Dose-response evaluation of the amnestic effects of triazolam and pentobarbital in normal subjects.

The effects of placebo, triazolam (0.25, 0.5, and 0.75 mg), and pentobarbital (100, 200, and 300 mg) were examined in seven normal male volunteers using a double-blind, Latin Square, balanced crossover design. Before and at hourly intervals after oral drug administration at approximately 10 a.m., subjects completed subject ratings of drug effects, psychomotor performance tasks, and two versions of a number recall task in which eight-digit number stimuli were recalled by reproduction on a numeric keypad. In one version, number stimuli were displayed on a video screen for varying lengths of time (3, 6, or 9 seconds) before immediate recall. In another version, subjects initially reproduced a continuously displayed number and then recalled the number following an immediate or a 16-second delay interval. Both triazolam and pentobarbital produced dose-related effects on all measures. Relative potency comparisons showed that triazolam was 270-384 times more potent than pentobarbital on subject ratings and psychomotor measures but was 406-647 times more potent than pentobarbital on measures of recall impairment. Recall performance deficits produced by both triazolam and pentobarbital at short (3-second) stimulus presentation times were attenuated at longer presentation times. With the variable delay task, triazolam but not pentobarbital interacted with the delay interval and produced impairments only at the 16-second delay condition. These data indicate that both triazolam and pentobarbital quantitatively impair acquisitional processes involved in short-term recall performance. Furthermore, triazolam may have a greater potential than pentobarbital to produce memory impairment, as reflected by its greater relative potency on these measures and its tendency to interfere with retention over short delay intervals.