Running Head: REWARD-RELATED DECISION MAKING Alterations in Reward-Related Decision Making in Boys with Current and Future Internalizing Disorders

Reward-related decision making was examined within a longitudinal study of 221 11-year-old boys, 50 of whom had current internalizing disorders. Participants completed a behavioral decision-making task involving varying probability and magnitude of obtaining reward. Concurrent internalizing but not externalizing diagnoses were related to reduced frequency of choosing the reward option with a large, likely reward. Furthermore, low frequency of choosing this reward option predicted internalizing diagnoses, self-reported depressive symptoms, and parent-rated internalizing problems one year later. These results are discussed with respect to emotion-based, motivational, and behavioral models of altered reward processing and diminished positive affect in depression. The study represents a step toward understanding some motivational and emotional aspects of early-onset internalizing disorders within a framework of affective neuroscience. Alterations in Reward-Related Decision Making in Boys with Current and Future Internalizing Disorders Approaches to understanding the development of affective disorders have emphasized not only clinical symptoms of increased negative affect, but also symptoms and behavior reflecting diminished positive affect. Emotion-based, motivational, and behavioral models of depression postulate reduced experience of positive affect (Clark & Watson, 1991), reduced activity in the behavioral facilitation system (BFS) (Depue & Iacono, 1989), and reduced frequency of experiencing positive reinforcement (Lewinsohn, Hoberman, Teri, & Hautzinger, 1985), respectively. From an affective neuroscience perspective, positive affect can be parsed into several components related to reward (Rolls, 1999). These include the motivation to obtain rewards, executing reward-seeking behaviors, and the hedonic aspects of experiencing a reward. It is not clear which aspects of positive affect may be altered in internalizing disorders. Depression could involve decreased motivation to obtain reward, reduced frequency of reward-seeking behavior, or diminished experience of rewarding outcomes. Anhedonia, social withdrawal, and reduced activity level, all central symptoms of internalizing disorders, can be understood within this framework. Investigations into specific components of reward processing cannot rely solely on selfreport of mood, but instead must examine reward-related choices and behavior. Yet, there are no such investigations to our knowledge in clinical samples of children, and only limited data from these types of behavioral studies in clinical samples of depressed adults. One study of healthy adults suggests that those with dysphoria do not display the typical enhanced response bias during the reward condition of a signal-detection task (Henriques, Glowacki, & Davidson, 1994). Boys’ Reward-Seeking 4 The authors interpreted this as indicating a conservative behavioral strategy that results from decreased reward saliency. Such a strategy may also lead those with depressive symptoms to be conservative when making decisions with potentially rewarding outcomes, even when the likelihood or amount of reward is high. The reward-contingent decision (RCD) paradigm was developed by Rogers et al. (Rogers et al., 2004; Rogers et al., 2003) to examine specific components of reward-related decisions within an affective neuroscience framework. The RCD is a game of chance that involves choosing one of two options under varying conditions of reward probability and magnitude. The “fixed” option always involves a moderate probability of receiving a low-magnitude reward, while the “risky” option varies in both the probability and magnitude of reward. Healthy adults choose the risky option more often under high-reward and high-probability circumstances than under low-reward or low-probability circumstances (Rogers et al., 2003, 2004). Recently, the RCD was used to test a model of internalizing disorders. Findings indicated that tryptophandepleted adults, who temporarily experienced reduced serotonin activity in the central nervous system, failed to discriminate between high-magnitude and low-magnitude rewards (Rogers et al., 2003). Given that low serotonin function has been implicated in depression and anxiety (van Praag et al., 1987) and is a pharmacologic target of treatment for depression, this outcome is consistent with the hypothesis that internalizing disorders involve altered reward processing. The RCD has yet to be used to examine psychopathology directly, but it has been applied to older children and adolescents (May et al., 2004) and offers the potential to examine how children with problem behavior vary in reward-related decisions. In addition to reflecting current internalizing problems, alterations in reward processing may play a role in the development and persistence of internalizing symptoms. This was Boys’ Reward-Seeking 5 suggested by a study examining the course of adults’ depression in relation to behavioral activation, which is presumed to involve reward motivation. Adults with lower self-reported behavioral activation functioning had a poorer outcome after eight months than did those with higher functioning (Kasch, Rottenberg, Arnow, & Gotlib, 2002). Perhaps reward processing influences the maintenance or recurrence of depressive symptomatology. The current study examined 11-year-old boys’ reward-seeking decisions in relation to concurrent internalizing diagnoses as well as diagnoses and symptomatology one year later. We hypothesized that boys with internalizing disorders would make reward-related choices reflecting diminished reward-seeking, by choosing a risky option less frequently when it involved high reward magnitude. In addition, we expected this style of decision-making to be predictive of higher rates of future depressive symptoms and internalizing disorders.