This study provides the first evidence for significant genetic contributions to individual differences in hair cortisol levels under baseline conditions and under conditions of significant environmental challenge. Individual differences in basal HPA activity and reactivity are notoriously difficult to measure, given the sensitivity of the system to time of day, diet, exercise, and sample collection procedures (Kudielka et al., 2009
; Kudielka and Wust, 2010
). The results of this study are consistent with evidence for heritability of the cortisol awakening response (Bartles et al., 2003
; Wust et al., 2000
; Kupper et al., 2005
; Ouellet et al., 2009
; Franz et al., 2010
) and suggest that hair cortisol may provide an effective integrated measure of long-term HPA activity that can be useful in the search for genetic influences on stress response pathways.
The results of this study also replicated the effects of a major environmental stressor on mean cortisol levels assessed in hair, and add to the growing body of evidence that hair cortisol is an effective and simply collected marker for long term activity of the HPA system in response to persistent environmental challenge. A similar increase in hair cortisol levels was found following relocation to a new environment in rhesus monkeys (Davenport et al., 2006
) and the stress of chronic pain, neonatal hospitalization and long term unemployment has been reflected in hair cortisol in humans (VanUum et al., 2008
; Yamada et al., 2007
; Dettenborn et al., 2010
). The present study also confirmed the sensitivity of vervet hair cortisol to increases in circulating cortisol in late pregnancy found in human studies (Kirschbaum et al., 2009
; D'Anna et al., in review).
We do not have a precise estimate of hair growth rates or the time course of cortisol accumulation in vervet hair. In their study of rhesus monkeys, Davenport et al. (2006)
shaved sections of hair 3–4 months prior to new hair collection, thus allowing them to specify the time window for the increase in hair cortisol following relocation to new housing more accurately. They also found no significant difference in cortisol levels between the proximal and distal portions of the hair shaft. The vervet hair samples, reported here, showed significantly higher levels of cortisol in hair sampled 25–29 weeks after the cross country transport, and approximately 4 months after the end of the quarantine period. Our methodology does not allow us to determine whether the higher post-move cortisol levels included these more intense stressors, or whether they simply measured response to the ongoing environmental disturbances following quarantine. The finding of higher levels of cortisol in hair of vervet females sampled in late pregnancy suggests the possible weighting of the prior 1–2 months in hair cortisol accumulation.
The results of the present study indicated a relatively high degree of overlap in the genetic influences on hair cortisol levels in both low and higher stress environments, and did not provide evidence for a major role for gene-environment interactions. There has been considerable interest in gene-environment interactions to explain inconsistency in candidate gene studies of risk for psychopathology (Rutter, 2009
). Most studies of gene-environment interactions have focused on extreme early adversity as a risk factor that interacts with genetic vulnerability to produce depression or other biobehavioral disorders (Caspi and Moffit, 2006
; Suomi, 2007
). In research on cortisol reactivity, studies with human infants and nonhuman primates have provided preliminary evidence for interactions of different genetic polymorphisms with early life stress (Barr et al., 2004
; Chen et al., 2010
; Luijk et al., 2010
), and recent stressful life events were found to interact with the serotonin transporter promoter polymorphism in cortisol response to an experimental psychosocial stressor (Alexander et al., 2009
The early adversity hypothesis could not be tested in the present study as none of the subjects had experienced mother-infant separation or abusive treatment during early development. The environments studied here were within the range of mild to moderate levels of stressor exposure. The monkeys were socially housed in stable matrilineal groups that were maintained after the move to the new environment, and mean levels of cortisol in hair were considerably lower than levels reported for individually housed monkeys in other studies (Davenport et al., 2006
). Under these circumstances, we found evidence for an additive role of genetic influences and current environmental stress on a long term measure of HPA activity. Almost all subjects showed an increase in hair cortisol in the more challenging environment, and subjects who had relatively high cortisol levels in the baseline environment tended to have high levels in the post move environment. Individual differences in cortisol response to the change in environment were influenced by pregnancy status and female age, but we did not find evidence for a significant genetic contribution to the individual pre-post change scores. This result, combined with the relatively high genetic correlation, does not dismiss the possibility of candidate gene-environment interactions in HPA activity, but it does not provide support for a major role for gene-environment interactions under these circumstances.
Longitudinal research on other neurobehavioral phenotypes has demonstrated similar stability of genetic influences over time and across circumstances. In studies of twins, shared genetic influences across developmental stages have been reported for internalizing and externalizing behavior problems (Haberstick et al., 2005
), reading difficulties (Astrom et al., 2007
) and brain morphology (Pfefferbaum et al., 2004
). Longitudinal bivariate analysis of body mass index (BMI) over a 25 year period showed significant increases over time, with a high genetic correlation suggesting that the same genes influenced variation in BMI at both time points (Martin et al., 2010
). These longitudinal studies are consistent with our current findings of relative consistency in the genetic influences on hair cortisol levels before and after a major environmental change.
In summary, the results of this study are important in several respects. They are the first demonstration of significant genetic contributions to individual differences in hair cortisol levels. They support the hypothesis that hair cortisol is a marker for increased HPA activity in response to persistent environmental stressors. They also show substantial overlap in the genes that influence hair cortisol levels in low and higher stress environments, a finding that is consistent with an additive model of genetic and environmental influences on long term HPA activity. These results support the value of hair cortisol as a useful measure for identifying genetic factors that add to chronic environmental stress in creating increased risk for psychiatric, immunological and cardiometabolic disorders.