Some earlier work, not rigorously controlled, suggested that conditional increases in sympathoadrenal stress hormones could occur. The purpose of this experiment was to test this idea further using an appropriately controlled design. To do this, we subjected rats living in a tether-type apparatus to differential fear conditioning. Chronic catheterization allowed us to sample blood before and after conditional stimulus probes without having to touch the rats. No evidence for conditional changes in norepinephrine and corticosterone was found. In contrast, differential conditioning of epinephrine responses was found. The conditional response, however, was not a simple one in that conditional increases in epinephrine following CS+ probes were not always seen. These data support the idea that learned changes in hormonal stress respondents can occur. But they leave open the question of why clear cut conditional changes in these visceral systems are so difficult to obtain.

Although it is known that the number of presentations of a stressor can influence the adrenocortical stress response, relatively little information exists on how stressor intensity affects this process. To evaluate this, we repeatedly presented rats with stressors of 3 different intensities and sampled blood for corticosterone. The first major finding was that the rat's initial adrenocortical responsiveness regardless of the stressor employed was a critical variable. Rats that showed a small corticosterone response showed no evidence of habituation or of differences due to stressor intensity. Rats that showed an initial robust response all showed partial habituation of their corticosterone response over time but the patterns varied with stressor intensity. Handled and prone restrained rats showed the same pattern but rats subjected to the more intense stressor of supine restraint showed delay in habituation and tonically elevated responses. These data indicate that individual differences in reactivity to stressors as well as stressor intensity can influence the pattern of the stress response over the course of repeated administration of the stressor.

On an every other day basis, chronically catheterized male rats were subjected to a 30 sec grid shock in either an ascending (0.0, 0.25, 1.0, 4.0 mA) or descending order. A third group was repeatedly subjected to 1.0 mA shock over the same time frame. In 85% of the shock trials, plasma prolactin increased from baseline levels, thus indicating that prolactin is a relatively reliable index of stress. However prolactin did not change in a step-wise fashion with stressor intensity for a significant number of rats. Data from the group given repeated exposure to the 1 mA stressor showed no evidence of habituation--a process which might have explained the findings. This study indicates that prolactin levels do not sensitively track stressor intensity for individual rats.

When individual rats were exposed to different intensities of a stressor, foot shock, plasma catecholamines were found to be sensitive and reliable indices of the stress. Plasma corticosterone did not perform as well. Similarly, levels of both plasma epinephrine and norepinephrine correlated highly significantly with a behavioral measure of the degree of stress—namely, the amount of movement about the cage seen during the 30 sec shock period. Importantly, this behavioral measure was as sensitive and reliable an index of stress as the catecholamines. However, use of either the catecholamines or this behavioral measure as a clinically useful measure of the level of stress was limited by the fact that their responses to the stressor were extremely short-lived. Nonetheless, because the catecholamines reliably and sensitively track the intensity of a stressor, they appear to be a good visceral measure of stress, perhaps the best currently available. But the behavioral concomitants of stress are quickly and easily quantifiable and present a wide range to study, starting with alerting, through a progression of more aroused motor activity, and ending with fight-flight. Because the behavioral concomitants of stress have not been as intensively studied as the endocrine ones, we believe that future efforts to find a clinically useful index of stress will be rewarded by a refocussing of attention away from the visceral respondent to the overt behavioral one.