Understanding of How an Apnea Ends
Our failure to find statistically significant discriminants of EDS in OSAS is also due to lack of understanding of how an apnea ends: The recording of an “alpha EEG arousal” at apnea termination may be an important factor in the report of severe sleepiness and abnormal MSLT scores. Apneas and hypopneas may not systematically lead to an alpha EEG arousal, but may terminate while the subject is still in a light sleep and thus be less detrimental than an alpha EEG arousal. It is also important to note that when we used criteria more sophisticated than those of Orr et al, we were unable to confirm that hypersomnolent OSAS patients had lower nocturnal Sa02 or were more obese. Within the ^5 MSLT score group of patients with severe sleepiness score and subjective complaint of marked somnolence, we identified subjects with a normal BMI of <27 who never desaturated below 87 percent Sa02 but who terminated each apneic event with a clear EEG arousal response. read only
This contrasted with patients presenting with Sa02 drops ^70 percent who continued in light sleep through resumption of respiration. It might have been interesting to evaluate the hypoxic and hypercapnic responses when each subject was studied. Factors such as obesity and sleep fragmentation may blunt the arousal response, leading to the recording of less abnormal MSLT scores. To date, despite more severe respiratory disturbances, data to confirm these hypotheses have never been systematically collected. Our study, on the other hand, has confirmed the findings of Roth et al, obtained on a much smaller group of patients. There is a relationship between the overall nocturnal sleep structure as defined by Rechtschaffen and Kales’ criteria and the MSLT scores for our population. The quality of sleep, defined by variables such as the percentages of NREM sleep stages, the timing of REM sleep, and the number of long awakenings, was critical for the definition of daytime somnolence. Although the group with severe EDS had a consistently longer nocturnal TST, they had the most disturbed sleep stage distribution.
The normal functional organization of the sleeping brain appears absent in severe OSAS. The circadian distribution of REM sleep also seems to be significantly disrupted, as indicated by the percentage of patients presenting one or more early REM sleep periods. There was no statistically significant time-of-day effect for appearance of REM sleep during the daytime naps, and this abnormal REM sleep schedule seems entirely linked to the nocturnal sleep disruption: All of our patients have now been treated and restudied polygraphically on a non-research protocol. We know that treatment of OSAS leads to complete disappearance of the SOREMPs, which differ from those noted with a REM-sleep-related disorder such as narcolepsy.
Our results thus stress the importance of investigating the sleep EEG and its organization systematically when evaluating breathing during sleep or sleep respiratory symptoms such as OSAS, as it will give information on the risks of daytime somnolence.