Urbanicity, Neighborhood Conditions, and Dementia Mortality.

Importance: As populations age and cities become more densely populated, urbanicity is transforming neighborhood environments that influence healthy aging. However, how dementia mortality varies across the urbanicity gradient and how neighborhood conditions contribute to this pattern remain unknown. Objective: To examine dose-response associations of urbanicity with dementia mortality, quantify the extent to which neighborhood socioeconomic and environmental factors attenuate this pattern, and estimate potential deaths averted under feasible improvements in modifiable conditions. Design, Setting, and Participants: This prospective cohort study used person-level mortality data linked to the national census across 32 844 lower super output areas (ie, small census-based geographic areas) in England from March 2011 to February 2023. Individuals aged 16 years and older with National Health Service general practice registration in England were included. Data were analyzed from July 2025 to January 2026. Exposure: Participants' residential addresses were linked to UK Census-based social and environmental data. Urbanicity was defined as neighborhood population density (persons per hectare), and neighborhood conditions included income, employment, education, crime, living environment, housing, and service accessibility. Main Outcomes and Measures: The primary outcome was mortality with dementia as the underlying cause, and the secondary outcome was mortality with dementia mentioned anywhere on the death certificate. Dose-response associations were examined using Cox proportional hazards models with hazard ratios and 95% CIs. Contribution of neighborhood conditions to these associations was assessed using the percentage of excess risk attenuated. Potential death averted under feasible improvements were estimated using parametric g-computation. Results: A total of 40 948 445 individuals (mean [SD] age, 46.9 [19.3] years, 21 263 986 females [51.9%]; 1 190 069 Black [2.9%]; 1 091 312 Indian [2.7%]; 35 973 092 White [87.9%]) were followed up for a median (IQR) of 8.68 (4.47-11.90) years. There were 5 309 719 deaths, including 621 756 underlying-cause and 926 502 any-mention dementia deaths. Dementia mortality showed a curvilinear inverted U-shaped association with urbanicity, peaking at mid-urbanicity (around 20 to 40 persons per hectare) and declining at both extremes. This pattern was consistent across subgroups, dementia subtypes, and all-cause mortality. For mortality with dementia as the underlying cause, adjustment for all factors largely eliminated the gradient (≥71% attenuated), with accessibility and living environment contributing the most. Scenario analyses shifting both service accessibility and outdoor living environment from the lowest to the second quintile estimated 65 572 (95% CI, 42 036-88 783) dementia deaths averted (10.5% reduction), compared with 43 452 (95% CI, 24 214-62 775) deaths averted (7.0% reduction) for accessibility-only, and 22 700 (95% CI, 11 086-34 201) deaths averted (3.7% reduction) for living environment-only improvements. Estimated gains were larger in males, adults aged 45 to 54 years, residents of private households (vs residents of care homes), and Black residents living under a scenario of improvement to both accessibility and environment. Conclusions and Relevance: In this cohort study, dementia mortality risk was greatest in mid-urbanicity and lower in dense urban cores and rural areas, and this gradient was largely attenuated by neighborhood conditions. These findings suggest that targeted improvements in modifiable neighborhood conditions may substantially reduce dementia mortality and advance equity, particularly in vulnerable communities.