Australasian Science: Australia's authority on science since 1938

When the World Is Turned Upside-Down

Sad boy

One in ten children will continue to suffer from serious distress after a traumatic experience. Image: iStockphoto

By Eva Alisic

When children are exposed to a traumatic event, most of them experience distress but eventually recover. Which factors predict persistent problems?

Norway is a beautiful country, but also a nation where two horrific attacks recently struck the community. On 22 July, Anders Breivik placed a bomb in the government quarter of Oslo. Eight people were killed and a further 89 were injured. As emergency services scrambled to respond to this devastation, Breivik went to the island of Utøya disguised as a policeman. He gathered the participants of a youth camp around him under the pretence of briefing them on the bombing and starting to shoot, chasing those who tried to flee. The massacre continued for one-and-a-half hours and resulted in 69 deaths. About 60 people were injured.

A traumatic event such as this has an enormous impact, not only on the survivors, witnesses and relatives of the victims but also on their friends, schoolmates, colleagues, neighbours, the wider community and, one could argue, on the identity of a country.

Children and adolescents involved in disasters and other traumatic events are a vulnerable group, and a traumatic event may interfere with their healthy development. We know from a number of studies that most children experience stress symptoms immediately after a traumatic event. These include re-experiencing the event (nightmares and flashbacks), avoidance of situations or people that relate to the event (avoiding places or people who were involved), and a state of hyperarousal (being jumpy and having difficulty concentrating). These reactions are normal after such a frightening experience, and they usually diminish within the first 6 weeks after the event.

However, a small but significant number will continue to experience post-traumatic stress. Although estimations of the proportion of children with long-term stress reactions vary, a number of studies have found this to be between 10% and 20%.

Current research focuses on identifying and understanding children’s trajectories of recovery. For example, Dr Robyne Le Brocque of the University of Queensland and colleagues recently identified three trajectories for children in Queensland who experienced accidental injury. The recovery trajectory (33% of cases) shows high initial stress symptoms that subside quickly. The resilient pathway (57% of cases) starts with somewhat elevated levels of clinical stress that diminish and are below cut-offs that would indicate a clinical level of distress. In the chronic trajectory (10% of cases) the children continuously show high levels of distress for up to 2 years after the event.

Le Brocque believes that the last figure is probably an underestimation and the earlier two overestimations because missing data were biased toward higher levels of symptoms. In addition, this study was carried out in children who had been exposed to accidental, and not intentional injury, which may lead to higher levels of stress.

Knowing that at least one in ten children will continue to suffer from serious distress after trauma beggars the question: how can we predict who these children are? If they can be quickly identified following an event, it will be easier to target psychological interventions to the survivors most in need.

With a team in The Netherlands, we set out to find variables related to long-term stress reactions in children and adolescents up to 18 years of age following any kind of traumatic event. Instead of starting a new study, we decided to statistically combine the findings of international studies that tracked children after trauma. This meta-analysis enabled us to include many more participants and predictor variables than we could incorporate in a single study.

The studies included in the meta-analysis had to meet specific criteria in order to be included. For example, they had to depict a natural process of recovery after trauma – that is, a situation in which some children and families will seek help and others will not, as happens in “normal” circumstances. Studies focusing on the provision of an intervention have not been included for this reason; they could be biased towards children with high symptom levels, and we would expect the interventions to have an effect on the symptom levels of the group. Other inclusion criteria considered the amount of exposure of the participants, the timing of the measurements (at least shortly after the event for the predictors, and at least once after 3 months or more for the measure of stress), and the nature of the post-traumatic stress reactions assessed.

After a systematic search, 40 longitudinal studies (including the one by Le Brocque) incorporating a total of 7039 participants satisfied the filtering conditions. The studies reported on stress after a wide range of traumatic events: accidents, disaster, life-threatening illness, war, terrorism and other violence. However, each study had its own focus and creative ideas, and measured a wide range of variables that may predict later distress, ranging from the number of X-rays taken in hospital to the experience of social support.

In order to come to robust conclusions, we included only those predictors that were measured in at least five separate studies in the meta-analysis. This reduced the number of variables from more than 80 to 12, including demographics, acute stress reactions, the severity of injuries, and parental distress.

So which were the strongest predictors? Would severity of injury play a bigger role than parental distress? And would demographics be of use when predicting persistent problems?

The most powerful factors were early distress symptoms and parental distress. Children and adolescents who had more acute stress (measured within 1 month post-trauma), post-traumatic stress (1–3 months post-trauma), depressive or anxiety symptoms showed higher levels of long-term distress. So, the amount of stress in the short-term often but not always tells us something about well-being later on.

Interestingly, parental distress was also a fairly strong predictor. The role of parents’ own well-being in children’s post-traumatic stress has only just begun to gain attention from researchers, but appears to be an important one. This is not only a finding in our meta-analysis; when we conducted interviews with parents this turned out to be an important theme as well.

We asked parents of children aged 8–12 years who had been exposed to a traumatic event some time ago to look back at their child’s recovery. One parent who had survived a serious car accident with her son said: “In terms of milestones in his recovery… he had quite a relapse at a certain moment. We thought: ‘We’ll have to start therapy’... He really had a tough time then, suffered from some serious blues. I think it is because I collapsed. I had been really strong before, but in that period I just couldn’t anymore. I completely lost touch.”

The meta-analysis also showed that girls have a slightly higher chance of suffering from long-term distress. Some researchers speculate whether girls really have more symptoms or whether they are socially conditioned to report them more than boys, or that internalising problems such as anxiety and depressive symptoms are just more common in girls.

Severity of injuries and the duration of hospitalisation had small effects as well. Elevated heart rate shortly after admission to a hospital may also be a marker of stress hormones that are released during the trauma.

Finally, a number of variables did not contribute to the prediction of the number of stress reactions. These included age, minority status and socioeconomic status. It means that by just looking at a few demographics we will not be able to distinguish those at risk from the children who will recover naturally.

What about constructing an assessment measure based on the nine variables that were significant? Due to the wide variety in topics, instruments and analyses in the studies, we were able to look at the relationship between post-traumatic stress and one variable at a time only, but with the number of studies with similar designs increasing, estimating the combined effect of these variables will be a possibility for the future.

Future research will also require more analysis and insight in the other 70 variables that we excluded from the study as they may be highly relevant. For example, a few studies in the normal population have shown that a history of trauma renders children more vulnerable to post-traumatic stress after a new event. A family history of mental disorders, or having had an earlier episode of anxiety disorder, may have the same effect.

In addition, it is essential to look at potentially helpful factors. Social support is an important protective variable in adults. We could not identify enough studies on the topic to analyse the results for children, but from interviews with children in the highest classes of primary school we think that it is essential for them as well.

In these sessions children talked about their recovery. Many of them spontaneously talked about the support they felt, and some of them recounted exactly how many postcards they had received from friends or brought scrapbooks with drawings by their classmates along to the interview.

Actively seeking support was a theme for a number of children as well. However, that was not their only way to cope with the experience. They also talked about focusing on positive things, avoiding risks, and commemorating the event. Getting a clearer picture of the impact of different styles of coping is an important task for the future.

For now we know that the majority of the children and adolescents involved in the tragic events in Norway will recover in a natural way from their experience – even though it will take time. We will need to closely watch those children who have high levels of symptoms in order to provide them with good services when their trajectory tends to be a chronic one. In addition, we need to take care of not only the children but also their parents and the wider community who have been affected by the terrible choices of one man.

Eva Alisic did her PhD research in The Netherlands and is currently a Larkins Fellow at Monash University.