Aim: To explore the effectiveness of the cardiometabolic index (CMI), lipid accumulation product (LAP), and waist-to-thigh ratio (WTR) in identifying insulin resistance (IR) in females who are clinically not having diabetes mellitus (DM). Methods: This cross-sectional study was conducted in Sri Lanka between 2020 and 2023. A total of 282 female participants aged 18 to 35 were recruited. Their anthropometric measurements, fasting blood sugar (FBS) levels, IR by HOMA-IR, lipid profile, and blood pressure were assessed. All females with FBS levels >100 mg/dl were excluded from the study. Results: The study participants were divided into two groups, normal-weight controls (N = 142, BMI:18.0-22.9 kg/m2) and overweight/obese cases (N = 140, BMI:23.0-24.9 kg/m2, Obesity: >25 kg/m2), based on their BMI values. The mean values of CMI, LAP, and WTR in cases were 3.56 +/- 0.63, 49.97 +/- 17.75, and 1.88 +/- 0.17, respectively. Significant associations were identified between all adiposity indexes, IR and lipid profile parameters (p < 0.05). In this female cohort, despite being clinically non DM, the prevalence of IR was significantly higher (52.83 %). ROC analysis results showed greater sensitivities for detecting IR with higher AUC as follows: CMI = 0.854(0.811-0.899), LAP = 0.869(0.829-0.910), and WTR = 0.828(0.779-0.877). The normal-weight controls were assessed to check the prevalence of metabolically-obese normal-weights (MONW,65.5 %) and metabolically-healthy normal-weights (MHNW,35.5 %) among the population, given the higher prevalence of IR. The detection accuracy of MONW females in the ROC assay was; CMI > LAP > WTR > VAI. Conclusions: CMI had the highest accuracy for detecting IR, with the accuracy of all four indexes ranked as follows: CMI > LAP > WTR > VAI.
