Predictive Models for Knee Pain in Middle-Aged and Elderly Individuals Based on Machine Learning Methods.

Area of Science:

• Computational epidemiology
• Biostatistics
• Geriatric medicine

Background:

• Knee pain is a prevalent issue affecting middle-aged and elderly populations.
• Accurate prediction of knee pain risk is crucial for timely intervention and management.

Purpose of the Study:

• To develop and validate machine learning models for predicting knee pain in adults aged 45 and older.
• To create a user-friendly nomogram for clinical application in assessing knee pain risk.

Main Methods:

• Utilized data from 5386 individuals from the National Health and Nutrition Examination Survey.
• Developed and compared logistic regression, random forest, and Extreme Gradient Boosting models.
• Validated model performance using metrics including Area Under the Curve (AUC), sensitivity, and specificity.

Main Results:

• Logistic regression identified female gender, pain elsewhere, and body mass index as significant risk factors.
• The logistic regression model achieved an AUC of 0.71 in the test set.
• A derived nomogram demonstrated good discrimination with an AUC of 0.75.

Conclusions:

• A logistic regression-based nomogram provides a convenient tool for evaluating knee pain risk in the US middle-aged and elderly population.
• The nomogram utilizes easily obtainable clinical variables, negating the need for additional radiologic assessments.
• This tool can aid primary care physicians in identifying individuals at higher risk for knee pain.