A pharmacodynamic Markov mixed-effects model for the effect of temazepam on sleep.

Background: A hypnogram shows how sleep travels through its various stages in the course of a night. The sleep stage changes can be quantified to study sedative drug effects.

Methods: Hypnograms from 21 patients with primary insomnia were collected during a randomized, placebo-controlled crossover study of 20 mg temazepam. A separate daytime session was performed to determine the pharmacokinetics of 20 mg temazepam and its effect on saccadic eye movement and electroencephalogram. A first-order Markov model was developed to describe the probability of sleep stage changes as a function of time after drug intake and time after last sleep stage change. The influence of temazepam concentration on the probability to change sleep stage was incorporated into the model.

Results: Transitions between sleep stages were profoundly influenced by the time of the night and by the time since the last change of sleep stage. Temazepam reduced the time spent awake. This effect could be attributed to four mechanisms: (1) transition to "deeper" sleep was facilitated, (2) transition to "lighter" sleep was inhibited, (3) regardless of sleep stage, the transition to wake state was inhibited, and (4) return to sleep was facilitated. Daytime sensitivities to temazepam, measured with the surrogate markers saccadic peak velocity and electroencephalogram beta activity, each correlated with one of the transition probabilities influenced by temazepam.

Conclusions: By the development of a Markov model for these non-ordered six categorical data, the effect of temazepam on the sleep-wake status could be interpreted in terms of known mechanisms for sleep generation and benzodiazepine pharmacology.