LAUSR

Introduction Although neighbourhood may predict late-life cognitive function, studies mostly relies on measurements at a single time point, with few investigations applying a life-course approach. Further, it is unclear whether the associations between neighbourhood and cognitive tests scores relate to specific cognitive domains or general ability. This study explored how neighbourhood deprivation across eight decades contributes to late-life cognitive function. Methods Data were drawn from the Lothian Birth Cohort 1936 (n=1091) with cognitive function measured through 10 tests at ages 70, 73, 76, 79 and 82. Participants residential history was gathered with lifegrid questionnaires and linked to neighbourhood deprivation in childhood, young adulthood, and mid-to-late adulthood. Associations were tested with latent growth curve models for levels and slopes of general (g) and domain-specific abilities (visuospatial ability, memory, and processing speed), life-course associations were explored with path analysis. Results Higher mid-to-late adulthood neighbourhood deprivation was associated with lower age 70 levels (beta=-0.113, 95%CI: -0.205, -0.021) and faster decline of g over 12 years (beta=-0.160, 95%CI: -0.290, -0.031). Initially-apparent findings with domain-specific cognitive functions (e.g. processing speed) were due to their shared variance with g. Path analyses suggested that childhood neighbourhood disadvantage is indirectly linked to late-life cognitive function through education and selective residential mobility. Conclusions To our knowledge, we provide the most comprehensive assessment of the relationship between life-course neighbourhood deprivation and cognitive ageing. Living in advantaged areas in mid-to-late adulthood may contribute to better cognitive function and slower decline, whereas advantaged childhood neighbourhood environment likely affects functioning through cognitive reserves.