Predictability, control, and the pituitary-adrenal response in rats.

Abstract

Five experiments were performed to compare to effects of signaled and unsignaled shock on the pituitary-adrenal response of rats. In Experimental 1, exposure to the two procedures yielded no difference in plasma corticosterone levels. In Experiment 2, the addition of a food-reinforced lever-pressing baseline produced conditioned suppression in the signaled condition but bo group difference in steroid values. In Experiment 3, in order to guard against steroid elevations produced by exposure to shock per se, blood samples were obtained during brief test sessions prior to the occurrence of shock. The procedure resulted in a significant elevation in the steroid levels of the signaled shock group. In Experiment 4, a within-subjects sampling procedure revealed that disparate group steroid values obtained earlier in the session had converged by the end of the test session. The final experiment replicated the original failure to obtain a steroid difference due to predictability in the absence of a behavioral baseline, despite the fact that blood samples were obtained by using the early "probe" sampling procedure. Collectively, these results suggest that (a) the effects of predictability are largely seen in the temporal pattern of steroid elevation and not in their terminal values, (b) the effects of predictability on steroids are modulated by the availability of control, and (c) control is not confined to the stimulus that is being predicted.