Socioeconomic and geographic patterning of under- and overnutrition among women in Bangladesh.

In Bangladesh, the prevalence of overweight among adults is increasing while underweight continues to be common. However, little is known about the pattern of underweight and overweight within Bangladesh and at the neighborhood level. The objective of this study was to assess the socioeconomic and geographic patterning of underweight and overweight in the population and determine if the burdens of these nutritional disorders coexist within neighborhoods in Bangladesh. A nationally representative sample of 10,589 ever-married women aged 15-49 y from 361 neighborhoods in Bangladesh was drawn from the 2004 Bangladesh Demographic and Health Survey. BMI (in kg/m(2)) was used to model nutritional status in a multinomial regression model with women classified as underweight (<18.5 kg/m(2)), overweight (≥25 kg/m(2)), or normal (18.5-24.9 kg/m(2)). Indicators of socioeconomic status and geography included household wealth, neighborhood wealth, and place of residence. Household wealth was related negatively to underweight (OR = 0.35 [95% credible interval (int) = 0.28-0.43] for the richest one-fifth vs. the poorest one-fifth) and positively to overweight [OR = 4.36 (95% int = 2.94-6.57) for the richest one-fifth vs. the poorest one-fifth] in a graded fashion. Neighborhood wealth was positively associated with overweight [OR = 1.75 (95% int = 1.25-2.44) for the top one-third vs. the lowest one-third] and negatively associated with underweight [OR = 0.81 (95% int = 0.69-0.96) for the top one-third vs. the lowest one-third]. Residence in rural neighborhoods was significantly associated with decreased levels of overweight [OR = 0.71 (95% int = 0.58-0.91)]. We observed an inverse relationship between the random effects associated with underweight and overweight at the neighborhood level (r = -0.66; P = 0.008). In conclusion, our results suggest burdens of underweight and overweight in Bangladesh are strongly related to individual socioeconomic position but geographically distinct. Neighborhoods where women were at a higher risk of being underweight were more likely to be those where women were at a lower risk of being overweight.