To evaluate plasma p-tau217 ability to detect pre-symptomatic Alzheimer disease (AD) in ethnoracially diverse cognitively normal individuals.

Plasma p-tau217 demonstrates strong performance in identifying AD neuropathology in symptomatic individuals. However, its performance in detecting pre-symptomatic AD remains unknown. Furthermore, cutoffs validated for symptomatic AD may not be suitable for pre-symptomatic disease.

The cohort comprised 58 Black (30%), 46 Hispanic/Latino (23%), and 93 non-Hispanic White participants (47%). Non-Hispanic White participants were older (75.9±9.5 years, p<0.001; 42% female) than Black (65.0±9.4 years; 64% female) and Hispanic/Latino participants (64.2±8.1 years; 61% female). Given similar demographics and biomarker positivity (amyloid and tau PETs, p-tau217), Black and Hispanic/Latino participants were merged. Plasma p-tau217 concentrations increased with age (Black and Hispanic/Latino: rho=0.252, p=0.010; non-Hispanic White: rho=0.552, p<0.001). P-tau217 correlated with amyloid Centiloids, tau SUVR, atrophy, and white matter disease in non-Hispanic White participants (rho=0.419; 0.500; -0.268; 0.342, all p≤0.001), with directionally consistent, but non-significant correlations in Black and Hispanic/Latino participants. An optimized cutoff (≥0.132 pg/mL) outperformed the established symptomatic cutoff (≥0.186 pg/mL) with higher sensitivity (Black: 83% vs 66%; Hispanic/Latino: 100% vs 66%; non-Hispanic White: 82% vs 57%; merged: 84% vs 60%) but lower specificity (87% vs 93%; 78% vs 93%; 62% vs 80%; 72% vs 88%).

Plasma p-tau217 performed similarly across ethnoracial cohorts in detecting at-risk individuals. An optimized threshold may improve preclinical AD detection and support representative enrollment in prevention trials.