Dietary micronutrients and habitual dietary patterns are associated with infertility risk among individuals with high insulin resistance: A machine learning analysis of NHANES and MyPyramid Equivalents Database

Background Insulin resistance is increasingly recognized as a key metabolic factor in reproductive dysfunction. However, the relationship between dietary factors and infertility risk among individuals with high insulin resistance remains insufficiently understood, especially from a data-driven perspective. Methods Data from the U.S. National Health and Nutrition Examination Survey (2013–2018) were analyzed in conjunction with the MyPyramid Equivalents Database and the What We Eat in America database. Correlation analyses and the Boruta algorithm were used to perform feature selection. Class imbalance was addressed using the Synthetic Minority Over-sampling Technique (SMOTE). Seven machine-learning models were constructed. SHapley Additive exPlanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME) were applied to interpret the associations between dietary factors and infertility risk in high insulin-resistant women. Results The random forest model demonstrated the strongest predictive performance across both databases. In the NHANES-based dietary micronutrient model, SHAP analysis revealed that among individuals with high insulin resistance, vitamin C, beta-carotene, and vitamin K were negatively associated with infertility risk, while moisture, phosphorus, and vitamin B-1 showed positive associations. In the MyPyramid Equivalents Database model, refined grains, other vegetables, tomatoes, and total fruits demonstrated a negative association with infertility, whereas cured meat, milk, and white potatoes exhibited a positive association. Conclusion Dietary micronutrient intake and habitual dietary patterns are closely associated with infertility risk in individuals with elevated insulin resistance. These results underscore the potential benefits of diet-based risk stratification and targeted nutritional interventions for this population.