Longitudinal associations of neighborhood environment features with pediatric body mass index.

Abstract

It has been posited that policies to promote child health and prevent obesity should target neighborhood environments but evidence on the impact of neighborhoods on child weight is conflicting and longitudinal studies (which have benefits for causal inference) are scarce. We used electronic health records (2007-2016) from an urban, pediatric integrated delivery system and linked children (N=51,873, ages 6-19 years, 77% African American) to neighborhood-level data to investigate how changes in neighborhood environments relate to changes in body mass index (BMI). Measures of neighborhood environment were resources for healthy foods and physical activity ('resources'), greenness, violent crime rate, perceived safety and social cohesion. Fixed effects models estimated associations between changes in neighborhood environment exposures and changes in BMI z-score and whether effects differed by sex, baseline age, neighborhood socioeconomic status and population density. Approximately 22% of the cohort was obese (BMI z-score ≥ 95th percentile). In adjusted models, increases in neighborhood greenness and perceived safety were associated with decreases in BMI z-score (mean change in BMI z-score for 1-SD increase for both: -0.012; 95% CI= (-0.018, -0.007)). Increases in neighborhood safety had a stronger effect in children ages 6-10 years than in older children. Increases in social cohesion were associated with increases in BMI z-score (mean change: 0.005 95% CI = (0.003, 0.008)) especially in boys. Increases in food and physical activity resources were not associated with changes in BMI. This study suggests that increasing neighborhood greenness and safety are potential approaches to reduce children's BMI.