The association between adverse childhood experiences, neighborhood greenspace, and body mass index: A cross-sectional study

An association between adverse childhood experiences (ACEs) and elevated body mass index (BMI) has been found in previous investigations. ACEs' effects on BMI have been primarily considered via individual-level physiological and behavioral frameworks. Neighborhood factors, such as greenspace, are also associated with BMI and may merit consideration in studies examining ACEs-BMI associations. This exploratory study examined associations of BMI with ACEs and neighborhood greenspace and tested whether greenspace moderated ACEs-BMI associations. Methods entailed secondary analysis of cross-sectional data. ACEs and BMI were captured from 2012/2013 Philadelphia ACE Survey and 2012 Southeastern Household Heath Survey data; greenspace percentage in participants' (n = 1,679 adults) home neighborhoods was calculated using National Land Cover Database data. Multi-level, multivariable linear regression 1) examined associations between BMI, ACEs, (0 ACEs [reference], 1-3 ACEs, 4 + ACEs), and neighborhood greenspace levels (high [reference], medium, low) and 2) tested whether greenspace moderated the ACEs-BMI association (assessed via additive interaction) before and after controlling for sociodemographic and health-related covariates. Experiencing 4 + ACEs (beta = 1.21; 95 %CI: 0.26, 2.15; p = 0.01), low neighborhood greenspace (beta = 1.51; 95 %CI: 0.67, 2.35; p < 0.01), and medium neighborhood greenspace (beta = 1.37; 95 %CI: 0.52, 2.21; p < 0.01) were associated with BMI in unadjusted models. Only low neighborhood greenspace was associated with BMI (beta = 0.95; 95 %CI: 0.14, 1.75; p = 0.02) in covariate-adjusted models. The ACEs-greenspace interaction was not significant in unadjusted (p = 0.89-0.99) or covariate-adjusted (p = 0.46-0.79) models. In conclusion, when considered simultaneously, low neighborhood greenspace, but not ACEs, was associated with BMI among urban-dwelling adults in covariate-adjusted models.