Sleep laboratory studies of hypnotic drugs: efficacy and withdrawal effects

Flurazepam, temazepam, and triazolam are compared in terms of initial and short term efficacy, effectiveness during intermediate and long term use, withdrawal effects, and general side effects. The usefulness of temazepam is considerably restricted since the drug is slowly absorbed; peak blood concentrations are not reached until 2 to 3 hours after ingestion. Consequently, while the majority of insomniac patients complain primarily of difficulty falling asleep, temazepam is not effective for this sleep complaint. Further, the drug has an intermediate elimination half-life and induces a significant degree of morning sleepiness (hang-over). Rebound insomnia of a moderate degree occurs with some frequency following withdrawal of temazepam. Triazolam is effective initially and with short term use both for inducing and maintaining sleep. However, much of this effectiveness is lost with continued nightly use over an intermediate period (2 weeks). The drug has a rapid elimination rate; during drug administration, sleep may worsen in the final hours of the night (early morning insomnia), and following drug withdrawal, rebound insomnia is frequent, immediate, and severe. Side effects are frequent and include some morning sleepiness (before tolerance develops) and significant memory impairment and even episodes of amnesia. Triazolam may have a narrow margin of safety in that serious behavioral symptoms have been reported even with a 1-mg dose. Flurazepam is effective both for initiating and maintaining sleep with initial and short term drug administration. Further, its efficacy is maintained not only with intermediate term use but with long term drug use (4 weeks). Flurazepam is a long elimination half-life drug, and there is significant daytime sedation during short term use; with continued use this effect diminishes. Rebound insomnia has not been noted following withdrawal of flurazepam; there is a carry-over effectiveness into the first and second nights of withdrawal, and any withdrawal sleep disturbance would be expected to be infrequent, delayed in appearance, and mild in degree.