Abstract
BACKGROUND: Obstructive sleep apnea (OSA) is associated with dyslipidemia, neurocognitive impairment, and cardiovascular morbidity. This study investigates the triglyceride-to-high-density lipoprotein cholesterol ratio (TG/HDL-C) as a composite biomarker for OSA prevalence, with implications for mitigating OSA-associated complications through metabolic modulation. METHODS: A nationally representative cohort of 3,270 adults (NHANES 2015-2018) was analyzed using weighted multivariate logistic regression. Restricted cubic splines (RCS) modeled nonlinear associations, while receiver operating characteristic (ROC) curves evaluated predictive performance against conventional lipid indices (TG, HDL-C) and composite lipid indices (TG/HDL-C, NHHR). The DeLong test further compares the predictive efficacy between models. The differences between groups in predicting OSA based on TG/HDL-C levels were further examined via subgroup analysis. RESULTS: Significant differences in lipid profiles were observed between OSA and non-OSA cohorts (p < 0.001). Each unit increase in TG/HDL-C conferred a 17% elevated OSA risk [OR = 1.17; 95% CI: 1.00-1.35; p = 0.045], with a nonlinear dose-response relationship (p for interaction = 0.017). TG/HDL-C demonstrated superior predictive accuracy (AUC = 0.589) versus NHHR (AUC = 0.572). Subgroup analyses indicated that TG/HDL-C as an indicator of OSA varied by the family income to poverty ratio (PIR) group, as well as among congestive heart failure and angina populations (p for interaction < 0.05). CONCLUSION: This study demonstrates that the TG/HDL-C ratio is a superior predictor of OSA risk compared to TG and NHHR. These findings underscore the clinical utility of composite lipid indices TG/HDL-C for early OSA detection and targeted intervention.