Elevated triglyceride-glucose index, a surrogate marker of insulin resistance, is associated with a high risk of obstructive sleep apnea

Abstract

This study aimed to investigate the association between representative Insulin resistance (IR) surrogates and obstructive sleep apnea (OSA). The study consisted of 8606 adults from the Korea National Health and Nutrition Examination Survey (KNHANES, 2019-2021), a nationwide population-based database. The STOP-Bang questionnaire was used for assessing the risk of OSA. Fasting insulin concentrations, triglyceride glucose index (TyG index), and homeostasis model assessment for insulin resistance (HOMA-IR) were used as surrogate IR markers. A multivariable logistic regression and a restricted cubic spline (RCS) model were applied to examine the association between IR and OSA risk. Multivariate logistic regression analysis revealed that only TyG index (odds ratio [OR] 2.02; 95% confidence interval [CI] 1.56-2.62) was significantly associated with an increased risk of OSA. Results of the RCS analysis showed a positive nonlinear relationship between HOMA-IR and OSA risk (p for non-linearity < 0.001). Subgroup RCS analyses indicated that the effect of elevated TyG index on increasing risk of OSA was more pronounced in males, those aged over 60 years, and smokers. These findings suggest that the TyG index may be a more promising surrogate marker for identifying individuals at increased risk of OSA than other IR indicators. It may serve as a practical tool for OSA risk assessment, particularly in populations with elevated metabolic burden.

Keywords: Insulin resistance; KNHANES; Obstructive sleep apnea; STOP-Bang; Triglyceride glucose index.

Fig. 1

Flow chart for the inclusion of the study participants. KNHANES Korean National Health and Nutrition Examination Survey, HbA1c Glycated hemoglobin, HDL-C High-density lipoprotein cholesterol. Note: Comorbidity status was determined based on diagnosis history or clinical criteria (e.g., blood pressure, glucose level, cholesterol).

Fig. 2

ROC curve for the STOP-Bang score as a predictor of OSA (AUC = 0.861). ROC receiver operating characteristic, OSA obstructive sleep apnea, AUC area under the curve, Sens sensitivity, Spec specificity, PPV positive predictive values, NPV negative predictive values.

Fig. 3

ROC curve for surrogate markers of insulin resistance to STOP-Bang score. ROC receiver operating characteristic, AUC area under the curve, Sens sensitivity, Spec specificity, PPV positive predictive values, NPV negative predictive values.

Fig. 4

The results of multivariable logistic regression analysis for OSA. OSA obstructive sleep apnea, TyG index triglyceride glucose index. Note: OR = Odds Ratio; CI = Confidence Interval. Asterisks indicate significance levels: *p < 0.05; **p < 0.01; **p < 0.001.

Fig. 5

Association between insulin resistance and OSA based on restricted cubic spline model. Shaded areas are 95% confidence intervals. The model is adjusted for age, sex, body mass index, marital status, educational level, household income level, smoking status, drinking status, aerobic physical activity, anaerobic physical activity, hypertension, diabetes, and dyslipidemia (Model 3) (a) fasting insulin concentrations (b). TyG index (c). HOMA-IR. OSA obstructive sleep apnea, HOMA-IR homeostasis model assessment for insulin resistance TyG, triglyceride glucose index.

Fig. 6

Association between TyG index and OSA based on restricted cubic spline model stratified by sex, age, smoking, and drinking status. Shaded areas are 95% confidence intervals. The model is adjusted for age, sex, body mass index, marital status, educational level, household income level, smoking status, drinking status, aerobic physical activity, anaerobic physical activity, hypertension, diabetes, and dyslipidemia (Model 3). OSA obstructive sleep apnea, TyG index triglyceride glucose index.