Semester

Fall

Date of Graduation

2020

Document Type

Dissertation

Degree Type

PhD

College

School of Public Health

Department

Epidemiology

Committee Chair

Kim E. Innes

Committee Co-Chair

Usha Sambamoorthi

Committee Member

Amna Umer

Committee Member

Christa Lilly

Committee Member

Diane Gross

Abstract

Objective: The purpose of this research was to study the collective and incremental association of common non-cancer chronic pain conditions (NCPC) to risk for Alzheimer’s disease and related dementias (ADRD); to explore the potential mediating influence of mood and sleep disorders; and to assess the potential synergistic effects of NCPC’s and multimorbidity on ADRD risk. This research also aimed to study the contribution of NCPC to ADRD risk in poor, rural, underserved United States populations who suffer a disproportionate burden of ADRD risk factors. To achieve these goals, this dissertation pursued three specific aims: 1) assess the association of common NCPCs to incident ADRD and examine the potential mediating influence of sleep and mood disorders on this association in a nationally representative sample of United States elders; 2) evaluate the relation of common NCPCs to incident ADRD in a large Appalachian population and evaluate the potential mediating effects of sleep and mood disorders on this association; and 3) examine the potential modifying role of multimorbidity in the association of common NCPCs to incident ADRD.

METHODS: The 3 studies employed a retrospective cohort design using data on two US populations of older (≥65), non-institutionalized, fee-for-service Medicare beneficiaries. Data were drawn from two large, population-based data sources: the Medicare current beneficiaries survey (MCBS) linked with the Medicare fee-for-service (FFS) claims database (N=16, 934 adults ADRD-free at baseline, AIMS 1,3); and the West Virginia Medicare insurance database (N=161, 573 adults ADRD -free at baseline, AIM 2). Incident ADRD was ascertained using both survey and claims-based algorithms (Aims 1,3) or claims-based algorithms alone (Aim 2). Analyses were conducted using logistic regression and adjusted for sociodemographics, lifestyle characteristics, medical history, and medication use.

RESULTS: AIM 1: The presence of NCPCs at baseline was significantly and positively associated with incident ADRD after adjustment for potential confounders, with risk for ADRD increasing significantly with rising number of NCPCs. Inclusion of sleep disorders and/or depression/anxiety substantially attenuated these risk estimates, suggesting these conditions may in part mediate the observed associations of NCPC’s to incident ADRD. AIM 2: In this large population of older West Virginia Medicare beneficiaries, presence of any NCPC at baseline was associated with significantly increased risk for incident ADRD after adjustment for covariates; the strength and magnitude of this association rose significantly with increasing number of diagnosed NCPCs. Inclusion of depression/anxiety, but not sleep disorders attenuated these risk estimates, suggesting a partial mediating role of mood disorders. Sensitivity analyses using multinomial logistic regressions accounting for risk of death yielded comparable findings. AIM 3: NCPC accompanied by multimorbidity (vs. absence of NCPC or multimorbidity) was significantly and positively associated to risk for incident ADRD, with ADRD risk more pronounced in those with a greater number of co-occurring chronic conditions.

CONCLUSIONS: In these two large cohorts of US and Appalachian Medicare beneficiaries, NCPCs showed a strong, dose-response association to incident ADRD that may be partially mediated by mood and/or sleep disorders. Risk for incident ADRD was amplified in the presence of concomitant NCPC and multimorbidity, with risk estimates particularly pronounced in elders with a greater number of co-occurring chronic conditions.

Embargo Reason

Publication Pending

Download

Share

COinS