Neighborhood environment profiles related to physical activity and weight status: A latent profile analysis

Background: Neighborhood built environments (BE) include combinations of co-existing stimuli influencing physical activity (PA). Dealing with numerous environmental variables and complexity presents a significant challenge. The current analysis explored whether a range of reported BE features associated with adults' physical activity produced distinct multivariate patterns, and tested whether adults' PA and body mass differed by BE profiles. Methods: Participants (20-65. years, 48.2% female, 26% ethnic minority) were recruited between 2002 and 2005 from 32 neighborhoods from Seattle-King County, WA (N= 1287) and Baltimore, MD-Washington, DC regions (N= 912). Independent Latent Profile Analyses were conducted in each region with 11 environmental variables from the Neighborhood Environment Walkability Scale. Validity of the neighborhood profiles was examined by their relationship to PA (accelerometer-derived moderate-to-vigorous minutes/day, self-reported minutes/week of walking for transportation and leisure) and self-reported BMI using ANCOVA models. Results: Neighborhood profiles for Seattle and Baltimore regions were visually similar, suggesting generalizability. High-walkable recreationally-dense neighborhoods differed significantly from other neighborhood types by as much as 13 MVPA minutes/day, almost 60 minutes/week of walking for transportation, and 75. min/week of leisure-time activity. Neighborhood profiles also differed significantly for BMI. Discussion: These findings could help identify optimal patterns of environmental attributes that facilitate physical activity and improve weight status.