Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Estimation of machine learning-based models to predict dementia risk in patients with atherosclerotic cardiovascular diseases: a UK Biobank study
ABSTRACT
Background:
The Atherosclerotic Cardiovascular Disease (ASCVD) is associated with dementia. However, the risk factors of dementia in ASCVD patients remain unclear, necessitating the development of accurate prediction models. Our aim was to develop a machine learning model for use in patients with ASCVD to predict dementia risk using available clinical and sociodemographic data.
Objective:
To develop a model for use in patients with ASCVD to predict dementia risk using available clinical and sociodemographic data.
Methods:
This prognostic study included patients with ASCVD between 2006 and 2010, with registration of follow-up data ending on April 2023 based on the UK Biobank. We implemented a data-driven strategy, identifying predictors from 316 variables and developing a machine learning model to predict risk of incident dementia, Alzheimer's Disease (AD), and Vascular Dementia (VD) within five, ten, and longer-term follow-up in ASCVD patients.
Results:
A total of 29 630 patients with ASCVD were included and 1336 (4.5%) developed dementia during a mean follow-up of 10.3 years. The best prediction model was light gradient boosting machine (LightGBM), achieving the following performance metrics for all incident dementia: AUC, 0.86; accuracy, 0.86; sensitivity, 0.67; specificity, 0.88; precision, 0.39; and FI-score, 0.50. The top 10 important features were examined, among which age, time to complete pairs matching tasks, neuroticism scores, glucose levels and spirits consumption had positive associations with incident dementia risk while Forced vital capacity (FVC), time engaging in activities, vitamin D levels, C-reactive protein levels and red wine consumption had negative associations.
Conclusions:
The findings of this study suggest that predictive modeling could inform patients and clinicians about ASCVD at risk for dementia.
Citation
