We aimed to identify key midlife predictors of incident dementia and develop a robust, machine learning-enabled risk prediction model. We analyzed longitudinal data from 9,266 participants (aged 45–64 years at baseline, 1987–1989) in the Atherosclerosis Risk in Communities (ARIC) study. Incident dementia was ascertained through December 2019. A machine learning-based LASSO-Cox proportional hazards model was applied to develop the multivariable risk prediction model. The results show that in a mean follow-up period of 25 years, 2,010 participants developed dementia. The LASSO-Cox model identified 12 key midlife predictors and achieved strong discrimination with a C-index of 0.77 (95% CI: 0.75-0.79) in the training set (n=6,248) and 0.78 (95% CI: 0.76-0.81) in the test set (n=3,108). These 12 predictors included: age, Digit Symbol Substitution Test (DSST), apolipoprotein E (APOE) ε4 status, HbA1c, brachial blood pressure, Factor VIII, Delayed Word Recall Test (DWRT), hypertension, stroke history, C-reactive protein, white blood cell count, and apolipoprotein B. Built upon these key predictors, the resulting nomogram demonstrated strong discrimination (AUC 0.77-0.86) and good calibration for dementia risk. Quartiles of the LASSO-Cox risk score effectively stratified participants into low, moderate, high, and very high dementia risk groups. In conclusion, the findings clearly demonstrate that midlife cardiometabolic and inflammatory disorders are significant independent predictors of dementia risk in late life. Furthermore, the newly developed machine learning-based LASSO-Cox model offers a robust and highly discriminative method for identifying individuals at high risk of subsequent dementia.

Authors:
Longjian Liu, Drexel University, United States

About the Presenter(s)
Dr. Longjian Liu, MD, PhD, MSc, FAHA is a tenured Full Professor of Epidemiology in at Drexel University, Philadelphia, PA, United States. Professor Liu has extensive research experience in CVD, dementia and healthy aging study.

See this presentation on the full schedule – Thursday Schedule

Bookmark