AbstractBackgroundProgression of fibrillar tau is a key driver of neurodegeneration and cognitive decline in Alzheimer’s disease (AD), however which factors predict the rate of tau accumulation at the patient‐level is unclear. Here we propose to employ a polygenic score (PGS) for the prediction of tau accumulation and cognitive changes in AD, which could be useful for selection of individuals with faster tau progression in clinical trials.MethodsWe included 231 ADNI participants with longitudinal measurements of tau‐PET, amyloid‐PET, cognitive data and genotype data. We computed a PGS based on 85 independent lead SNPs from two recent large GWAS (DOI:10.1101/2020.10.01.20200659; DOI:10.1038/s41588‐021‐00921‐z), excluding all APOE variants. Using linear mixed effect models, we computed the individual rates of change in each of the biomarkers. In linear regression models we tested whether PGS can predict the rate of tau‐PET changes in Braak‐stage ROIs (Figure 1A) and cognitive changes. In order to assess whether amyloid mediates the effect of PGS on tau accumulation we performed sensitivity analyses in subgroups categorized by amyloid status as well as tested the interaction between the PGS and global amyloid‐PET. We further estimated sample size required for detection of hypothetical treatment effect on the rate tau accumulation using power analysis.ResultsHigher PGS was associated with higher accumulation rates of tau‐PET in cortical regions (Braak 3+4:β=0.306, pFDR<0.001; Braak 5+6:β=0.262, pFDR<0.001; Figure 1B). A higher PGS was further associated with faster decline in episodic memory (β=‐0.223, pFDR<0.001; Figure 1C) and global cognition (β=0.280, pFDR<0.001; Figure 1C). The observed effects of PGS on cognitive changes were mediated by higher tau‐PET accumulation (Figure 2). We found a synergistic effect between PGS and elevated amyloid‐PET levels on the rate of tau‐PET accumulation such that the PGS effects on tau‐PET accumulation were stronger particularly at elevated amyloid levels (Figure 3). Selection of individuals with highest PGS yielded a 32‐33% saving in required sample size to detect a treatment effect on tau accumulation.ConclusionHigher polygenic score is associated with faster tau accumulation and may be a useful tool for risk stratification in disease‐modifying treatments on tau.