It is very important for knowing the drug effects on sleep to search for not only the self-assessments of sleep on the nights when the drugs were administered but also the electrophysiologically recorded sleep on these nights. Concerned with these studies, many kinds of experimental designs have been used.
The authors undertook the cross over design for investigating the effects of the new triazolobenzodiazepine derivative, estazolam on the all-night sleep pattern. In the present study, problems on choice of the baseline nights and the values of sleep parameters which offered more suitable informations on drug effects, were reported.
Four healthy, male adults, ranging in ages from 29 to 40 years, were subjected to this study. The subjects were administered placebo on the first and the second nights and the active drugs on the third night and placebo on the fourth night, again by double-blind method.
Sleep records were obtained on these four consecutive nights. These trials were repeated four times for each subject.
In the second placebo nights (P₂), total sleep time and stage REM tended to increase and stage 4 tended to decrease, compared with the first placebo nights (P₁). So, it was suggested to be more suitable to choose P₂ as baseline value rather (P₁ + P₂) /2, for P₁ was more biased than P₂.
The two values: (D-P₂) /P₂ (D: drug night) and (D-P₂) were compared to obtain such values that offered more abundant informations on drug efficacies.
There were no great differences between these two values. But, (D-P₂) /P₂were unlikely to be suitably corrected, because for example, as to total sleep time, P₂ was inversely correlated to (D-P₂).
From these results, it is concluded that in the present study it is suitable to use P₂ as the baseline values and to use (D-P₂) for studying the drug efficacies.

抄録全体を表示