Claudia Rasmuseen and Dr. James Dee Higley, Psychology Department
Temperament is defined as a variation between individuals’ affective-motivational reactivity and self-regulation (Rothbart & Bates, 1998; Rothbart & Derryberry, 1981; Rettew &McKee, 2005). Temperament can be measured as early as birth and has been found to remain relatively stable across the lifespan (Goldsmith, 1987; Lansade, Bouissou & Erhard, 2008;Chronis-Tuscano, et.al, 2009). Although most temperaments used to be grouped into reactivity and self-regulation, new temperament taxonomies have been found to belong to two higher order dimensions: degree of sensitivity to signals of punishment and a propensity to experience negative emotions (Rettew & McKee, 2005). The variety in taxonomies extends to non-human primates and primatologists are no foreigners to the study of temperament.
The link between temperament and Psychopathology is a well-established finding on both human and non-human primate research. Affective temperaments have been associated with specific clusters of psychopathologies; for instance, predominant affective disorders such as Cyclothymic and irritable are significantly related to depression and paranoia (Iasevoli, et.al.,2013). In addition, Cyclothymic/irritable and hyperthymic temperaments were found to predict both total cases and new cases of bipolar spectrum disorders in a 3-year follow up on college students (DeGeorge, et. al., 2014). Negative affectivity, extraversion, effortful control, and mechanism of association temperaments have been associated to ADHD, major Depression, Anxiety Disorders and Substance abuse (Rettew & McKee, 2005). In addition to mood disorders, there is an established relationship between temperament and personality disorders. Temperaments such as high harm avoidance and novelty seeking are related to borderline disorder (Svrakic, et.al., 2002). There are animal models for certain psychopathologies such as depression and anxiety in rhesus macaques and nervousness temperament is related to anxiety in rhesus macaques (Capitanio, 2012).
The goal for this study was to see if patterns of cortisol could identify temperament extremes. The following discussion will relate back to the BBA database which is John Capitanio’s bio-behavioral assessment on rhesus macaques located in the UC Davis Primate Research Center. This database contains the information of 2,733 rhesus macaques between three and four months of age. Last summer, I participated in the following assessment. For the BBA, infant rhesus macaques are separated from their moms for 2 days, during which 4 blood samples are obtained and later assayed for plasma cortisol levels. Sample 1 reflects an initial response of the infant to separation/relocation. Sample 2 reflects their adaptation to the separation/relocation. Sample 3 reflects the results of an overnight dexamethasone suppression test (Dex). And sample 4 reflects an ACTH stimulation test obtained 30 minutes after Sample 3. On the second day, the infants are returned to their moms, and then reinstated in their colonies. By analyzing the cortisol’s average percent change from sample to sample, I was able to create patterns of stress response. There are 26 possible response patterns that demonstrate the potential plasma cortisol changes the subjects may experience over the four sampling periods. Although during both simple linear regression and analysis of variance, sample 3 seems to be driving most of the relationship between the shapes and other behaviors, when seen individually, sample 3 does not predict temperament such as vigilance. This suggests that the response of the infant before the dexamethasone shot and after the dexamethasone shot is essential to the predictive power of the shapes for behaviors. The creation of these classifications themselves is innovative, for BBA cortisol levels have never been looked at in this holistic form encompassing all individual samples. In addition, these classifications have allowed for the identification of pathology in the HPA axis of these individuals; for instance, shape 7 (Figure 1) denotes individuals whose system did not respond to any of the Dex and ACTH conditions. In addition, to cortisol samples during the BBA, infants are rated on 4 temperaments: vigilance, nervousness, confidence and gentleness.
In addition, to data collection, I learned Mplus a statistical analytical tool that allowed me allow me to use profile analysis. The latent profile analysis basically estimates the probability that any observation will fall into a class and resolves when the classes have the highest probability of membership. I analyzed the 4 cortisol measures and included sex and weight as covariates. The resulting classes were the following 4 illustrated in figure 2. These resulting classes were then analyzed with the patterns of stress response I had created earlier. The results illustrate that the normal patterns of response belong in shape 4; whereas, the most abnormal patterns belong to shape 1.
Once, the stress patterns were validated, I ran an analysis of variance to see if there was a difference on a particular temperament depending on pattern membership. The results indicate that cortisol patterns (p=.02) are significantly correlated to temperament ratings. In addition, the infants who had the most deviant from the norm patterns also displayed extremes in vigilance, nervousness, confidence and gentleness temperaments. The results illustrate that cortisol can be used as a bio-marker for extremes in temperament. Because extremes in temperaments such as nervousness have been used to predict psychopathologies such as anxiety in adulthood, Cortisol can be an additional and more reliable tool to test for psychopathologies.