Association between late‐life neighborhood built environment and cognitive impairment in older adults.

Background: Elders' mobility and social activities are closely related to their living environments and cognitive function. Built environments and social activities tend to vary across populations and geographic regions. Therefore, this study aims to explore the association between late‐life neighborhood built environment and cognitive function in community‐dwelling older adults. Methods: This prospective cohort study recruited 605 participants aged 65 or older from senior health checkup program at a tertiary medical center at baseline (2011‐2013) with three biennial follow‐ups. After excluding participants with suspected dementia, stroke, brain tumor, incomplete or changing residential address between 2005 and 2011, 465 participants remained for analyses. Global and domain‐specific cognition (memory, attention, executive function, and language) were assessed at baseline and each follow‐up. Built environment data was obtained from 2005 land use survey in Taiwan. Two buffer zones (500 m and 1000 m) centering at each participant's residential address were created to assess built environment by the following indices: (1) the percentage of public open space, (2) the number of cultural resources, and (3) sports venue. Generalized linear mixed models were used to explore the association between built environment and global or domain‐specific cognitive function adjusting for age, sex, socioeconomic status, apolipoprotein E e4 status, physical activity, physical function and chronic conditions. Results: Over six years, more public open space was associated with better logical memory delaying theme recall (within a 1000‐m buffer: β=2.17), more cultural resources were associated with poor executive function (Trail Making Test A, within a 500‐m buffer: β=‐0.078). For every 1 year increment, more public open space was associated with poor executive function (Trail Making Test B, within a 1000‐m buffer: β=‐0.52), more cultural resources was associated with better executive function (within a 1000‐m buffer: β=0.006) and poor logical memory delayed theme recall (within a 500‐m buffer: β=‐0.019). No significant association was found for sports venues in relation to cognition. Conclusion: We found that different neighborhood built environment was associated with the performance of specific cognitive domains. Findings from this study indicate that choosing proper neighborhood environment may prevent cognitive impairment in the preclinical stage of dementia. [ABSTRACT FROM AUTHOR]