Weight gain in adulthood is now common in many populations, ranging from modest gains in developing countries to a substantial percentage of body weight in some Western societies. To examine the rate of change across the spectrum of low to high-income countries we compared rates of weight change in samples drawn from three countries, Nigeria, Jamaica and the United States.
Population samples from Nigeria (n = 1,242), Jamaica (n = 1,409), and the US (n = 809) were selected during the period 1995–1999 in adults over the age of 19; participation rates in the original survey were 96%, 60%, and 60%, respectively. Weight in (kg) was measured on 3 different occasions, ending in 2005. Multi-level regression models were used to estimate weight change over time and pattern-mixture models were applied to assess the potential effect of missing data on estimates of the model parameters.
The unadjusted weight gain rate (standard error) was 0.34(0.06), 1.26(0.12), 0.34(0.19) kg/year among men and 0.43(0.06), 1.28(0.10), 0.40(0.15) kg/year among women in Nigeria, Jamaica, US, respectively. Regression-adjusted weight change rates were significantly different across country, sex, and baseline BMI. Adjusted weight gain in Nigeria, Jamaica and US was 0.31(0.05), 1.37(.04), and 0.52(0.05) kg/year respectively. Women in Nigeria and the US had higher weight gains than men, with the converse observed among Jamaicans. The obese experienced weight loss across all three samples, whereas the normal weight (BMI < 25) had significant weight gains. Missing data patterns had an effect on the rates of weight change.
Weight change in sample cohorts from a middle-income country was greater than in cohorts from either of the low- or high-income countries. The steep trajectory of weight gain in Jamaica, relative to Nigeria and the US, is most likely attributable to the accelerating effects of the cultural and behavioral shifts which have come to bear on transitional societies.