Educational Attainment and Risk of All‐Cause Dementia, Vascular Dementia, and Alzheimer’s Disease – A Population‐Based Twin Study

Abstract

AbstractBackgroundMost studies on the association between educational attainment and dementia have been conducted in unrelated individuals. Dementia‐discordant twin pairs can inform if the protective effect of education on dementia is evident when controlling for environmental and genetic effects. We investigated the association of educational attainment on all‐cause dementia, vascular dementia (VaD), and Alzheimer’s disease (AD) mortality in a population‐based sample of dizygotic (DZ) and monozygotic (MZ) pairs.MethodParticipants were from a population‐based longitudinal older Finnish Twin Cohort study with baseline questionnaire in 1975 (N = 29,784; mean age = 36.9±14.9years). Education was coded as <12 years versus ≥12 years. Registry‐based cause of death data until year 2020 was based on the International Classification of Diseases. We ran individual‐level cox proportional hazard models with sex as a covariate and conditional within‐twin‐pair models for education predicting all‐cause dementia, VaD, and AD mortality. We report hazard ratios (HR) with 95% confidence intervals (CI).ResultIn individual‐level analyses higher education was associated with a decreased risk of all‐cause dementia (HR = .935, 95%CI = .934;.936, p<.001) and VaD (HR = .638, 95%CI = .637;.640, p<.001), but with an increased risk of AD (HR = 1.276, 95%CI = 1.275;1.277, p<.001). In co‐twin analyses of discordant DZ and MZ twin pairs, co‐twins with higher education had lower – although non‐significant – risk of all‐cause dementia (N = 52, HR = .688, 95%CI = .319;1.481, p = .339) or AD (N = 36, HR = .600, 95%CI: .218;1.651, p = .323). Co‐twin analyses were not done for VaD due to small number of discordant pairs (N = 6).ConclusionThis large population‐based sample showed that higher education is related to lower risk of all‐cause dementia and VaD mortality. Unexpectedly, higher education was related to higher AD mortality, but discordant twin pair analyses were informative regarding this counterintuitive result. This case‐control design, controlling for genetic and environmental effects, did not support any causality from higher education to higher AD risk. Despite the large sample, twin pairs – especially MZ twins – discordant for dementia are rare. Meta‐ or mega‐analysis of multiple large twin studies is needed to clarify if secondary education is related to lower risk of dementia, VaD, and AD mortality when controlling for genetic effects.