Pre–Post Tornado Effects on Aggressive Children’s Psychological and Behavioral Adjustment Through One-Year Postdisaster

Abstract

Using a risk-resilience framework, this study examined how varying levels of exposure to a natural disaster (EF-4 tornado) and children’s characteristics (sex; anxiety) influenced the behavioral and psychological adjustment of children who shared a common risk factor predisaster (elevated aggression) prior to exposure through 1-year postdisaster. Participants included 360 children in Grades 4–6 (65% male; 78% African American) and their parents from predominantly low-income households who were already participating in a longitudinal study of indicated prevention effects for externalizing outcomes when the tornado occurred in 2011. Fourth-grade children who were screened for overt aggressive behavior were recruited in 3 annual cohorts (120 per year, beginning in 2009). Parent-rated aggression and internalizing problems were assessed prior to the tornado (Wave 1), within a half-year after the tornado (Wave 2), and at a 1-year follow-up (Wave 3). Children and parents rated their exposure to aspects of tornado-related traumatic experiences at Wave 3. Children displayed less reduction on aggression and internalizing problems if the children had experienced distress after the tornado or fears for their life, in combination with their pre-tornado level of anxiety. Higher levels of children’s and parents’ exposure to the tornado interacted with children’s lower baseline child anxiety to predict less reduction in aggression and internalizing problems 1 year after the tornado. Higher levels of disaster exposure negatively affected at-risk children’s level of improvement in aggression and internalizing problems, when life threat (parent- and child-reported) and child-reported distress after the tornado were moderated by baseline anxiety.

FUNDING

The research in this article has been supported by grants from the National Institute of Child Health & Human Development (R01 HD079273) and the National Institute of Drug Abuse (R01 DA023156)

Notes

¹ The growth curves were rerun without parental depression in as a covariate. Because caretaker depression is related to child psychopathology, as well as potentially marking biased reporting, the inclusion of the term might artificially deflate the predictive role of tornado exposure. When caretaker depression was removed, the same statistically significant results were found in the growth curve analyses as when caretaker depression was in the model, indicating that it had not confounded the interpretation of the effects of degree of exposure.

² The growth model results for the recomputed Internalizing score (retaining Depression and Somatization, but dropping Anxiety) indicated there was a significant overall decline in the modified Internalizing score, t(59) = −5.22, p < .001; a significant effect of sex on the change rate, t(164) = −2.46, p < .02; a significant effect of child anxiety on the slope, t(164) = −3.38, p < .01; and a significant effect for parents’ TORTE Life Threat score interacting with children’s baseline Anxiety, t(164) = 2.79, p < .01. This pattern of effects precisely parallels the pattern found in for the full Internalizing score, and the effects are in the same directions, indicating that the Internalizing results are not simply due to the effects of the Anxiety score being a baseline predictor and being a part of the Internalizing composite over time.

Additional information

Funding

The research in this article has been supported by grants from the National Institute of Child Health & Human Development (R01 HD079273) and the National Institute of Drug Abuse (R01 DA023156)