Importance of serum albumin in machine learning-based prediction of cognitive function in the elderly using a basic...

Area of Science:

• Gerontology
• Neuroscience
• Biochemistry

Background:

• Cognitive decline is a growing concern in aging populations.
• Nutritional status, indicated by serum albumin, may influence cognitive health.
• Predictive models for cognitive function are crucial for early intervention.

Purpose of the Study:

• To investigate the correlation between serum albumin and cognitive function.
• To assess the impact of serum albumin on the accuracy of cognitive function prediction models.
• To explore the utility of machine learning in estimating cognitive function.

Main Methods:

• Analysis of electronic health records from 1,352 patients (2014-2021).
• Cognitive function assessed using the Mini-Mental State Examination (MMSE).
• Machine learning models (DLM, SVM, Random Forest, XGBoost) used to estimate MMSE scores from blood test data, with and without albumin levels.

Main Results:

• A positive correlation was found between serum albumin levels and cognitive function.
• Inclusion of serum albumin improved prediction accuracy (Mean Squared Error) across multiple machine learning models for all age groups.
• Serum albumin was a significant explanatory variable for cognitive function in individuals aged 65 and above.

Conclusions:

• Serum albumin is a valuable biomarker for cognitive function, particularly in the elderly.
• Incorporating serum albumin into predictive models enhances their accuracy.
• Findings suggest a link between nutritional status and cognitive health in older adults.