Dietary patterns and chronic prostatitis: a symptom severity prediction model based on nutritional clustering and...

Area of Science:

• Urology and Nutritional Science
• Computational Biology and Bioinformatics

Background:

• Chronic prostatitis/chronic pelvic pain syndrome (CP/CPPS) has complex causes, with diet playing a significant role.
• Understanding dietary influences is crucial for managing CP/CPPS symptoms.

Purpose of the Study:

• To develop a predictive machine learning model for CP/CPPS symptom severity.
• To analyze dietary patterns using food frequency questionnaire (FFQ) data.
• To establish a foundation for personalized nutritional interventions in CP/CPPS patients.

Main Methods:

• Principal Component Analysis (PCA) was used to identify dietary patterns from FFQ data.
• LASSO regression selected key dietary predictors of symptom severity.
• Six machine learning models were trained and evaluated, with XGBoost selected as the optimal model (AUC = 0.883).
• SHapley Additive exPlanations (SHAP) were employed for model interpretability.

Main Results:

• Two primary dietary patterns were identified: "Red Meat and Processed Food" and "Dairy-rich".
• The "Red Meat and Processed Food" dietary pattern showed the strongest positive association with CP/CPPS symptom severity.
• The XGBoost model demonstrated superior performance in predicting symptom severity.

Conclusions:

• A robust machine learning model was developed to predict CP/CPPS symptom severity based on dietary patterns.
• The findings highlight the significant link between nutrition and CP/CPPS management.
• This model offers a novel approach for precision nutrition strategies in CP/CPPS care.