LAUSR

The American Heart Association has encouraged networks research focused on cardiovascular disease and its risk factors, such as obesity. However, little network research has focused on minorities or low-income populations. Our objective was to characterize the relationship between body mass index (BMI) with social network overweight/obesity among public housing residents in Baltimore, MD - a predominantly black, low-income group. We conducted a cross-sectional survey of randomly selected public housing residences (8/2014–8/2015). Adults had their height and weight measured and reported their network members' weight statuses using pictograms. Our dependent variable was respondents' BMI, and independent variable was perceived exposure to overweight/obesity in the social network. We also explored network exposure to overweight/obesity among 1) family members and 2) friends. We used multivariable linear regression adjusted for significant covariates. Our sample included 255 adults with mean age of 44.4 years, 85.5% women, 95.7% black, and mean BMI of 33.2 kg/m2. Most network members were overweight/obese (56.1%). For every 1% increase in network exposure to overweight/obesity, individuals' BMI decreased by 0.05 kg/m2 (p = 0.06). As network exposure to overweight/obesity among friends increased, individuals' BMI significantly decreased by 0.06 kg/m2 (p = 0.04). There was no significant relationship between BMI and network exposure to overweight/obesity among family members. In conclusion, among Baltimore public housing residents, a statistically significant, inverse association existed between individuals' BMI and overweight/obesity among friends in their social networks. Our results differ from relationships seen in prior studies of other populations, which may be due to racial and/or contextual differences between studies.