Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 13 - Advances in Longitudinal Methods in Research on Aging and Dementia from the MEthods for LOngitudinal Studies of DEMentia (MELODEM) Initiative
Type: Invited
Date/Time: Sunday, August 8, 2021 : 1:30 PM to 3:20 PM
Sponsor: Biometrics Section
Abstract #316624
Title: Estimating Heterogenous Effects of Education on Dementia Using Bayesian Trees
Author(s): Jordan Weiss* and Sameer Deshpande
Companies: University of California, Berkeley and Massachusetts Institute of Technology - CSAIL
Keywords: Bayesian trees; Effect modification; Health inequalities; Heterogeneous treatment effects

Educational gradients in dementia are well documented in the United States and have been implicated as an important contributor to racial disparities in the risk of dementia. However, this body of work has focused primarily on assessing average treatment effects across subgroups; relatively little is known about variation within demographic subgroups and how educational attainment may differentially affect dementia risk along a set of established risk factors for dementia. We take an intersectional, Bayesian treed approach together with data from the Health and Retirement Study to (i) estimate treatment effect heterogeneity in the education-dementia gradient across and within subgroups defined by gender, race/ethnicity, and socioeconomic position in a time-varying setting that allows us to (ii) identify intervenable characteristics that may explain variation in the effects of education on dementia across and within these groups. Our findings reveal substantial heterogeneity in how education affects dementia risk, for whom education provides the most benefit, and, in tests of effect modification, how and for whom effect modifiers buffered the ill-effects of low education.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2021 program