Diagnostic Performance of Triglyceride/High-Density Lipoprotein Cholesterol (TG/HDL-C) Ratio in Detecting Insulin Resistance Among Obese Children and Adolescents

Background We evaluated the diagnostic performance of the triglyceride/high-density lipoprotein cholesterol (TG/HDL-C) ratio for detecting insulin resistance (IR) in obese children and adolescents using sex- and puberty-specific homeostasis model assessment for insulin resistance (HOMA-IR) cutoffs. Methods This retrospective cross-sectional study included 207 obese patients (5-18 years old; 127 girls and 80 boys) assessed at a tertiary pediatric endocrinology clinic. IR was defined using sex-specific and pubertal stage-specific HOMA-IR thresholds. Anthropometric data and fasting glucose, insulin, lipid profile, glycated hemoglobin (HbA1c), and thyroid-stimulating hormone (TSH) were recorded. Receiver operating characteristic (ROC) analysis evaluated the TG/HDL-C ratio for IR discrimination. Results Compared with the IR-negative group, IR-positive participants had higher weight, body mass index (BMI) (and standard deviation score {SDS}), TG/HDL-C ratio, fasting insulin, HOMA-IR, HbA1c, fasting glucose, triglycerides, and TSH (all p<0.05). The best threshold for TG/HDL-C to identify IR was around 1.7 (area under the curve {AUC}=0.601; p=0.019), with sensitivity and specificity of 71.2% and 55.1%, respectively. Diagnostic accuracy was consistent across sex and pubertal groups. Conclusions The TG/HDL-C ratio demonstrates moderate diagnostic performance in detecting insulin resistance in obese children and adolescents, and this performance is independent of pubertal status or sex. In clinical settings, such as primary care, where insulin measurement and HOMA-IR calculation are not feasible, the TG/HDL-C ratio can be used as an independent marker. However, when insulin measurements are possible, this ratio should be considered a supportive marker alongside HOMA-IR. Additional investigations should assess the potential of this ratio to enhance diagnostic accuracy in multivariate analysis.