Benzodiazepine-induced impairment of matching-to-sample performance in humans

The effects of benzodiazepines on a visual pattern matching-to-sample (MTS) task were examined in nine healthy male volunteers. The MTS task employed randomly generated checkerboard-like stimuli presented on a video display. The sample and two comparison stimuli were simultaneously presented. Nonmatching comparison stimuli were randomly generated to be 3.125, 6.25, 12.5, 25.0, 37.5, or 50.0 percent different from the sample. Subjects responded on left or right button manipulanda to identify the matching comparison stimulus. The nonmatching stimulus condition was maintained constant for a 60-sec component and the percentage difference of the nonmatching stimuli was systematically varied across multiple components. The effects of triazolam (2.25-9.0 micrograms/kg) and lorazepam (7.5-45 micrograms/kg) were examined in a within-subjects, double-blind, placebo-controlled study. Under placebo conditions, response rates and accuracy were a positive function of the nonmatching stimulus discriminability. Triazolam produced dose-related decreases in response rate at nonmatching stimulus conditions greater than or equal to 25%. Only the 9.0 micrograms/kg dose of triazolam decreased accuracy and this occurred across all nonmatching stimulus conditions. Lorazepam effects were qualitatively similar but less robust than those of triazolam.