To systematically evaluate the association between insulin resistance indices and the risk of kidney stones and their recurrence in U.S. non-diabetic individuals, while identifying predictive indicators.

This cross-sectional study analyzed data from the 2007-2018 NHANES. Five IR indices were calculated. Weighted logistic regression, RCS, and mediation analyses were used to assess the independent associations between these indices and the risk of kidney stones and recurrence in non-diabetic individuals.

This study of 9605 non-diabetic participants showed an overall kidney stones incidence of 8.63% and a recurrence rate of 2.70%. Weighted logistic regression and RCS analyses revealed significant positive associations between METs-IR, HOMA-IR, TyG-BMI, and the risk of kidney stones and their recurrence. Every unit increase in METs-IR was linked to a 2% rise in the incidence of kidney stones (95% CI: 1.014-1.027, P<0.001) and a 3.3% rise in recurrence (95% CI: 1.018-1.048, P<0.001); each unit increase in HOMA-IR raised incidence by 5% (95% CI: 1.025-1.078, P<0.001) and recurrence by 7.9% (95% CI: 1.041-1.118, P<0.001). Adjusting for confounders shifted these relationships from nonlinear to linear (P>0.05). METs-IR demonstrated the strongest diagnostic accuracy for predicting recurrence (AUC=0.713), with uric acid and vitamin D mediating associations between IR indices and the risk of kidney stones and their recurrence in non-diabetic individuals.

This study found that elevated IR indices (METs-IR, HOMA-IR, TyG-BMI) significantly increased kidney stone risk in a non-diabetic population. Serum uric acid and vitamin D mediated this association, with METs-IR best predicting kidney stones incidence and recurrence.