This retrospective cohort study used the US All of Us Research Program (n=187,341 adults ≥50 years, mean follow-up 10.26 years) to examine whether EHR-documented ADHD diagnosis and ADHD polygenic risk score (PRS) were each independently associated with incident mild cognitive impairment (MCI) or dementia, and whether comorbid ADHD worsened outcomes among those who developed cognitive impairment.
ADHD was associated with a >4-fold higher hazard of MCI or dementia (adjusted HR 4.60; 95% CI 3.99–5.29); high ADHD PRS (vs. low) added a 35% higher hazard (HR 1.35; 95% CI 1.25–1.46). Among those who developed cognitive impairment, comorbid ADHD was linked to ~2-fold higher odds of suboptimal mental health (OR 1.69), quality of life (OR 1.66), and social satisfaction (OR 2.39), plus 20% more total healthcare utilization.
Reverse causation cannot be excluded given the long dementia prodrome — lagged analyses showed the ADHD–cognitive impairment association became non-significant at 9-year minimum intervals. EHR-based diagnosis is subject to underascertainment and differential healthcare access bias. The ADHD PRS was derived predominantly from European-ancestry GWAS data, limiting applicability to other ancestry groups.
Adults with ADHD appear to carry a substantially elevated risk of late-life cognitive impairment, supported by both clinical and genetic evidence; clinicians should consider proactive cognitive monitoring in older patients with ADHD. Those who develop MCI or dementia alongside ADHD need integrated mental health support and careful care coordination, given their greater symptom burden and healthcare use.