by Michael D.Anestis, M.S.
This morning, I came across one of those studies that makes me wonder how some researchers are just ridiculously good at collecting cool datasets. In a study published in Addictive Behaviors by Lela Rankin Williams, Nathan Fox, Carl Lejuez, Elizabeth Reynolds, Heather Henderson, Koraly Perez-Edgar, Laurence Steinberg, and Daniel Pine, the authors looked at the contributions of infant temperament and adolescent risk-taking to substance use in adolescence. To do this, the examined data from 137 adolescents who were recruited into temperament groups at four months of age, assessed at 14 and 24 months, and then assessed again in adolescence (average age 15.3).
In this study, the authors were interested in examining a few different variables. The first was the temperamental trait of behavioral inhibition (BI), which they assessed at infancy. BI is characterized by elevated levels of negative emotions, vigilance, avoidant responses to novel situations (particularly social ones) and heightened autonomic reactivity and stress hormone responses (Calkins, Fox, & Marshall, 1996; Kagan, Reznick, & Snidman, 1988; Kagan & Snidman, 1991). The authors noted that BI has been associated with substance-use problems (e.g., Caspi et al., 1996); however, they also noted that BI is an imperfect predictor of risk and that, as such, research may be better off considering BI in the context of other variables (moderators) rather than considering it alone.
Along these lines, the authors were also interested in examining risk-taking, which they assessed when the participants were adolescents. They wondered whether infant BI would be a greater predictor of adolescent substance use for participants with high levels of risk-taking. Additionally, they wondered whether these effects would be more pronounced for one sex versus the other. The results of the analyses were somewhat mixed, but presented an interesting picture to consider:
First of all, there was a significant two-way interaction of sex and BI in the prediction of adolescent substance use. What that means is that the manner in which and degree to which infant BI was associated with adolescent substance use depended upon whether the participant was a boy or a girl. In this case, boys who exhibited higher levels of BI as infants demonstrated greater levels of adolescent substance use and girls who exhibited lower levels of BI as infants demonstrated greater levels of adolescent substance use.
Second, there was a significant two-way interaction between infant BI and adolescent risk-taking in the prediction of adolescent substance use. In this case, this means that adolescent risk-taking was predictive of adolescent substance use, but only in folks who had exhibited high levels of BI as infants.
So what can we take from findings like this? First of all, the authors did an admirable job of trying to refine an overly simple question: does temperment in infancy predict later behavior? Few of us doubt that stable characteristics from infancy are relevant in terms of later outcomes, but considering only one characteristic ignores the potential impact of others and considering any characteristic without any consideration of environmental influences is also likely to diminish our understanding of any relationship. After all, although they can enhance our predictions, specific temperaments in infancy do not always result in particular outcomes in adolescence. In this case, what the authors demonstrated is that infant BI and adolescent risk-taking are relevant in predicting adolescent substance use, but they are more important in some contexts than in others. This idea is important in all areas of research in clinical psychology. Looking at interactions (moderation effects) allows us to understand things like "are certain people more likely than others to respond to particular treatments for specific diagnoses?" "does childhood abuse impact certain individuals more than others?" and "which individuals who engage in non-suicidal self-injury are most likely to eventually attempt suicide?" In other words, such questions lend a level of specificity to our understanding of things and keep us from making global generations about all people when an effect is more relevant to some than it is to others. Now, keep in mind that interactions should be replicated in multiple diverse samples before we become confident that they represent a general tendency, but studies like this offer us an initial glimpe at a potentially clearer understanding of risk for one particular outcome.
************
If you would like to learn more about this or other topics discussed on PBB, we recommend that you consult our online store for scientifically-based psychological resources.
Mike Anestis is a psychology resident at the University of Mississippi Medical Center and a doctoral candidate in the clinical psychology department at Florida State University.
Subscribe in a reader
