Significance of plasma p-tau217 in predicting long-term dementia risk in older community residents: Insights from machine learning approaches

Zhenxu Xiao,Xiaowen Zhou,Qianhua Zhao,Yang Cao,Ding Ding

DOI: https://doi.org/10.1002/alz.14178

2024-08-08

Abstract:Introduction: Whether plasma biomarkers play roles in predicting incident dementia among the general population is worth exploring. Methods: A total of 1857 baseline dementia-free older adults with follow-ups up to 13.5 years were included from a community-based cohort. The Recursive Feature Elimination (RFE) algorithm aided in feature selection from 90 candidate predictors to construct logistic regression, naive Bayes, bagged trees, and random forest models. Area under the curve (AUC) was used to assess the model performance for predicting incident dementia. Results: During the follow-up of 12,716 person-years, 207 participants developed dementia. Four predictive models, incorporated plasma p-tau217, age, and scores of MMSE, STICK, and AVLT, exhibited AUCs ranging from 0.79 to 0.96 in testing datasets. These models maintained robustness across various subgroups and sensitivity analyses. Discussion: Plasma p-tau217 outperforms most traditional variables and may be used to preliminarily screen older individuals at high risk of dementia. Highlights: Plasma p-tau217 showed comparable importance with age and cognitive tests in predicting incident dementia among community older adults. Machine learning models combining plasma p-tau217, age, and cognitive tests exhibited excellent performance in predicting incident dementia. The training models demonstrated robustness in subgroup and sensitivity analysis.