Relationship between social factors and pituitary-adrenocortical activity in female rhesus monkeys (Macaca mulatta)

Psychosocial stress in humans has been related to the occurrence or progression of certain diseases and a positive social environment has been shown, in some cases, to ameliorate this effect. In many experimental studies changes in serum cortisol levels have been used as an endocrine-defined stress response, however, the variation in this measure under unchanging conditions in socially housed animals is not well documented. The present study was designed to examine the relationship between specific social behaviors and concentrations of serum cortisol in female rhesus monkeys. Subjects were members of one of two social groups. One was an established long-term (8 years) group with members which had from one to many kin, while the second was a group formed 5 months prior to the onset of the present study composed of initially unfamiliar animals. Blood samples were collected weekly and a total of 16 hr of behavioral data was collected on each subject. For purposes of analyses, the year-long study was organized according to calendar quarters (3 months) with cortisol concentrations averaged and behavioral categories cumulated. When regression techniques were applied, it is of note that neither group nor dominance rank was a predictor of cortisol levels. However, when a correlation analysis was applied to each group separately to assess the effect of dominance rank, rank significantly correlated with cortisol levels in the established but not in the recently formed group. In each quarter, analyses revealed behavioral categories which combined to account for a significant proportion of the variance in cortisol levels, but these behaviors were not identical over the four quarters. For three of the four quarters analyzed, and for the year as a whole, an affiliative behavioral category was predictive of cortisol levels in addition to agonistic behavioral categories. These data suggest that cortisol levels are influenced by not only negative interactions, such as receiving bites, but also by positive interactions such as receiving grooms. Results also showed that reconciliatory behaviors occurred significantly more frequently following dyadic agonistic episodes in the recently formed versus the long-term group. The higher rate of reconciliatory behaviors exhibited by the recently formed group may account for the lack of a dominance rank/cortisol relationship.