Disconnected Neighborhoods: Segregated Interactions and Cognitive Health in Middle-Aged and Older Adults

Residential segregation constrains interaction across social lines, operating as an established place-based risk factor for health. Yet, studies conventionally operationalize segregation through static residential composition rather than active patterns of interaction. This study examines segregation by capturing both residents’ characteristics and their movement through broader neighborhood networks, as inferred from collective travel patterns. We use the Spatial Preference in Local Interaction (SPLIT) index – a mobility-based, tract-level measure of cross-neighborhood segregation – and link it to American Community Survey data on tract-level prevalence of cognitive impairment among middle-aged and older adults from 2018 to 2023. We find that cognitive impairment is significantly lower in tracts where mobility networks are economically diverse yet racially and ethnically homogeneous – a pattern consistent with the protective effects of in-group cohesion (bonding) and cross-area resource acquisition (bridging). These findings suggest that the cognitive consequences of cross-neighborhood interaction resist simple characterization, underscoring how residential ties regulate access to both proximate community supports and wider economic opportunity structures.