A machine learning framework for predicting cognitive impairment in aging populations using urinary metal and...

Area of Science:

• Environmental Health
• Neuroscience
• Computational Biology

Background:

• Cognitive impairment in older adults is a growing public health issue.
• Environmental metal exposure is a significant risk factor, but combined effects and demographic influences are not well understood.

Purpose of the Study:

• To investigate the association between exposure to multiple environmental metals and cognitive impairment in older adults.
• To identify demographic and metal-related predictors of cognitive decline using machine learning.
• To develop a predictive model and accessible tool for risk assessment.

Main Methods:

• Analysis of data from 1,230 older adults (≥ 60 years) across four NHANES cycles (1999-2014).
• Quantification of urinary metals and creatinine, with cognitive status assessed using standardized tests.
• Application and evaluation of six machine learning algorithms, including eXtreme Gradient Boosting (XGBoost).

Main Results:

• The XGBoost model achieved high accuracy (ACC=0.81) and AUC (0.90) in predicting cognitive impairment.
• Key predictors identified include educational level, age, race/ethnicity, creatinine, and urinary levels of thallium, molybdenum, and barium.
• A publicly accessible webserver was developed for the predictive model.

Conclusions:

• Machine learning effectively identifies environmental metal exposure and demographic factors associated with cognitive impairment.
• The developed webserver facilitates accessible risk screening for precision prevention strategies in aging populations.