Obesity‐Induced Breathing Variability During Sleep Is Independent of Apneas and Sleep Fragmentation

Abstract

IntroductionObesity is a major cause of sleep‐disordered breathing (SDB). Conventional metrics of SDB, such as the apnea‐hypopnea index, can be confounded by the effects of obesity on oxygenation and lack of standard definitions. Thus, the impact of SBD on sleep is often difficult to quantify. Sleep fragmentation is frequently observed in obese individuals, but whether it occurs independently of SDB remains unknown. Quantitative analysis of ventilation may delineate the effects of obesity on breathing patterns and sleep fragmentation. We aimed to examine the effects of obesity on respiratory patterns during sleep and the relationship between obesity‐related respiratory variability and sleep fragmentation.MethodsSleep recordings were performed in 15 male lean C57BL/6J mice and 17 male diet‐induced obese (DIO) mice on the same genetic background. We applied Poincaré analysis of minute ventilation (VE) during sleep to estimate the breathing variability. Apneas were scored as ≥90% reductions in airflow for two breath cycles or ≥0.7s regardless the presence of arousals. Arousals were defined as single waking events lasting less than 16s. Arousals were classified as respiratory when associated with ≥30% drops in VEfrom baseline.ResultsBreathing variability was significantly higher in DIO mice during NREM sleep, but not during REM sleep. Obesity was associated with an increased apnea severity and 45% greater sleep fragmentation. However, obesity‐induced breathing variability was not affected by apnea or arousal indexes. Respiratory arousals comprised 15% of the arousals in both strains.ConclusionsObesity increased respiratory variability during NREM sleep independently of apneas and macro‐sleep architecture. Obesity caused sleep fragmentation that was not fully explained by SDB severity. Our quantitative analysis of VE identified differences in breathing variability in obesity that were not captured by traditional SDB metrics, which may be applicable for human SDB.