Early detection and risk stratification of insulin resistance in children and adolescents with excess weight: a sdLDL/oxLDL–Based model

Childhood excess weight is a significant risk factor for insulin resistance (IR), a key precursor to metabolic syndrome and type 2 diabetes. In epidemiologic studies, IR is commonly assessed by the homeostatic model assessment for insulin resistance (HOMA-IR). This study aimed to develop a small dense low-density lipoprotein (sdLDL) and oxidized low-density lipoprotein (oxLDL)-based model for the early detection and risk stratification of HOMA-IR-defined IR in children and adolescents with excess weight. A total of 510 children and adolescents with excess weight (aged 7-17 years) were enrolled and stratified into IR (n = 173) and non-IR (n = 337) groups according to HOMA-IR. Comprehensive anthropometric and biochemical parameters were collected, including comparative analyses of sdLDL, oxLDL, and triglyceride to high-density lipoprotein cholesterol (TG/HDL) ratio. Independent risk factors were identified through univariate and multivariate logistic regression, followed by development of a nomogram prediction model incorporating three lipid parameters and adjusted for age and sex. Model performance was evaluated using receiver operating characteristic curves, calibration analysis with bootstrap internal validation and decision curve analysis (DCA). Univariate analysis identified significant associations between IR and mean arterial pressure, TG/HDL, total cholesterol to HDL (TC/HDL) ratio, LDL to HDL (LDL/HDL) ratio, oxLDL, and sdLDL (all P < 0.05). Multivariate analysis confirmed TG/HDL (adjusted odds ratio [aOR] = 5.61, 95% confidence interval [CI]: 3.060-10.639, P < 0.001), oxLDL (aOR = 1.092, 95% CI: 1.065-1.120, P < 0.001), and sdLDL (aOR = 1.109, 95% CI: 1.076-1.146, P < 0.001) as independent risk factors. The final age- and sex-adjusted nomogram demonstrated strong discrimination (AUC = 0.836), good calibration (bootstrap-corrected mean absolute error = 0.012), and clear net clinical benefit across a wide range of decision thresholds based on DCA. Probability-based stratification enabled both continuous scoring and visual classification of individual IR risk. The developed age- and sex-adjusted nomogram combining TG/HDL ratio, sdLDL, and oxLDL demonstrates superior performance for predicting IR in children and adolescents with excess weight. By integrating continuous risk scoring with clinically actionable risk stratification, this tool supports early identification and tiered management in both screening and clinical decision-making settings.