Do parent posttraumatic stress symptoms (PTSS) predict later child PTSS?

Background: Although research suggests that parent and child PTSS are associated, the magnitude of this association at different time points and in the context of various covariates has been difficult to quantify in a robust manner due to methodological limitations of past studies (e.g., small sample sizes, reliance on meta-analyses of aggregate data). A more complete understanding of the relationship between parent and child PTSS is necessary to elucidate the etiology of child PTSD and inform screening and early intervention efforts. Aims: The present study aimed to clarify whether and how strongly parent PTSS predicted later child PTSS using pooled cross-study individual participant data. The study also sought to confirm the cross-sectional association between parent and child PTSS during the acute and post-trauma periods. It was hypothesized that parent and child PTSS would be cross-sectionally associated, parent acute PTSS, and acute DSM-IV avoidance symptoms specifically, would each predict later child PTSS, and parent acute PTSS and acute DSM-IV avoidance symptoms would each increase the likelihood of later child subthreshold PTSD. Additionally, it was hypothesized that the parent and child PTSS relationship would be moderated by gender and trauma type, whereby female (vs. male) caregivers would have a stronger influence on child PTSS, and that parent PTSS would more strongly relate to child PTSS after a potentially traumatic event (PTE) which involved interpersonal violence (vs. other types of PTEs). Method: Data was obtained from an international data archive of child traumatic stress. Item-level parent and child PTSS data was harmonized by dichotomizing each item for the presence/absence of DSM-IV PTSS. PTSS severity was determined by counting the number of present DSM-IV PTSS. Spearman correlations were calculated to determine the relationship between parent and child total PTSS and between acute parent DSM-IV avoidance symptoms and later child PTSS. Negative binomial regression analyses were conducted to examine the associations between parent and child PTSS when controlling for confounding factors, such as child ethnic minority status and (when predicting later PTSS) child acute PTSS. Binomial logistic regression analyses examined if parent acute PTSS and acute DSM-IV avoidance symptoms each predicted subthreshold child PTSD. Regression analyses were repeated with the interactions of parent gender and parent PTSS, and PTE type and parent PTSS. Results: The final sample included data from 1,817 cases from 16 studies. Parent and child PTSS had small to moderate cross-sectional ([rho]s = .23 - .26) and longitudinal associations ([rho] = .31). These associations all maintained significance when including covariates in the model (odds ratios = 1.06 - 1.07). Parent acute DSM-IV avoidance symptoms were moderately associated with later child PTSS ([rho] = .31), but this relationship lost significance when including demographic covariates. Parent PTSS significantly predicted later child subthreshold PTSD. Parent acute DSM-IV avoidance symptoms only predicted subthreshold PTSD excluding demographic covariates from the model. Regression analyses indicated that parent gender and trauma type did not moderate the parent and child PTSS relationship. Discussion: Findings suggest that parent and child PTSS are cross-sectionally and longitudinally associated, and that parent PTSS is one of multiple risk factors for child PTSS, including child ethnic minority status and child acute PTSS. There was mixed support for parent DSM-IV avoidance symptoms predicting later child PTSS, and no support for moderation effects of trauma type and parent gender. Early intervention addressing this multitude of factors, such as systemic challenges faced by ethnic minorities, parent acute PTSS, and child acute PTSS, may optimally reduce risk for later child PTSS. Additional research is needed to understand how parent PTSS relates to later child PTSS using more diverse samples and varying study designs.