Motor speech markers of Alzheimer’s risk

Abstract Motor changes are increasingly recognized as part of the Alzheimer’s disease (AD) phenotype, yet motor speech function and its clinical utility for early AD risk stratification remain underexplored. This cross-sectional study investigated whether measures of motor speech performance in home-based rapid syllable repetition (diadochokinesis [DDK]) tasks are associated with the established AD risk indicators of subjective cognitive decline (SCD), apolipoprotein E ε4 ( APOE ε4) carriage, and plasma phosphorylated tau 181 (p-tau181). A community sample of 339 cognitively unimpaired adults (238 asymptomatic control (AC), 66.3 ± 6.5 years, 71.4% female; 101 SCD, 66.0 ± 6.6 years, 72.3% female) from the Island Study Linking Aging and Neurodegenerative Disease remotely self-administered four DDK tasks (“pa”, “ta”, “ka”, “pataka”) from the self-administered online Tasmanian Test protocol. Self-reported cognitive changes were used to classify AC or SCD groups. We conducted nonparametric analyses and percentile-LASSO multiple regression to predict risk indicators from motor speech performance measures, allowing for age, sex, education, anxiety, and depression. No nonparametric associations with SCD or APOE ε4 remained significant after adjustment for multiple comparisons. One measure of relative repetition variability remained significantly associated with plasma p-tau181. In covariate-adjusted regression analyses, motor speech measures were associated with SCD (AUC 0.700), APOE ε4 (AUC 0.715), and natural log-transformed p-tau181 (MAE 0.323). Sensitivity analyses for hearing loss, incomplete data, and depression-screened subsets revealed potentially mitigating effects of depression and blood biomarker data availability bias. This indicates that additional work, especially longitudinal follow-up, is needed to verify these preliminary findings in the search for scalable screening tools for preclinical dementia risk.