Urban Mobility and Population Health in The Hague: A Multi-Scale Analysis
In the authors' words
Urban mobility may provide insight into population health by reflecting how residents connect to services, resources, and urban systems. However, mobility patterns are also shaped by contextual factors, including socioeconomic, environmental, behavioural, and demographic conditions, complicating interpretation of mobility- health associations, particularly when mobility and health are measured at different spatial scales. This ecological cross-sectional study examined associations between mobile-phone-derived mobility and population health in The Hague, the Netherlands, from January to July 2019. Mobility was measured as mean outgoing distance across eight regions, while contextual and health indicators were primarily available for 44 neighbourhoods. Health outcomes included cardiometabolic medication use, polypharmacy, and syndemic-based disease burden. Using a residual-based adjustment framework, mobility and health outcomes were modelled separately using the same contextual domains, and residual correlations assessed whether mobility retained health-relevant variation beyond shared contextual conditions. Outgoing mobility varied substantially across regions and was strongly patterned by contextual factors. At the regional level, higher-than-expected mobility tended to coincide with lower-than-expected disease burden after contextual adjustment ({rho} = -0.26 to -0.48). In cross-scale neighbourhood analyses, however, residual correlations were considerably weaker ({rho} = -0.11 to 0.10). Cluster-bootstrap 95% confidence intervals included zero for all outcomes, and sensitivity analyses yielded similar results. These findings suggest that aggregated mobility is better interpreted as a spatially and contextually embedded indicator of urban connectivity and population composition than as a standalone indicator of population health. The contrast between regional and neighbourhood-level findings highlights how geographic units can influence interpretation of mobility-health relationships.
Appeared: Thursday, September 24. medRxiv. Preprint, not yet peer-reviewed.