Supplementary Material for: Triglyceride-glucose index, systolic blood pressure and risk of tubular atrophy/interstitial fibrosis in IgA nephropathy

Introduction: While the triglyceride-glucose (TyG) index, a validated surrogate for insulin resistance, has demonstrated prognostic value in IgA nephropathy (IgAN) progression, its specific relationship with tubular atrophy/interstitial fibrosis (TA/IF) remains undetermined. Given the established association between insulin resistance and hypertension in IgAN, we aimed to investigate the association of the TyG index and systolic blood pressure (SBP) with TA/IF, and to develope a predictive model for early TA/IF detection.
Methods: This cross-sectional study included 691 patients with primary IgAN. Exposures examined included the TyG index and SBP at the time of kidney biopsy, the former being the logarithmized product of fasting triglyceride and glucose concentrations. We tested association between TyG index and TA/IF, defined as Oxford T1-2 scores, using logistic regression models. Mediation analysis was performed to assess the potential mediating role of SBP in this relationship. A novel model was established based on the identified variables to predict risk of TA/IF. The performance of this model was evaluated for discrimination (receiver operating characteristic curves), calibration (calibration curve) and clinical utility (decision curve analysis).
Results: Patients in the highest tertile of TyG index had 3.09-fold higher risk for TA/IF compared with those in the lowest tertile. The TyG index was independently and positively associated with the risk of TA/IF (odds ratio(OR):3.830, 95% confidence interval(CI) 2.578-5.691; P < 0.001). SBP was found to mediate the association between TyG index and TA/IF, with a proportion mediated of 20.7% observed in the highest TyG index tertile (OR [indirect association]: 1.319, 95% CI 1.118-1.558). The developed predictive nomogram model incorporated SBP, estimated glomerular filtration rate, TyG index, high-density lipoprotein cholesterol and proteinuria; it demonstrated good predictive performance with strong discrimination (area under the curve: 0.864; bootstrap-corrected: 0.859) and calibration (calibration curves). Decision curve analysis confirmed the model’s clinical utility showing a positive net benefit over a wide range of threshold probabilities.
Conclusion: In patients with IgAN, the TyG index was independently associated with the risk of TA/IF, and SBP partially mediating this relationship. The developed nomogram, consisting of TyG index, SBP, and other conventional risk factors, provides a practical tool for risk stratification of TA/IF and guidance on IgAN management.