Plasma phosphorylated tau 217 concentrations, APOE genotype, and timing of cognitive impairment in individuals across diverse racial and ethnic groups: a pooled analysis of prospective cohort studies.

BACKGROUND: Plasma tau phosphorylated at threonine 217 (p-tau217), a biomarker of Alzheimer's disease neuropathological changes, can increase before overt symptoms occur. However, individuals with similar p-tau217 concentrations but different genetic backgrounds might differ in their risk or timing of cognitive decline. We aimed to determine whether APOE genotype provides additional prognostic information beyond plasma p-tau217 concentration for estimating the risk and timing of cognitive impairment. METHODS: We did a pooled analysis of participant-level data from seven multi-ethnic prospective cohorts of older adults, including individuals with no cognitive impairment at increased risk of dementia and individuals with mild cognitive impairment or dementia, recruited through academic research centres and community-based settings in Canada, Dominican Republic, and USA. Data were collected from 1992 to 2025. This study included data from participants with plasma p-tau217 measurements, clinical cognitive status, APOE genotype, and complete covariate data. The primary outcome was cognitive impairment, defined as mild cognitive impairment or dementia. Baseline analyses included all eligible participants; longitudinal analyses were restricted to participants with no cognitive impairment at baseline and with at least one follow-up clinical assessment. We evaluated whether APOE-ε4 carrier status modified the risk and timing of cognitive impairment associated with plasma p-tau217 concentration modelled as a continuous variable. Associations of baseline p-tau217 concentration with prevalent and incident cognitive impairment were assessed using logistic regression and Cox models, stratified by APOE-ε4 and with interaction terms. Timing and prognostic performance were evaluated using survival analyses, discrimination measures, and random survival forests. FINDINGS: Among 8873 participants with available plasma p-tau217 concentration and clinical data who were examined for eligibility, 8582 participants (5652 [65·9%] female and 2930 [34·1%] male) were included in the baseline analysis. Mean age was 70 years (SD 10), and 1380 (16·1%) were Black, 4056 (47·3%) were non-Hispanic White, and 3146 (36·7%) were of Hispanic or other ethnic origin. 4569 (53·2%) participants had no cognitive impairment at baseline and were included in the longitudinal analysis. At baseline, higher p-tau217 concentrations were associated with cognitive impairment (odds ratio 1·77, 95% CI 1·42-2·20) and this association was stronger in APOE-ε4 carriers (2·25, 1·52-3·34) than in non-carriers (1·52, 1·35-1·72). Higher p-tau217 concentrations at baseline were also associated with incident cognitive impairment (hazard ratio 1·41, 95% CI 1·22-1·64; for 1-SD increase of p-tau217 levels), with a stronger association in APOE-ε4 carriers (1·76, 1·36-2·26) than in non-carriers (1·26, 1·12-1·42). Each 1-SD increase in p-tau217 concentration was associated with a 24% shorter time to cognitive impairment among APOE-ε4 carriers, compared with 13% among non-carriers. Differences in cognitive-impairment-free survival by p-tau217 concentration emerged 3-4 years after biomarker assessment and before symptom onset. INTERPRETATION: Plasma p-tau217 concentrations, considered together with APOE genotype, might help stratify risk and estimate the timing of future cognitive impairment in asymptomatic individuals who are at increased risk for Alzheimer's disease, such as those with a family history of dementia or known APOE-ε4 carrier status. These findings support the use of biomarker-informed approaches to guide monitoring and therapeutic intervention in appropriately selected at-risk individuals before the onset of clinical symptoms. FUNDING: National Institutes of Health.