Dementia and handedness: Meta-analyses

Abstract

Hemispheric asymmetries are a core feature of brain organization and may influence neurodegenerative processes. Handedness has been proposed as a behavioral marker of lateralization, but its relevance in dementia remains unclear. To determine whether individuals with dementia, including Alzheimer’s disease, differ from healthy controls in the prevalence of left-, mixed-, and non-right-handedness (left- and mixed-handedness), frequentist and Bayesian meta-analytic approaches were used. First, a systematic literature search following PRISMA guidelines was conducted in PubMed and Google Scholar through March 2025 using the terms: “((dementia) OR (Alzheimer's disease)) AND (handedness).” Inclusion criteria required studies to (1) report handedness for both dementia and control groups, (2) not match participants for handedness, and (3) provide sufficient statistical data. Eighteen studies (N = 13,282 dementia patients, N = 18,540 controls) met the criteria. Frequentist meta-analyses showed that individuals with dementia were significantly less likely to be non-right-handed (OR = 0.90, 95 % CI: 0.83–0.97, p = 0.01) and mixed-handed (OR = 0.71, 95 % CI: 0.59–0.85, p = 0.01) compared to controls. However, sensitivity analyses excluding the largest study rendered these associations non-significant, indicating limited robustness. Bayesian meta-analyses supported the null hypothesis with moderate-to-strong evidence across all handedness categories (BF₁₀ = 0.06–1.06). Exploratory Bayesian meta-regressions tested dementia subtype as a moderator and revealed no significant effects. In conclusion, while handedness differences in dementia appear small and sensitive to study design, emerging patterns, particularly in mixed-handedness, may reflect subtle disruptions in hemispheric specialization. Further research may clarify whether such traits provide early behavioral signals of neurodegenerative changes.