Free versus forced exposure to an elevated plus-maze: evidence for new behavioral interpretations during test and retest

The rodent elevated plus-maze is based on an approach/avoidance conflict between secure closed arms and aversive open arms that can be measured to assess anxiety. Despite this apparent simplicity, several discrepancies emerge from the interpretation of an animal's behavior in the maze, especially when considering the one-trial tolerance effect. In order to bring new elements of interpretation, we compared the behavior of rats exposed to the standard version of the test (forced exposure) to the behavior of rats that were allowed to freely explore the apparatus. We also compared the effects of testing/retesting and chlordiazepoxide in these two situations. Our results confirm that open-arm avoidance is a natural tendency and therefore that it is not learned during initial exposure to the maze. In addition, comparison of the two situations suggests that some of the open-arm entries during a forced confrontation with the maze are better interpreted as attempts to avoid the whole situation, rather than as indications of a low level of anxiety. Finally, the one-trial tolerance effect was partially reduced in the free-exposure situation. Our results contradict the hypothesis that there is acquisition of a phobic-like response to open arms during trial 1. Rather, they are discussed in line with the hypotheses by Rodgers and Shepherd (Psychopharmacology (Berl) 113:237-242, 1993) and Bertoglio and Carobrez (Behav Brain Res 108:197-203, 2000) concerning the acquisition of spatial information about the whole apparatus, leading on trial 2 to an unbalanced approach/avoidance conflict and to the inefficiency of anxiolytic drugs.