Machine learning-based LASSO-Cox model for dementia prediction: the role of midlife cardiometabolic, inflammatory, and genetic risk factors in a US cohort

Longjian Liu; Jintong Hou

doi:10.1016/j.annepidem.2026.01.007

Back

Machine learning-based LASSO-Cox model for dementia prediction: the role of midlife cardiometabolic, inflammatory, and genetic risk factors in a US cohort

Journal article

Open access

Peer reviewed

Machine learning-based LASSO-Cox model for dementia prediction: the role of midlife cardiometabolic, inflammatory, and genetic risk factors in a US cohort

Longjian Liu and Jintong Hou

Annals of epidemiology, Forthcoming

23 Jan 2026

DOI: https://doi.org/10.1016/j.annepidem.2026.01.007

PMID: 41581843

Featured in Collection : Research Supported by Drexel Libraries' OA Programs

Files and links (1)

url

https://doi.org/10.1016/j.annepidem.2026.01.007View

Published, Version of Record (VoR)Open Access via Drexel Libraries Read and Publish Program 2026CC BY V4.0, Open

Abstract

Midlife Features and Dementia Risk

LASSO-Cox Modeling

Machine Learning

We aimed to identify key midlife dementia predictors and develop a novel machine learning (ML) -enabled risk prediction model. Using data from 9,266 Atherosclerosis Risk in Communities study participants (aged 45-64 years at baseline, 1987-1989). Incident dementia was ascertained through December 2019. A ML-based LASSO-Cox model was applied to develop the risk prediction model. Over a 25-year mean follow-up, 2,010 participants developed dementia. The LASSO-Cox model identified 12 key predictors and achieved C-indices (95%CI) of 0.77 (0.75-0.79) in the training set (n = 6,182) and 0.78 (0.76-0.81) in the test set (n = 3,084). Predictors included age, Digit Symbol Substitution Test, apolipoprotein E ε4, HbA1c, brachial blood pressure, Factor VIII, Delayed Word Recall Test, hypertension, stroke history, C-reactive protein, white blood cell count, and apolipoprotein B. The resulting nomogram demonstrated strong discrimination (AUC 0.77-0.86) and good calibration. LASSO-Cox risk score quartiles effectively stratified participants into low, moderate, high, and very high dementia risk groups. The findings demonstrate that the newly developed machine learning-based LASSO-Cox model provides a robust method to predict individuals at high risk of dementia.

Metrics

1 Record Views

Details

Title: Machine learning-based LASSO-Cox model for dementia prediction: the role of midlife cardiometabolic, inflammatory, and genetic risk factors in a US cohort
Creators: Longjian Liu (Corresponding Author) - Drexel University
Jintong Hou - Drexel University
Publication Details: Annals of epidemiology, Forthcoming
Publisher: Elsevier
Grants: The predicative values of vascular and metabolic disorders for risk of incident mild cognitive impairment and dementia, R21AG082210, National Institutes of Health (United States, Bethesda) - NIH
Resource Type: Journal article
Language: English
Academic Unit: Epidemiology and Biostatistics
Other Identifier: 991022155550604721

Machine learning-based LASSO-Cox model for dementia prediction: the role of midlife cardiometabolic, inflammatory, and genetic risk factors in a US cohort

Files and links (1)

Abstract

Metrics

Details

Drexel University Social media