Deriving life-course residential histories in brain bank cohorts: A feasibility study.

Abstract

The exposome is theorized to interact with biological mechanisms to influence risk for Alzheimer's disease but is not well-integrated into existing Alzheimer's Disease Research Center (ADRC) brain bank data collection. We apply public data tracing, an iterative, dual abstraction and validation process rooted in rigorous historic archival methods, to develop life-course residential histories for 1254 ADRC decedents. The median percentage of the life course with an address is 78.1% (IQR 24.9); 56.5% of the sample has an address for at least 75% of their life course. Archivists had 89.7% agreement at the address level. This method matched current residential survey methodology 97.4% on average. This novel method demonstrates feasibility, reproducibility, and rigor for historic data collection. To our knowledge, this is the first study to show that public data tracing methods for brain bank decedent residential history development can be used to better integrate the social exposome with biobank specimens. Public data tracing compares favorably to survey-based residential history. Public data tracing is feasible and reproducible between archivists. Archivists achieved 89.7% agreement at the address level. This method identifies residences for nearly 80% of life-years, on average. This novel method enables brain banks to add social characterizations.