Abstract
BACKGROUND AND OBJECTIVES: Although neighborhood environments are increasingly recognized as determinants of cognitive aging, less is known about how home and neighborhood conditions jointly influence dementia in later life or whether these associations differ by gender. RESEARCH DESIGN AND METHODS: Guided by ecological models of aging, I analyzed 9 years (2015-2021) of National Health and Aging Trends Study data from 5,389 community-dwelling adults aged ≥65. Survey-weighted multilevel mixed-effects logistic models estimated 4 models: home-only (indoor and exterior hazards), neighborhood-only (social cohesion and physical disorder), joint (home and neighborhood), and interactions with gender. RESULTS: At baseline, 14.0% met criteria for probable/possible dementia; 10.6% had indoor hazards; and 48.5% exterior hazards. In adjusted models, indoor hazards were associated with higher odds of dementia (OR = 1.71, 95% CI: 1.31-2.23), whereas exterior hazards were not. Neighborhood physical disorder predicted dementia in neighborhood-only models (OR = 1.18, CI: 1.05-1.34), but this association became nonsignificant after including home hazards. Women had lower overall odds of dementia, yet the indoor hazards × gender interaction was significant (OR = 1.75, CI: 1.01-2.05), indicating stronger risk among women. Lagged analyses reinforced temporality: indoor hazards at t - 1 predicted next-wave dementia status with stronger effect sizes, while social cohesion emerged as protective only in lagged models. DISCUSSION AND IMPLICATIONS: Findings underscore the proximal salience of the dwelling for late-life cognition and highlight gendered exposure patterns. Systematic screening for indoor hazards, particularly for older women, could strengthen aging-in-place strategies. Neighborhood investments to reduce visible disorder may complement but not replace direct home-level interventions to support cognitive health.