women. Lower household wealth was associated with higher rates of change in BMI distributions in women. Education and household wealth were associated with BMI distributions and their change over time. Observed sex differences in these associations have implications for designing relevant policies and programs to approach target populations effectively. The BCPE-GAMLSS method can provide a useful visual assessment of time-varying measures.