What are the causes of excessive daytime sleepiness in patients with sleep-disordered breathing?

To the Editors:

Recently, the European Respiratory Journal published a study by Mediano et al. 1 on the determinants of excessive daytime sleepiness (EDS) in obstructive sleep apnoea (OSA) patients. The study did not find differences in body mass index (BMI), age or apnoea/hypopnoea index (AHI) between OSA patients with and without EDS, but found significantly more pronounced nocturnal hypoxaemia in patients with EDS, and concluded that nocturnal oxygen desaturation is involved in the pathogenesis of EDS. In his letter commenting on this work, Bahammam 2 pointed out that the possible role of obesity hypoventilation syndrome (OHS), a known cause of EDS, was not ruled out by Mediano et al. 1. OHS patients have the same AHI of pure OSA patients matched by age and BMI, but are characterised by apnoeas of longer duration and more severe nocturnal hypoxaemia 3, like the EDS patients studied by Mediano et al. 1. In other words, Bahammam 2 suggests that the lower nocturnal oxygen saturation and the associated daytime somnolence observed by Mediano et al. 1 in the EDS group might be due to OHS. Thus, Bahammam 2 also concludes that low nocturnal blood oxygen saturation has a fundamental role in EDS, whether or not it is related to OHS.

However, data we have collected from patients with sleep-disordered breathing support the hypothesis that EDS is not always associated with low nocturnal blood oxygen saturation in sleep-related breathing disorders, and that other factors are involved. We have recently demonstrated an association between EDS and impaired autonomic cardiac modulation, suggesting that autonomic arousals may be an additional cause of EDS 4. Autonomic arousals involve brainstem neurons controlling both sleep/vigilance and cardiovascular regulation. They might not produce detectable electroencephalogram (EEG) changes but are responsible for changes in cardiac autonomic regulation, as reflected by baroreflex sensitivity 5 and heart rate variability indices 6. To test our hypothesis, we studied sleep-disordered breathing patients with a wide spectrum of disease severity, from simple snoring to OSA 4. We found that EDS is actually associated with reduced baroreflex sensitivity occurring at night and with an increased ratio of the low- to high-frequency (LF/HF) heart rate power, the latter of which is an indirect index of sympatho/vagal balance 6. Baroreflex sensitivity and the LF/HF power ratio were inversely correlated, and only the LF/HF power ratio was found to be a statistically independent predictor of daytime somnolence at multivariate analysis. Compared to the population studied by Mediano et al. 1, overall, our patients were characterised by lower BMI (29 versus 32 kg·m⁻²) and AHI (35 versus 61 n·h⁻¹). Like Mediano et al. 1, we did not find differences in age, BMI and AHI between patients with and without EDS. Nor, however, did we find the significant differences between non-EDS and EDS patients in apnoea durations and in the minimum value of arterial oxygen saturation at night that Mediano et al. 1 reported in their study.

The discrepancy between our study and that of Mediano et al. 1 could be related to a lower prevalence of OHS in our EDS group. Although we too cannot completely exclude the presence of OHS in our study group, this is unlikely because the mean BMI was <30 kg·m⁻² and because analysis of daytime blood gases did not reveal hypercapnia. To better assess this issue, and to completely rule out the presence of OHS in our EDS patients, we have reanalysed our data after excluding all the obese patients, i.e. all subjects with a BMI ≥30 kg·m⁻², in the EDS group. The mean±sd BMI of this selected subgroup of our EDS patients decreased to 26.6±2.2 kg·m⁻². However, in spite of such careful exclusion of OHS, a significantly higher LF/HF power ratio was still found as compared to that in non-EDS patients, while AHI, apnoea duration and the minimum nocturnal oxygen saturation were practically the same as in patients without EDS (table 1⇓).

In conclusion, the study of Montano et al. 1 and our study 4 together suggest that at least two independent factors are associated with the pathogenesis of excessive daytime sleepiness in sleep-disordered breathing patients. One is a low nocturnal oxygen saturation, probably caused by obesity hypoventilation syndrome, as Bahammam 2 pointed out; another is an enhanced sympathetic cardiac modulation at night, probably caused by repeated nocturnal autonomic arousals, as we have hypothesised 4.

Statement of interest

None declared.