Longitudinal associations between residential density and body mass index: The mediating role of walking for transport and the moderating effect of neighbourhood disadvantage

Background: Longitudinal evidence regarding the associations between residential density, walking for transport (WfT) and body mass index (BMI) is limited, and how these relationships vary across different neighbourhood socioeconomic contexts is not yet known. Therefore, we examined longitudinal associations between residential density and BMI: the mediating role of WfT and the moderating effects of neighbourhood disadvantage. Methods: Data from the HABITAT (How Areas in Brisbane Influence HealTh and AcTivity) multilevel longitudinal (2007-2016) study of 11 035 middle-aged adult residents of 200 neighbourhoods in Brisbane, Australia, were used. Residential density was objectively assessed within a 1 km network buffer around residents' homes, and BMI was calculated from self-reported height and weight. Participants self-reported their previous week's minutes of WfT. Generalised structural equation modelling was used to estimate the mediating effect of WfT in explaining residential density-BMI associations across the quintiles of neighbourhood disadvantage, adjusting for time-varying and time-invariant confounders, including distance to Brisbane Central Business District (CBD) and residential self-selection. Results: Residents living in high-density neighbourhoods were found to engage in more WfT and maintain a lower BMI. WfT mediated approximately 6.44% of the density-BMI relationships. These longitudinal relationships remained consistent in moderately disadvantaged neighbourhoods but differed in areas with other levels of disadvantage and when accounting for CBD distance. Conclusion: Residentially dense neighbourhoods facilitate transport-related walking, potentially reducing the risk of weight-related chronic conditions. However, these effects vary across different neighbourhood socioeconomic contexts. Urban planning and policy efforts should focus on context-specific measures to leverage the anticipated benefits of residential density.