Causal Mediation Analysis of the Timing of Risk Factors for Incident Cognitive Impairment

• 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
• Author(s): Nicole Armstrong*
• Companies: Alpert Medical School of Brown University
• Keywords: causal mediation; incident cognitive impairment; cardiovascular disease; depression; counterfactual; time-to-event

Abstract

Cardiovascular disease (CVD) and depression are risk factors for incident cognitive impairment. Given their bidirectional relationship, the examination of subclinical CVD, marker of vascular damage, prior to the development of clinical CVD may elucidate the associations among CVD, depression, and incident cognitive impairment. This study examines the mediating role of late-life depressive symptoms in the association of subclinical CVD with cognitive impairment among those without clinical CVD. We implemented the counterfactual approach to causal mediation using time-to-event data from the Cardiovascular Health Study. Two-year lags were introduced between subclinical CVD and depressive symptoms as well as between depressive symptoms and incident cognitive impairment. The total effect was decomposed into direct and indirect effects (via late-life depressive symptoms), obtained from an accelerated failure time model with weights derived from multivariable logistic regression of late-life depressive symptoms on subclinical CVD. The findings suggest that late-life depressive symptoms partially mediate the association of subclinical CVD with onset of incident cognitive impairment.