The current study tested a set of variables that would account for the relationship between child maltreatment and PTSD symptoms. To do so, a multiple mediator model was tested using three variables, RSA reactivity, cortisol reactivity and experiential avoidance, with prior empirical support to explain the relationship between child maltreatment and PTSD symptoms. The set of variables did indeed mediate the relationship between child maltreatment and PTSD symptoms, suggesting that these variables play a key role in explaining how maltreated children develop PTSD symptoms. However, experiential avoidance was the only variable to contribute significantly to the mediational model, exerting a significantly stronger indirect effect when compared to the indirect effects of RSA and cortisol reactivity. This indicates that the more maltreated participants avoided painful thoughts, emotions, memories, and physiology, the more PTSD symptoms they reported. Moreover, experiential avoidance was positively related to all three PTSD symptom clusters, suggesting that it is associated with a range of PTSD symptoms not one cluster in particular. These results raise the possibility that avoidance plays a role in the development of PTSD symptoms while preventing maltreated children from experiencing and processing the abuse in an effective manner. Alternatively, maltreated children who do not commonly use strategies to avoid painful private events, or who engage in experiencing private events with awareness and acceptance, may be more resilient to the trauma and less likely to develop PTSD symptoms. Thus, focusing on the extent to which adolescents are willing to experience painful private events appears to have considerable utility in research and intervention models.
A particular advantage to studying psychological processes is that they have direct implications for psychological interventions. For instance, secondary prevention programs can incorporate a focus on disrupting experiential avoidance as part of an overall program assisting adolescents recovering from child maltreatment. This method of prevention could potentially disrupt the pathway from maltreatment to PTSD symptoms and thereby prevent or mitigate the development of PTSD symptoms. Intervention models could also be adapted to disrupt experiential avoidance as an alternative means for achieving treatment goals when maltreated adolescents are experiencing PTSD symptoms. For example, by promoting a willingness to experience difficult thoughts, emotions and memories adolescents may be able to counteract patterns of avoidance and promote the experiencing and processing of the abuse to achieve recovery. Moreover, experiential avoidance may have broader implications beyond PTSD symptoms. Experiential avoidance is an approach to all painful private experiences, not just those that are related to traumatic events or PTSD symptoms. Interventions that promote acceptance of private events in conjunction with behavioral change strategies, such as dialectical behavior therapy (Linehan, 1993
) and acceptance and commitment therapy (Hayes, Strosahl, & Wilson, 1999
), have a growing empirical base. Thus, the extension of these treatments to PTSD symptoms may be an avenue to explore in future research.
The lack of findings for cortisol and RSA were surprising. Once group differences on demographic and health-related behaviors were controlled, the maltreated group did display a blunted cortisol reaction to stress. This finding is consistent with previous research showing a relationship between child maltreatment and blunted cortisol reactions in stress paradigms (Hart, et al., 1995
). However, contrary to expectations results of this study failed to find a relationship between cortisol reactivity and PTSD symptoms. While a relationship between cortisol reactivity and PTSD symptoms has been reported in adult samples (Heim, et al., 2000
), this association is not always found with adolescent samples (MacMillan, et al., 2009
). One explanation for this may come from longitudinal research showing that women who were sexually abused have higher resting cortisol concentrations during childhood, comparable estimates to non-sexually abused peers during adolescence, and significantly lower concentrations in young adulthood (Trickett, et al., 2010
). Thus, an adjustment of the HPA axis may be occurring in maltreated samples across development with profiles assessed in younger and older developmental stages more strongly tied to PTSD symptoms. In addition, RSA estimates declined significantly from resting to stress conditions, however, these estimates did not vary significantly by group nor were they systematically related to PTSD symptoms. There may be several explanations for these findings. Like cortisol, there is some variation in studies examining the relationship between RSA and PTSD using adult (Sahar, Shalev, & Porges, 2001
) and child or adolescent samples (Scheeringa, Zeanah, Myers, & Putnam, 2004
). Thus, there may be a similar adjustment in parasympathetic control over stress occurring through development, although future research would need to demonstrate such a change. RSA estimates may also vary by type of trauma although again more research is needed to support such a claim.
Several limitations should be considered when interpreting the results of this study. First, these data are cross-sectional and therefore causal attributions are not appropriate despite the use of mediational modeling. Mediational modeling was used to examine the contributions of several empirically-tested variables in a single model in order to further inform the research literature and theoretical perspectives of maltreatment and PTSD. While this same model may hold in future research, longitudinal models will be needed to promote generalized causal inferences about the effects of RSA, cortisol, and experiential avoidance in mediating the relationship between child maltreatment and PTSD. Moreover, alternative models, such as ones that view cortisol and RSA as moderators of the relationship between child maltreatment and PTSD symptoms may yield additional explanatory value as opposed to viewing these variables as mediators of this relationship. Second, the sample for this study consists entirely of females. While the focus for the study was to examine an adolescent sample at highest risk for PTSD, results cannot necessarily generalize to maltreated males. Third, the effects of maltreatment in this sample are limited to those with a substantiated case of maltreatment, which tend to represent the more severe cases of abuse, in the past 12 months. Thus, results cannot be generalized to chronic cases of abuse per se, abuse that has occurred prior to the past 12 months, or other methods such as self-report. Fourth, while diurnal changes in cortisol stabilize as early as 11 a.m. (Kupper, et al., 2005
; Wessa, Rohleder, Kirschbaum, & Flor, 2006
), sampling cortisol at later times in the day may provide more reliable estimates as the diurnal rhythm continues to stabilize across the day. Finally, timing within the menstrual cycle was not assessed and can influence hormone concentrations, particularly cortisol. Future research should control the effect of menstrual cycling on cortisol concentrations. Despite these limitations, the results of this study suggest that experiential avoidance may be an important variable explaining how maltreated adolescents develop PTSD symptoms. This study provided a strong test of several potential mediators of the relationship between child maltreatment and PTSD symptoms and offers viable ideas for future prevention or intervention research of PTSD symptoms. The results suggest that when maltreated adolescents are willing to experience painful private events, they are more likely to experience fewer PTSD symptoms.