Sparing of perceptual implicit memory despite suppression of the electrophysiological word repetition effect: a study with general anesthetic isoflurane

Objective: Much effort has been devoted to the search for the neurophysiological correlates of implicit memory. A commonly held view is that the early portion (250–500 ms) of the event-related potential (ERP) word repetition effect reflects processes important for perceptual implicit memory whereas the latter portion reflects processes implicated in explicit memory. It is, however, difficult to disentangle with certainty the relative contributions of each form of memory on ERPs since both forms co-exist in normal subjects. To dissociate ERP effect related to implicit and explicit memory, we used isoflurane sedation in normal subjects to suppress explicit remembering while sparing implicit memory. These ERPs were compared with those of non-medicated control subjects. Methods: Thirteen subjects performed an incidental encoding task for words presented auditorily during the inhalation of a subanesthetic dose of isoflurane. After termination of isoflurane administration, we assessed free recall and recorded ERPs during a syllable completion task (implicit memory) and during a passive listening task (ERP repetition effect). Eleven non-medicated control subjects were tested in a similar manner. Results: The controls showed robust early and late ERP repetition effect. The isoflurane group had implicit memory without free recall and showed no ERP repetition effect. Conclusions: These findings failed to show an association between any part of the repetition effect and perceptual implicit memory. The results are consistent with the view that processes linked to explicit memory contribute to the ERP repetition effect since there was a marked difference in free recall between the control and isoflurane groups. Significance: The present study shows that the reversible alterations of memory by general anesthetics can be used to study the neurophysiological correlates of memory processes