Neighbourhood environments and cognitive health in the longitudinal Personality and Total Health (PATH) through life study: A 12-year follow-up of older Australians

Background: Urban neighbourhood environments may impact older adults' cognitive health. However, longitudinal studies examining key environmental correlates of cognitive health are lacking. We estimated crosssectional and longitudinal associations of neighbourhood built and natural environments and ambient air pollution with multiple cognitive health outcomes in Australian urban dwellers aged 60+ years. Methods: The study included 1160 participants of the PATH Through Life study (60+ cohort) who were followed up for 12 years (four assessments; 2001/02 to 2013/15) and with data on socio-demographics, health, cognitive functions and diagnoses, and full residential address. Neighbourhood environmental features encompassed population and street-intersection densities, non-commercial land use mix, transit points, presence of blue space, percentages of commercial land, parkland and tree cover, and annual average PM2.5 and NO2 concentrations. All exposures except for tree cover were assessed at two time points. Generalised additive mixed models estimated associations of person-level average, and within-person changes in, exposures with cognitive functions. Multistate hidden Markov models estimated the associations of neighbourhood attributes with transitions to/from mild cognitive impairment (MCI). Results: Dense, destination-rich neighbourhoods were associated with a lower likelihood of transition to MCI and reversal to no MCI. Positive cross-sectional and longitudinal associations of non-commercial land use mix, street