The difference may not necessarily reflect a true difference in PaC02 but may be due to a combination of the more interrupted sleep on placebo and the capnometers inability to follow rapid variations in PaC02, which causes an underestimation of the peak level.
The sleep-induced decrease in Sa02 in our patients was affected very little by nitrazepam or flunitraze-pam. This, somewhat unexpected, result of the study may, in part, be explained by the fact that all our patients took oral theophylline (which was also the case in most of the patients of Block et al). We found it, however, unwise to withdraw theophylline during the study since we wanted to avoid a deterioration in lung function. As most COPD patients use theophylline regularly, our study design also is more pertinent to the everyday clinical situation.

Another explanation for the small effects of benzodiazepines may be that the patients were clinically stable, not severely hypoxemic during sleep (Sa02 >70 percent), not hypercapnic, not obese and not allowed to consume alcohol prior to the study. These factors might otherwise have enhanced the respiratory depressant effect of the benzodiazepines. Most reports on serious nocturnal hypoventilation after benzodiazepine sedation concern acutely ill, severely hypoxemic and hypercapnic patients who may be abnormally sensitive to the respiratory depressant effects.
We finally wish to emphasize that we studied the effects of single doses of benzodiazepines. A harmful effect on oxygenation during sleep from long-term treatment with benzodiazepines in COPD patients cannot be excluded from our study. This is especially true for slowly eliminated benzodiazepines, such as nitrazepam and flunitrazepam.
With these reservations in mind, we conclude that oxygenation during sleep in nonoverweight patients with severe but stable hypoxemic nonhypercapnic COPD treated with theophylline seems to be affected very little by single therapeutic doses of nitrazepam or flunitrazepam.