Background Obstructive sleep apnea (OSA) has a high prevalence, and it has been shown to be an independent risk factor for various diseases. Therefore, it is important to strengthen screening for population at highrisk of OSA. OSA patients are prone to combine with lipid metabolism disorders, but it remains unclear whether the atherogenic index of plasma (AIP), visceral adiposity index (VAI), lipid accumulation product (LAP), cardiometabolic index (CMI), and Chinese visceral adiposity index (CVAI), which are used asmetabolic indexes, can be used to predict OSA.
Methods A total of 2 968 inpatients with suspected OSA and aged ≥18 years who completed polysomnography (PSG) in the First Affiliated Hospital of Xinjiang Medical University from March 2017 to June 2022 were selected, with 2 850 patients finally included based on the inclusion and exclusion criteria and divided into the OSA group 〔apnea-hypopnea index (AHI) ≥5 times/h, n=2 193〕 and non-OSA group (AHI<5 times/h, n=657) according to the AHI. The clinical data and laboratory test results of these patients were collected through the electronic medical record system. Univariate and multivariate Logistic regression analyses were used to investigate the correlation of AIP, VAI, LAP, CMI, and CVAI with OSA. The receiver operating characteristic (ROC) curve was plotted to analyze the efficacy of metabolic indexes in predicting OSA. A gender-stratified analysis was performed to explore the relationship between metabolic indexes and OSA in different populations.
Results Age, gender (male proportion), neck circumference, height, total cholesterol, triacylglycerol, AHI, AIP, VAI, LAP, CMI, and CVAI were significantly higher in the OSA group than the non-OSA group, high-density lipoprotein cholesterol (HDL-C), mean oxygen saturation and minimum oxygen saturation were significantly lower than the non-OSA group (P<0.05). After dividing the five metabolic indexes into quartiles (Q1 to Q4), them ultivariate Logistic regression analysis showed that AIP〔OR=2.241, 95%CI (1.689, 2.972), P<0.001〕, VAI〔OR=2.517, 95%CI (1.919, 3.301), P<0.001〕, LAP〔OR=2.313, 95%CI (1.761, 3.038), P<0.001〕, CMI〔OR=2.732, 95%CI (2.054, 3.633), P<0.001〕, and CVAI〔OR=6.060, 95%CI (4.411, 8.324), P<0.001〕 were associated with the risk of OSA (P<0.05). Further analysis stratified by gender showed that in female patients, AIP, VAI, LAP, CMI, and CVAI were associated with the risk of OSA (P<0.05) ; in male patients, CMI, LAP, and VAI were not associated with OSA (P>0.05), but AIP and CVAI were associated with OSA (P<0.05). The areas under the ROC curves (AUCs) of AIP, VAI, LAP, CMI, and CVAI for predicting OSA were〔0.593, 95%CI (0.568, 0.618) 〕〔0.607, 95%CI (0.583, 0.632) 〕〔0.594, 95%CI (0.569, 0.619) 〕〔0.616, 95%CI (0.591, 0.640) 〕, and〔0.728, 95%CI (0.706, 0.751) 〕, respectively.Further analysis stratified by gender for the clarification of the predictive efficacy of five metabolic indexes for OSA showed that the AUCs of the five metabolic indices for predicting OSA were higher in the female population than the total population, and the AUCs of the five metabolic indexes were lower in the male population than the total population. The AUC of CVAI was higher than other indexes in the total population, male and female populations (AUC=0.728 for the overall population, AUC=0.764 for the female population, AUC=0.681 for the male population) .
Conclusion As the quartiles of AIP, VAI, LAP, CMI, and CVAI increase, the risk of OSA rises. CVAI has a better predictive efficacy for OSA than other indexes, therefore, CVAI may be used as a predictor for screening of population at high risk of OSA.