Background. Both biological and social factors exert an influence on adolescent behavior, including risk-taking behaviors.  However, little research has examined how such biological and social factors interact to influence adolescent behavior, or how common genetic variability may contribute to the observed behavioral variability in risk-taking among adolescents.  The purpose of this study was to assess individual differences in psychobiological and social-cognitive processes that underlie adolescents’ orientations toward pursuing reward versus avoiding risk.

Aims of the pilot study. This report documents our progress on the pilot study over the past year.  The study incorporated three fMRI paradigms (a gambling task to measure the impact of peers on risk-taking, a face perception task to measure affect regulation, and a goal priming task to measure self-regulatory capacity) plus DNA collection to test hypotheses regarding genetic influences on risk-taking.  A behavioral pilot study combining individual difference measures, genotyping, and a signal-detection reward sensitivity task yielded strong evidence of significant interactions between individual differences in self-regulation and COMT genotype on how well participants learned to respond preferentially to high-reward versus low-reward stimuli on the signal detection task. The pilot data were entirely consistent with our initial hypotheses. Thus, the aims of this larger pilot study were: (1) to identify the neural and psychological factors associated with peer influence on risky decision making among adolescents, and (2) to compare the association between those factors and risky decision making outcomes between adolescents and young adults.

Relation to the aims of the Center. The pilot study is relevant to two aims of the larger Center: (1) to formulate theory and hypotheses regarding the multidisciplinary prevention science of regulatory processes, and (2) to conduct basic scientific studies of regulatory processes that can be translated into novel prevention approaches to adolescent substance use. More specifically, by identifying neurobiological and psychological factors that allow adolescents to be differentially vulnerable to the influence of their peers on decision-making and behavior, we hpe to advance the science of prevention and set the stage for subsequent studies in which interventions targeting self-regulatory processes can be tested in schools and other settings where adolescents encounter challenges to avoid substance use specifically and risky behaviors in general.

New R01. As noted previously, the pilot study is now also being supported by an R01 grant to Strauman and Hariri (1 R01 DA031579-01, “Self-Regulation Failure: Identifying and Modifying a Risk Phenotype”), which was awarded under RFA RM-10-002, Science of Behavior Change: Finding Mechanisms of Change in the Laboratory and the Field.  The RFA itself focuses on translational research linking basic mechanisms underlying problematic behavior and novel interventions to alter such behavior.  The RFA is an excellent fit to this study, particularly in terms of our multidisciplinary, theory-based approach to identify the influences on adolescents’ risk behaviors and our ultimate interest in developing and testing new methods for preventing and modifying those behaviors.

Additional sample.  Because of the availability of the grant funds, we were able to add a young adult/college student sample drawn from the Duke Neurogenetics Study (DNS; Hariri, PI) and so we will have comparable data on adolescents and young adults.  Likewise, because additional funds are available through the R01, we increased our enrollment goals to 75 adolescents and 75 young adults, which will provide sufficient statistical power to test our main hypotheses regarding individual differences in risk mechanisms among adolescents as well as between-group comparisons of the two age groups.

Data collection. The proposed enrollment for the pilot study, expanded once the R01 became available, was 75 college students and 75 adolescents. As of July 1, we have completed at least partial data collection on 60 college students and 54 adolescents. Given the improved rate of data collection over the summer (which has been considerably greater for the adolescents compared to during the school year), we anticipate that by the end of September we will have completed data collection on all 75 college students and all 75 adolescents.

Findings to date. Because the DNA is being processed in batches, we do not yet have genotyping data available on all of the participants run as of 7/1.  However, we have several preliminary findings to report based on quality control analyses of fMRI data obtained to date.  Those findings include:

1. For both college students and adolescents, promotion goal failure feedback is associated with decreased left frontal activation as predicted, compared with activation at the same site following promotion goal success feedback. (See Figure 1.)  In addition, the difference in left frontal activation between promotion success and promotion failure feedback is correlated significantly (r = .48) with self-reported overall promotion goal pursuit success.  That is, individuals reporting greater success in attaining their promotion goals have greater left frontal activation following promotion success priming compared to promotion failure priming. This finding, if validated in the entire combined sample, is consistent with our hypotheses regarding the psychological and neural components of the hypothesized risk phenotype.

1. Both promotion and prevention goal priming (across the success and failure conditions) are associated with differences in posterior activation (including precuneus) when comparing the two samples.  (See Figure 2.) The college students show significantly greater precuneus activation for both types of goal priming compared to the adolescents. This finding, if validated after data collection is complete, indicates potentially important age differences in brain regions recruited by exposure to personal goals, consistent with prior reports in the literature and our own neurodevelopmental hypotheses.

The definitive tests of our hypotheses concerning neural and psychological mechanisms underlying risky choices in adolescents (compared with young adults) will be conducted in early fall, once data collection and genotyping are complete.  Nonetheless, in combination with our original pilot data, the data in hand are consistent with our model for how an a priori defined subset of individuals could become vulnerable to self-regulation failure and the behavioral and emotional consequences that result.

Data entry and preparation (e.g., preprocessing of fMRI data as well as genotyping) are being conducted as data are obtained, allowing us to keep the dataset current.  We anticipate conducting initial analyses to test hypotheses within the adolescent sample (and parallel analyses within the young adult sample) over this fall. We then anticipate final data analyses and manuscript writing to take place in spring 2012.  The dataset will be available to other Center investigators as well as to colleagues conducting similar research at other institutions, under the guidance of our local Institutional Review Board.

Figure 1. Activation following promotion goal success feedback (combined across the college student and adolescent samples).