ORIGINAL INVESTIGATION Midazolam does not inhibit association formation, just its storage and strengthening

Rationale Although there have been many studies examin-ing the effects of benzodiazepines on memory performance, their effects on working memory are equivocal and little is known about whether they affect the efficacy of practice of already learned material. Objectives The objectives in two experiments were to examine (a) whether midazolam impairs performance on a working memory task designed to minimize mnemonic strategies such as rehearsal or chunking of information to be recalled and (b) the effect of midazolam on repeated practice of paired associates that were learned before drug administration. Materials and methods Both experiments involved subcu-taneous administration of 0.03 mg/kg of body weight of saline or midazolam in within-subject, placebo-controlled designs, involving 23 subjects in (a) and 31 in (b). Results The drug had no effect on the ability to recall the digits in serial order even though the encoding task prevented the digits from being rehearsed or maintained in an articulatory buffer. Paired associates that were learned before the injection showed a benefit of subsequent practice under saline but not under midazolam. Conclusions The results suggest that (a) midazolam does not affect the formation of new associations in short-term memory (STM) provided that the presentation rate is not too fast to form these associations when sedated, despite the evidence that the drug blocks long-term memory (LTM) retention of associations; and (b) the potential for over-learning with practice of learned associations in LTM is adversely affected by midazolam such that repeated exposures do not strengthen new learning