Exploring the potential relationship between kidney disease index and cognitive dysfunction: a machine learning...

Area of Science:

• Gerontology
• Nephrology
• Neurology

Background:

• Cognitive impairment is a growing concern in aging populations.
• Kidney function is increasingly recognized as a factor influencing brain health.
• Novel biomarkers are needed to predict cognitive decline effectively.

Purpose of the Study:

• To investigate the association between the Kidney Disease Index (KDI) and cognitive function in older adults.
• To evaluate KDI's potential as a predictive marker for cognitive impairment.
• To compare KDI's predictive performance against established kidney function markers (eGFR, UACR).

Main Methods:

• Analysis of data from the National Health and Nutrition Examination Survey (NHANES) for adults aged 60+.
• Utilized multivariable regression, restricted cubic splines, and machine learning (LASSO, XGBoost, random forests) for predictive modeling.
• Evaluated model performance using ROC curves, decision curve analysis, calibration curves, and SHAP values.

Main Results:

• Higher KDI levels were significantly associated with lower cognitive scores and increased risk of cognitive impairment.
• Each 0.1 unit increase in KDI correlated with a 0.17-point decrease in cognitive scores and a 15% increase in cognitive impairment probability.
• Machine learning models demonstrated that KDI (AUC > 0.7) outperformed eGFR and UACR in predicting cognitive impairment.

Conclusions:

• KDI shows promise as a reliable indicator for predicting cognitive function and impairment risk in the elderly.
• Elevated KDI levels may represent a risk factor for cognitive decline, highlighting the importance of KDI management for cognitive health.
• Further longitudinal research and clinical interventions are recommended to confirm KDI's utility and develop targeted treatment strategies.