The amnestic effect of benzodiazepines, first described in 1965, and the subsequent attempts to identify the precise nature of this effect, are reviewed. The difficulty in deciding to what extent this effect is secondary to the sedative action of these drugs is shown by the lack of agreement between studies. Nevertheless, it is concluded that, given the right experimental design, all benzodiazepines can be shown to cause an anterograde amnesia which is probably primarily a result of reduced attention or rehearsal and secondary to sedation. Its onset, degree and duration are influenced by dose, rate of absorption, route of administration, potency and the receptor occupancy rate of the particular benzodiazepine involved, but plasma elimination t,,, appears to be relatively unimportant. The clinical relevance of this for the long-term use of hypnotics and anxiolytics is not clear. Tolerance appears to be greater than for the anxiolytic but less than the sedative or anticonvulsant effect of benzodiazepines. It seems that transient amnestic effects could occur in chronic users related to post-dose, peak benzodiazepine levels. The great variability in individual response means that transient amnesia is a potential adverse drug reaction in certain individuals taking benzodiazepines.KEY woms-Benzodiazepines, amnesia, sedation, tolerance, receptor occupancy.
HISTORICAL BACKGROUNDThe amnestic effect of benzodiazepines appears first to have been noticed by Brandt and Oakes (1965), who found that half their patients given diazepam (20mg) orally as a premedication prior to general anaesthesia, compared to a quarter of those given pentobarbital (lOOmg), were unable to remember the events prior to induction on the day after their operation. This effect, which was regarded by anaesthetists as a desirable property, was extensively studied and reported in the anaesthetic literature in the 1970s. The psychological methodology was necessarily simple and rather crude, consisting of the recall of events or recognition of picture cards, sometimes tested before the administration of general anaesthesia but more often postoperatively. Nevertheless, several useful reproducible findings emerged.Anterograde amnesia was commonly found after intravenous (Brown and Dundee, 1968;Fox et al., 1968;Pandit et al., 1971;Dundee and George, 1976;George and Dundee, 1977) but not intramuscular (Pandit and Dundee, 1970) or oral (Harry and Richards, 1972;Wilson and Ellis, 1973) diazepam. Lorazepam became preferred for this effect since it occurred reliably after intravenous (Heisterkamp and Cohen, 1975;Conner et al., 1976;Dundee and George, 1976;Pandit et al., 1976), intramuscular (Cormack et al., 1976;Fragen and Caldwell, 1976;Dundee et al., 1977) and oral (Turner, 1973; Magbagbeola, 1974;Dundee et al., 1977;McKay and Dundee, 1980) administration. By showing a series of cards at frequent intervals during premedication George and Dundee (1 977) showed that intravenous diazepam (10, 20mg) and flunitrazepam (1, 2mg) had a peak amnestic effect at 2min and la...