Parsing Childhood Symptoms With More Specificity Can Clarify Differential Outcomes

One of the most frequent reasons parents bring their latency-age children for mental health treatment is the child's oppositional and defiant behavior. These children typically present with a range of difficulties, including moodiness, argumentativeness, difficulty with peers and teachers, and “getting into trouble.” Many of these children grow into teenagers who develop substance abuse, other forms of delinquency, depression, and so on. Two articles in this issue of the Journal take a closer look at this common progression and, with the tools of behavior genetics and brain imaging, give us a more precise understanding of the nuances to this pattern, yielding some interesting questions as to how we may eventually assess and treat such patients with more specificity.

Stringaris and colleagues (1) examine teenagers' self-reports of emotional and behavioral traits at two time points separated on average by about 2 years. Through the use of a genetically informative sample consisting of siblings and twins, the authors were able to examine both the phenotypic association of traits and the underlying genetic association of traits. They focused on the concept of “oppositionality,” which has two primary components: an irritability dimension, which includes such symptoms as a hot temper and sudden changes of mood, and a headstrong/hurtful dimension, which comprises argumentativeness, rule breaking, and spiteful behaviors. In many studies, the irritability symptoms in childhood are associated with depression in adulthood, whereas the headstrong/hurtful symptoms predict later antisocial behaviors. Stringaris et al. replicated this finding over the shorter follow-up period of their sample. Even though irritability and headstrong/hurtful symptoms were moderately correlated at time 1 (r=0.46), irritability symptoms were related more to depression than to delinquency, whereas hurtful/headstrong symptoms were related more to delinquency than to depression. These relationships were true both cross-sectionally and longitudinally. The major new contribution of this study is that it also showed that while irritability and headstrong/hurtful behaviors are themselves only moderately influenced by heritable factors (0.31 and 0.45, respectively), the relationship between irritability at time 1 and depression at time 2 was due primarily to shared genetics (a genetic correlation of 0.2), with a similar finding for headstrong/hurtful behaviors at time 1 and delinquency at time 2 (a genetic correlation of 0.22). The much smaller relationship of irritability at time 1 to delinquency at time 2 showed no genetic correlation; the findings were similar for the relationship of headstrong/hurtful behaviors at time 1 to depression at time 2. Thus, while irritability and headstrong behaviors often show up together in youths, these behaviors each have a different genetic contribution.

The second study, by Schneider et al. (2), looked at two other often correlated traits—risk taking and substance abuse. It is known that addictive behaviors are linked to reduced activation of the ventral striatum, the “reward system” of the brain. However, it is not known whether risk-seeking behaviors, separate from addiction, are also associated with changes in the brain's reward system, which may explain their common co-occurrence. Schneider et al. chose to investigate this question by studying a community sample of young teens, among whom many may have higher levels of risk-seeking behaviors, but not yet addictive behaviors. Their sample consisted of 266 healthy 14-year-olds, with an additional subset of 31 teens with substance use. Risk-taking bias was assessed prior to MRI scanning, by the Cambridge Gamble Task. During functional MRI, reward anticipation was stimulated by a modified version of the monetary incentive delay task. The structural MRI analysis focused on determining the total gray matter volume for each subject, normalizing the scans, and then comparing them voxel by voxel to determine whether risk taking is correlated with structural variability in gray matter. The statistical models controlled for gender (given gender differences in risk taking), pubertal status, handedness, an intelligence estimate, and scanning site.

Schneider et al. showed that in teens with increased risk taking, despite having no substance use problems, activation of the ventral striatum was decreased bilaterally during the reward anticipation task, with a medium effect size (d=0.57 and d=0.52). Furthermore, greater risk taking was also associated with lower gray matter density bilaterally in the same regions, also with a medium effect size (d=0.44 and d=0.57). In the left striatum, but not the right, the association of decreased activation with risk-taking bias was partially mediated by the decrease in gray matter density, but both striatal activation and gray matter density contributed to the association of increased risk taking. In a comparison of the subsample of youths with significant substance use to matched comparison subjects with no substance use, there was significantly less activation in the left ventral striatum in the substance-using group relative to comparison subjects. This study thus extends the research linking behavioral patterns such as risk taking to a vulnerability to developing substance addictions, suggesting that in individuals with high levels of risk seeking, the reward system of the brain does not respond as well to natural rewards, and thus these individuals may be more drawn to seeking out the high rewards that substance use offers. While this study was cross-sectional and thus does not offer any information on whether there is a causal relationship, the research group will continue to follow this sample of teens to see whether these differences in striatal structure as well as activation are indeed risk factors for future addictive illness.

What do we as clinicians make of these reports? Both studies offer evidence that behavioral patterns seen in childhood can indeed be identified as risk factors for future adult psychiatric disorders. The first study suggests that we should consider the two components of oppositional defiant behaviors more specifically within a particular child. If irritability dominates, treatment targeting mood, which may include antidepressants as well as cognitive-behavioral therapy, may be more helpful than behavior modification alone. Conversely, if the defiant behaviors are the primary driver, behavioral interventions with both child and family may be most effective. For youths with a tendency to seek out risky behaviors, parents need to be educated about their children being at higher risk of developing a substance use problem. Engaging these youths in more acceptable risky behaviors, such as high-intensity sports or exotic travel, may give them the stimulation they seek without their feeling a need to resort to harmful psychoactive substances.

These types of reports also give us the data we need, as advocates for our patients, to push for earlier prevention programs for youths with behaviors that we now see are early predictors or risk factors of future illness. Much as cardiology has carved out new indicators for the treatment of high blood pressure, which is a risk factor for ischemic disease, we as psychiatrists can advocate that children with oppositional behaviors beyond the norm warrant earlier intervention.