Serial-probe recognition in rhesus macaques: effects of midazolam.

A serial-probe recognition task was used to assess the effects of midazolam on visual attention and short-term memory in three rhesus monkeys. On each trial, six unique alphanumeric sample stimuli (list items) were presented sequentially followed by a choice period. Choosing the 'probe' stimulus was correct if the probe matched one of the list items; otherwise, choosing the 'default' stimulus (a white square) was correct. Behavior was examined under a range of doses of midazolam (0.065, 0.13, 0.26, and 0.52 mg/kg IM). Midazolam did not significantly reduce choice accuracy or change the shape of the serial position function and did not significantly reduce choice responding. However, choice reaction time was significantly increased by the two highest doses of midazolam. Responding directed at the sample stimuli was reduced at the two highest doses of midazolam. Furthermore, 0.52 mg/kg midazolam significantly increased sample-stimulus reaction time at all six serial positions. Overall, these data suggest that the two highest doses of midazolam tested increase reaction time, but do not directly impair short-term visual recognition memory. This is noteworthy because such doses appear capable of protecting against nerve agent-induced seizures.