Effect of benzodiazepines on structural and conceptual/lexical priming

Impaired perceptual priming, as assessed by naming reaction times and accuracy, for briefly presented contour-deleted pictures under lorazepam has been documented in several studies but whether the nature of this impairment is visual versus conceptual/lexical is not clear. We used a previously developed paradigm to examine whether lorazepam affects visual processes involved in the coding of contour information or conceptual/lexical representations.

Three groups were tested (lorazepam, diazepam and placebo). In the study phase, participants named line drawings, presented for 500 ms, in which 50% of the contour was deleted by removing every other line and vertex from each part. In the test phase, participants saw the identical picture, its complement (the other 50% of the contour), or a samename/different-shape picture, each presented for 200 ms.

For all three groups, the magnitude of priming, as assessed by naming RTs and error rates, was equivalent in the identical and in the complementary conditions and the amount of facilitation was reduced in the same-name condition. Perceptual priming occurred both in the lorazepam and in the diazepam groups, though reduced, compared to placebo, for RTs. No conceptual priming (i.e. facilitated performance in the same-name condition) was observed in the benzodiazepine groups.

Equivalent priming for identical and complementary pictures in the three groups suggests that benzodiazepines do not affect the activation of the geon structural description. The lack of priming for same-name pictures suggests that benzodiazepines affect access, retrieval or selection of conceptual/lexical information.