Ageing Research Reviews · Nov 19, 2026

Genetic Variants and Domain-Specific Cognitive Function in the Alzheimer's Disease Continuum: A Systematic Review

By Lee D, Kim JP, Nho K et al.

Background APOE ε4 is the strongest common genetic risk factor for late-onset Alzheimer's disease and is associated with age-related cognitive decline. Whether its association with cognition, and that of other candidate variants, is uniform across cognitive domains or domain-specific remains unclear. Objective To synthesize single-variant candidate-gene evidence on domain-specific cognitive function (episodic memory, executive function, language, visuospatial function, attention, processing speed) in the Alzheimer's disease-indexed literature, beyond APOE alone. Methods We searched PubMed, Embase, and Scopus on 18 June 2026 without date restriction (PRISMA 2020; PROSPERO CRD420261446003). Of 6,453 records, we included 111 studies; all search strings required an Alzheimer's disease term. We applied two risk-of-bias tools to every study. Two reviewers extracted and appraised data using AI assistance and verified them against the source text, without reporting agreement statistics. Synthesis was narrative by domain, with an exploratory effect-size meta-analysis for APOE. Results APOE was examined in 70 of 111 studies (63%), and 30 further genes were examined in two or more studies each. Detrimental APOE associations were most consistent for episodic memory (48 of 75 findings; sign test p=3.3×10⁻⁹) and executive function (28 of 54; p=5.1×10⁻⁴). The direction of association did not differ between domains (χ²(5)=2.13, p=0.83; 84% detrimental), but detection frequency did (null results 28% to 65%; χ²(5)=20.16, p=0.001). In an exploratory meta-analysis of 36 studies, pooled ε4 effects on performance were small and negative in all six domains, significantly so only for episodic memory (Hedges' g=−0.14, 95% CI −0.21 to −0.07; executive function −0.07, −0.13 to 0.00), and within the same studies the episodic memory effect exceeded the executive function and language effects. The episodic memory result was robust to square-root sample-size weighting, age restriction and significance-threshold re-analysis; the executive function result was less robust, and the two risk-of-bias tools disagreed on whether either survives quality stratification. Conclusions The direction of APOE's association with cognition appears uniform across domains; it is detected most often on episodic memory measures, where its standardized effect also exceeds that on executive function and language measures, and part of the between-domain variation in detection, and possibly in effect size, may reflect the instruments used. Evidence beyond APOE is fragmented, with 93 of 123 non-APOE genes (76%) supported by a single study. Single-variant evidence from the Alzheimer's disease literature thus converges on APOE, with only preliminary evidence for a few other Alzheimer's disease risk loci, in contrast to the highly polygenic architecture that genome-wide studies describe for general cognitive function.

doi.org/10.1016/j.arr.2026.103448PubMed 43057396Populations

AgingResearch.News