Sleep fragmentation during REM and NREM sleep and glucose metabolism in patients with obstructive sleep apnea

Background Although numerous studies have demonstrated the positive relationships between apnea-hypopnea index (AHI), micro-arousal index (MAI) and glucose metabolism, the associations between sleep fragmentation during REM-phase and NREM-phase and abnormal glucose metabolism is not clear yet. Methods Subjects with initially suspected obstructive sleep apnea (OSA) in the Shanghai Sleep Health Study (SSHS) were selected. For each subject, MAI-REM and MAI-NREM data from overnight standard polysomnography; fasting plasma glucose and insulin (FINs) were also collected. The relationships between MAI and glucose metabolism in REM and NREM phases were assessed by logistic regression and multivariable linear regressions. Results A total of 4448 participants (mean age: 41.3 ± 12.4 years and mean BMI: 26.3 ± 3.7kg/ m ² ) were finally included in this study. Of all the subjects, the mean MAI-REM was 25.5 ± 19.3 events/h, and the mean MAI-NREM was 25.5 ± 18.2 events/h. Linear regression analyses revealed that MAI-NREM was independently associated with FINs (β = 0.014, P = 0.011). This association remained significant after sensitivity analyses excluding subjects with MAI-REM higher than the median (β = 0.032, P = 0.026). However, logistic regressions showed that no significant correlations were found between MAI-REM, MAI-NREM and hyperglycemia, hyperinsulinemia or insulin resistance (all P > 0.05), after adjustment for confounders including age, sex, BMI, smoking and alcohol consumption. Conclusion In OSA patients, there might exist a significant positive correlation between MAI-NREM and FINs. Further prospective cohort studies and animal experiments are needed to further determine such relationships.