A method for measuring neighborhood conditions across the life course using self-reported addresses

Understanding how neighborhood environments shape health across the life course requires historically grounded data on where people lived and neighborhood conditions at the time of residence. The Think PHRESH study examines how neighborhood contexts across the life course influence risk for Alzheimer’s disease and related dementias among adults in two predominantly Black neighborhoods. Most prior studies have relied on prospective cohorts or commercial address databases, which often provide limited temporal coverage and underrepresent Black populations. This paper describes methods used in Think PHRESH to collect retrospectively self-reported residential histories, process and geocode life-course addresses, and link them to historical census tract data dating back to 1940. We analyzed data from 814 predominantly Black adults enrolled in Think PHRESH. Participants reported addresses where they lived across their life course. Reported addresses were geocoded and linked to contemporaneous census tract data. Four tract-level neighborhood measures were derived for each address and time period: neighborhood deprivation, residential instability, segregation, and housing conditions. We calculated the coverage rate, or proportion of each participant’s life covered by usable geocoded addresses linked to census data within specific life periods (e.g., childhood). We examined variation in coverage by participant sociodemographic characteristics and by historical period. Retrospective self-report yielded substantial life-course coverage of residential and neighborhood data. Overall, 78% of participants’ lives were covered by usable addresses linked to census tract characteristics, a rate comparable to methods based on commercial databases or prospective surveys with narrower observation windows. Coverage was higher among men, younger participants, those with higher educational attainment and income, or those employed. Coverage rates were higher for more recent addresses than for earlier addresses. Across the life course, participants lived in neighborhoods with higher deprivation and segregation than national averages, while residential instability and adverse housing conditions were more pronounced in later life periods. Retrospectively self-reported residential histories linked to historical census data provide a feasible method for reconstructing life-course neighborhood exposures in underrepresented populations. This approach supports longitudinal assessment of changing neighborhood conditions over decades and research on how cumulative social and physical environments may influence late-life health outcomes.