Independent Relationships Between Cardinal Features of Obstructive Sleep Apnea and Glycometabolism: a Cross-Sectional Study.
Juanjuan Zou,Yunyan Xia,Huajun Xu,Yiqun Fu,Yingjun Qian,Xinyi Li,Xiaolong Zhao,Jianyin Zou,Lili Meng,Suru Liu,Huaming Zhu,Hongliang Yi,Jian Guan,Bin Chen,Shankai Yin
DOI: https://doi.org/10.1016/j.metabol.2017.11.021
2018-01-01
Abstract:BACKGROUND:Obstructive sleep apnea (OSA) is associated with abnormal glycometabolism; however, the cardinal features of OSA, such as sleep fragmentation (SF) and intermittent hypoxia (IH), have yet to show clear, independent associations with glycometabolism.METHODS:We enrolled 1834 participants with suspected OSA from July 2008 to July 2013 to participate in this study. Polysomnographic variables, biochemical indicators, and physical measurements were collected for each participant. Multiple linear regression analyses were used to evaluate independent associations between cardinal features of OSA and glycometabolism. Logistic regressions were used to determine the odds ratios (ORs) for abnormal glucose metabolism across microarousal index (MAI) and oxygen desaturation index (ODI) quartiles. The effect of the interaction between MAI and ODI on glycometabolism was also evaluated.RESULTS:The MAI was independently associated with fasting insulin levels (β = 0.024, p = 0.001) and the homeostasis model assessment of insulin resistance (HOMA-IR; β = 0.006, p = 0.002) after multiple adjustments of confounding factors. In addition, the ORs for hyperinsulinemia across higher MAI quartiles were 1.081, 1.349, and 1.656, compared with the lowest quartile (p = 0.015 for a linear trend). Similarly, the ODI was independently associated with fasting glucose levels (β = 0.003, p < 0.001), fasting insulin levels (β = 0.037, p < 0.001), and the HOMA-IR (β = 0.010, p < 0.001) after adjusting for multiple factors. The ORs for hyperglycemia across higher ODI quartiles were 1.362, 1.231, and 2.184, compared with the lowest quartile (p < 0.05 for a linear trend). In addition, the ORs for hyperinsulinemia and abnormal HOMA-IR across ODI quartiles had the same trends. There was no interaction between MAI and ODI with respect to glycometabolism.CONCLUSION:SF was independently associated with hyperinsulinemia, and IH was independently associated with hyperglycemia, hyperinsulinemia, and an abnormal HOMA-IR. We found no interaction between SF and IH with respect to OSA-related abnormal glycometabolism.