PTSD in Head Start Children After Exposure to Hurricane Andrew

Presenter:
Alan M. Delamater, Ph.D.
University of Miami School of Medicine


This study examined post-traumatic stress disorder (PTSD), behavioral adjustment, and
developmental outcomes in 101 preschool-aged children exposed to Hurricane Andrew in 
1992, compared with 74 children who were not significantly exposed. In addition, we
measured mothers' PTSD, depression, and optimism, as predictors of children's responses.      
Children were recruited from Head Start programs in Homestead, FL and in north Dade
County, and were studied 12 and 18 months after the hurricane. The sample consisted
primarily of African-American and Hispanic children from single-parent families (64%).
Structured interviews were conducted with mothers to determine exposure to stress and
children's PTSD symptoms. Behavioral ratings of children were made by mothers and
teachers using the Conners Rating Scales. Developmental ratings were made by mothers.
Mothers completed standardized questionnaires for measurement of depression, PTSD, and
optimism.

Hurricane-exposed (HE) mothers reported more stressful events both during and after the
storm (p's<.001) than mothers living in the north part of the county. At 12 months post-
hurricane, mothers reported more PTSD symptoms in exposed children (p<.0001), with 
16.5% meeting DSM-IV diagnostic criteria for PTSD. At 18 months, HE mothers continued 
to report greater PTSD symptoms in their children (p<.0001), with 11.6% meeting 
diagnostic criteria for PTSD. Results of repeated measures analyses of variance indicat-
ed increased levels of teacher-rated conduct (p<.001), daydreaming (p<.0001), emotional
indulgent (p<.0004), asocial (p<.02), and hyperactivity (p<.0001) behavior problems
among HE children, compared with those in the north of the county. Results from mothers'
behavior ratings indicated increased anxiety (p<.01) and conduct problems (p<.02) at
12 months post hurricane, but not 18 months post.

On self-report measures of PTSD, HE mothers had marginally more symptoms both 12 and
18 months post (p<.075), but at 12 months post, more HE mothers met diagnostic criteria
for PTSD (31% vs 12%, p<.005). Repeated measures analyses showed greater depression 
among HE mothers (p<.02), with significantly more HE mothers meeting criteria for 
moderate or greater depression (27.6% vs 13.7%, p<.03) at 12 months post.

Regression analyses were performed to identify predictors of children's PTSD. The 
analysis at 12 months showed PTSD was predicted by children's history of previous trauma
(p<.05), stress during (p<.01) and after (p<.01) the hurricane, mothers' PTSD (p<.01),
and mothers' ratings of children's anxiety (p<.01). The model accounted for 54% of the
variance in children's PTSD symptoms (p<.0001). The regression for children's PTSD at
18 months post-hurricane revealed that children's PTSD at 12 months post (p<.02), stress
after the hurricane (p<.02), mothers' PTSD at 18 months post (p<.0001), and mothers'
ratings of children's anxiety at 18 months (p<.02) were each independent predictors.
The model accounted for 40% of the variance in children's PTSD scores (p<.0001).

We also examined maternal optimism (as measured by the Life Orientation Test) as a 
correlate of and predictor of depression and PTSD. Maternal optimism at 12 months was
related with children's PTSD symptoms (p<.01), and mothers' own depression (p<.0001)
and PTSD symptoms (p<.0001). Similarly, optimism at 18 months post-hurricane was 
related with children's PTSD symptoms (p<.001), and mothers' depression (p<.0001) and
PTSD (p<.0001). Mothers' optimism ar 12 months also predicted children's PTSD at 18
months (p<.04), and mothers' depression (p<.01) and PTSD symptoms (p<.0001) at 18
months.

Children's developmental outcomes were rated by mothers using the Minnesota Preschool
Development Inventory. At 12 months post-hurricane, 31% of the sample were rated as
being significantly delayed in their overall development, but there was no relation-
ship between delay and hurricane exposure or PTSD. At 18 months, only 15% of children
were considered delayed. However, there was a significant relationship between delay
at 18 months and PTSD status at 12 months: those with PTSD at 12 months were more
likely to be delayed at 18 months (odds ratio=4.5, 95% confidence intervals, 1.7 to
12.3). Similarly, developmental delay at 18 months was associated with more PTSD at
18 months (odds ratio=7.9, 95% confidence intervals, 1.6 to 38.7).

Based on these findings, we conclude that 1) many young children can be expected to
exhibit PTSD symptoms, and other behavioral disruption, for at least 18 months after 
experiencing a natural disaster; 2) these symptoms are related to their previous 
trauma history, the degree of stress experienced during and after the event, and 
their mothers' PTSD; 3) maternal optimism is associated with children's and mothers'
emotional responses to disaster; 4) children with PTSD are at risk for developmental
delay; and 5) interventions should be provided soon after exposure and should target 
both children and their mothers.


Return To Conference Page