Assessment of nocturnal versus daytime gas exchange in stable COPD. With emphasis on hypoventilation during spontaneous sleep and in sleep influenced by alcohol or zopiclone

Abstract

Background/purpose: Chronic hypercapnic respiratory failure (CHRF) is associated with increased mortality in patients with chronic obstructive pulmonary disease (COPD), and sleep hypoventilation (SH) has been proposed as a possible predictor for CHRF in COPD. SH was previously found in COPD patients with CHRF using long term oxygen therapy (LTOT). However, SH in normocapnic, non-LTOT subjects have not been described. More than half of COPD patients have difficulties in initiating or maintaining sleep or are excessively sleepy at daytime. Hypnotics and alcohol are often used although both are known to depress the respiratory drive to breathe, and little is known regarding SH due to these agents. We have explored the associations between sleep architecture and nocturnal and daytime blood gases in stable COPD, both in spontaneous sleep and during sleep influenced by alcohol or the hypnotic zopiclone. Material/methods: Paper I is an observational sleep study in a pulmonary rehabilitation hospital of 100 (39 male) stable COPD inpatients, mean FEV1 1.1 L (42% of predicted), mean age 64 years, using polysomnography with transcutaneous measurement of carbon dioxide pressure increase (ΔptcCO2). Paper II and paper III presents data from interventional sleep recordings from 26 (9 male) and 31 (10 male) of the same subjects described in paper I, influenced by 0.5 mg ethanol/kg bodyweight or a pill of 5 mg zopiclone, respectively. Results: SH in spontaneous sleep was found in 15%, and although most had CHRF, six subjects were daytime normocapnic. Alcohol induced a mean (95% confidence interval) increase in the ΔptcCO2 during sleep of only 0.1 kPa (0.0-0.2, p=0.047) with no significant increase in the frequency of SH, whereas zopiclone increased the mean (SD) ΔptcCO2 with 0.23 (0.33) kPa, and the frequency of SH from 19% to 42% (p=0.020). Conclusions/consequences: SH is found both in hypercapnic and normocapnic COPD subjects. Whether it is a real predictor of CHRF should be investigated by prospective case-control studies. A moderate dose of alcohol has only minor effects on breathing at sleep whereas zopiclone increases the frequency of SH in COPD.