Effects of diazepam or chronic alcohol treatment on spatial reversal learning in mice.

Mice submitted to chronic alcohol consumption (CAC; 11 months) or to systemic diazepam administration were trained in a spatial reversal learning task. Although CAC-treated mice were able to learn the initial acquisition at normal rates, they were impaired during the first reversal of the discrimination and subsequent reversal sessions. In contrast, diazepam administration induced no deficits for any behavioral measure. In conclusion, CAC, but not diazepam administration, induces an exaggerated sensitivity to proactive interference. The two treatments spared, however, the development of the learning set curve. These results are congruent with clinical data showing that nondeclarative or implicit forms of memory processes are spared in diazepam-treated subjects or in chronic alcoholics.