Pilot controlled double-blind study of the hypnotic effects of zolpidem in patients with chronic 'learned' insomnia: psychometric and polysomnographic evaluation.

In a pilot double-blind trial in 21 patients with learned or idiopathic insomnia (DSM-IIIR), patients received placebo for 1 week (nights 1-7), either active (zolpidem, 10 mg) or placebo treatment for 2 weeks (nights 8-21) and then placebo for a further week (nights 22-28). Variables to measure efficacy, rebound and withdrawal were assessed daily from day 1 to day 28. Polysomnographic recordings together with sleep cycle analysis were performed on nights 7, 21 and 28. Patients treated with 10 mg zolpidem for 2 weeks had significantly improved sleep efficiency at the end of the randomised double-blind phase compared with the placebo group. Fractionated sleep-cycle analysis showed an increase in slow-wave sleep during the first 2-hour cycle after sleep onset. During the withdrawal placebo week, most of the main sleep variables remained relatively stable in the zolpidem group (nights 22-28), and deteriorated further in the placebo group. At the end of the withdrawal phase, there was a statistically significant difference between groups, in favour of the zolpidem treatment, in sleep efficiency, total sleep time, absolute and percentage of time awake, and percentage of REM sleep. REM sleep, which was normal in both groups at baseline, decreased significantly in the placebo group between nights 22 and 28 (during the withdrawal placebo week) compared with the zolpidem treatment group, and the number of periods of time awake increased. Minor subjective complaints were recorded under zolpidem and were comparable with those under placebo. Zolpidem seemed to improve some important sleep variables, when assessed both objectively and subjectively. The sleep cycle analysis suggested a possible shift of slow-wave sleep to an earlier period of the night, with a more physiological sleep structure. There was no evidence for withdrawal or rebound after stopping the 2 weeks of zolpidem treatment, but rather signs that the effect of zolpidem outlasted active treatment. The present pilot study justifies a prospective confirmatory comparison of zolpidem with benzodiazepines in an adequate number of patients and withdrawal after 6-8 weeks of treatment.