Abstract

OBJECTIVE: To evaluate whether white matter hyperintensities (WMH) and apolipoprotein E (APOE) ε4 status have an additive or multiplicative effect on the risk of incident all-cause dementia. METHODS: We conducted a prospective cohort study in the Atherosclerosis Risk in Communities (ARIC) study and confirmed findings in the UK Biobank (UKB). The exposures were APOE ε4 status (0 vs. ≥1 allele) and WMH on magnetic resonance imaging (MRI). The primary outcome was incident all-cause dementia. After confirming an additive interaction, we created combined exposure groups: WMH-/ε4-, WMH+/ε4-, WMH-/ε4+, and WMH+/ε4+. Cox proportional hazards models were adjusted for age, sex, race, education, cognition (ARIC only), hypertension, diabetes, and prior stroke. RESULTS: In ARIC (n = 1,736, mean age 63, 58.8% female, 48.7% non-Hispanic White individuals, median follow-up 18.6 years), the dementia incidence rate was 10.4 (95% CI, 9.2-11.6) per 1,000 person-years. Compared to WMH-/ε4-, adjusted hazard ratios (HRs) for dementia were: WMH-/ε4+, 1.5 (95% CI, 1.1-2.1); WMH+/ε4-, 2.0 (95% CI, 1.4-2.7); and WMH+/ε4+, 3.2 (95% CI, 2.2-4.6). In UKB (n = 40,307, mean age 55, 52.7% female, 97.1% non-Hispanic White individuals, median follow-up 3.2 years), the dementia incidence rate was 0.42 (95% CI, 0.32-0.55) per 1,000 person-years. Adjusted HRs were: WMH-/ε4+, 2.3 (95% CI, 1.2-4.5); WMH+/ε4-, 2.1 (95% CI, 1.0-4.6); and WMH+/ε4+, 6.7 (95% CI, 3.2-13.9). INTERPRETATION: WMH burden and APOE ε4 status additively increase dementia risk. These findings support the potential benefit of vascular risk management to reduce WMH and delay dementia onset, even among genetically at-risk individuals. ANN NEUROL 2026;99:656-667.