The purpose of this study was to extend the sizeable literature documenting reward system dysfunction in MDD to individuals with rMDD (i.e., a history of MDD but without current MDD). This approach has the potential to inform whether aberrant frontostriatal responses to rewards may represent a trait-like marker of vulnerability to MDD, given that individuals with a history of MDD are at increased risk of developing subsequent episodes of MDD (Hollon, DeRubeis, Shelton, Amsterdam, Salomon et al., 2005
). This approach may also aid in elucidating potential neurobiological mechanisms of MDD while mitigating the possible confounding effects of current mood state, illness severity, nonspecific effects of chronic illness and stress, and of psychotropic medication usage (McCabe et al., 2010
; Kerestes et al., 2011
; Peterson & Weissman, 2011
Based on data from our own research group (Smoski et al., 2009
; Dichter et al., 2009
; Smoski et al., in press
) and others (Forbes et al., 2006b
; Forbes et al., 2008
; Mitterschiffthaler et al., 2003
; Schaefer et al., 2006
; Kumari et al., 2003
; Keedwell et al., 2005a
; McCabe et al., 2009
) demonstrating frontostriatal hypoactivation to rewards in MDD, we hypothesized that the rMDD group would be characterized by frontostriatal hypoactivation during both temporal phases of reward responding. Results from the anticipatory phase of the task were contrary to this prediction: there were no brain regions with significantly decreased activation in the rMDD group, relative to the control group, during reward anticipation. However, there were a number of frontostriatal regions known to be responsive to rewards with relatively greater activation in the rMDD group, including the pregenual aspect of the ACG, the right MFG, and the right cerebellum.
The pregenual anterior cingulate has a central role in processing emotion (Etkin, Egner, & Kalisch, 2011
) and rewards (Liu, Hairston, Schrier, & Fan, 2011
). This region in particular codes for deriving the specific value of an expected reward and for value representations of forthcoming rewards (Wallis & Kennerley, 2010
). The midfrontal gyrus plays a critical role in monitoring incentive-based behavioral responses (Haber & Knutson, 2010
), and activation of this region has been found to be decreased in MDD during reward-based decision making and to predict depression severity in MDD (Smoski et al., 2009
). Finally, although the cerebellum is not typically considered part of the reward network, it has been shown to be involved in aspects of emotion regulation and cognition (Fusar-Poli, Placentino, Carletti, Landi, Allen et al., 2009
) and to be functionally impaired in a range of psychiatric disorders (Baldacara, Borgio, Lacerda, & Jackowski, 2008
). Our finding of increased activation in this region in rMDD requires replication, but may be linked to the extensive projections from this region to aspects of the reward network (Schmahmann, 2010
Although the overall direction of effects during the anticipatory phase of the task (i.e., greater activation in the rMDD group relative to the control group) was not predicted, it should be noted that there is evidence of ACG hyperactivation during reward anticipation in individuals with frank MDD, though in the dorsal rather than pregenual aspect of the ACG (Knutson et al., 2008
), a finding interpreted to reflect possibly increased uncertainty and conflict during anticipation of attainable gains. Given the localization of the present finding to the pregenual ACC, it may be that case that rMDD is characterized by relatively greater neural resources recruited to represent the value of anticipated rewards. Further, given that rewards were uncertain during the anticipation phase of the task, greater responses in this region in the rMDD group may reflect greater on-line monitoring of speeded button responses to obtain the forthcoming reward (Knutson et al., 2008
Analyses of outcome phase responses were consistent with hypotheses of reward network hypoactivation in rMDD and revealed a number of frontostriatal brain regions with relatively decreased activation in the rMDD group, including the OFC, right frontal pole, left insular cortex, and left thalamus. The OFC codes the magnitude and affective value of positive and negative rewards and primary reinforcers (Bechara, Damasio, & Damasio, 2000
), tracks the subjective utility of delayed rewards (Kable & Glimcher, 2007
), and facilitates decision-making based on cost-benefit gradients (de Lafuente & Romo, 2006
), particularly in ambiguous contexts (Hsu, Bhatt, Adolphs, Tranel, & Camerer, 2005
). As such, the OFC codes hedonic value and abstract representations of positive and negative outcomes and responds similarly to obtained rewards and avoided losses (Rolls, 1996
; Kim, Shimojo, & O’Doherty, 2006
). Thus, decreased OFC activation in the rMDD group may reflect diminished tagging of this reward stimulus with affective value. Because a major function of the OFC in incentive contexts is to influence future decision making (Deco & Rolls, 2006
), this has implications for the downstream effects of decreased OFC activation on goal-oriented behaviors. We note that this results requires replication given the possibility of susceptibility artifact above the sinus cavities.
In reward contexts, the frontal pole in believed to code not for incentive motivation or reward-based decision making, but rather for monitoring and evaluating decisions after the presentation of reward or punishment (Tsujimoto, Genovesio, & Wise, 2010
). As such, this region is believed to promote learning associations between behaviors to attain goals as well as costs to attain them (Tsujimoto, Genovesio, & Wise, 2011
The insular cortex mediates coding both the anticipation and experience of negative outcomes (Knutson, Rick, Wimmer, Prelec, & Loewenstein, 2007
; Samanez-Larkin, Hollon, Carstensen, & Knutson, 2008
) and insula activity in reward contexts has been linked to anxiety and avoidance learning (Paulus & Stein, 2006
; Samanez-Larkin et al., 2008
). The thalamus is an integral component of the cortico-basal ganglia system and holds a large glutamatergic projection to the ventral striatum, medial prefrontal cortex, and amygdala (Akert & Hartmann-von Monakow, 1980
) and that mediates motivation and emotional drive, planning and cognition for the development and expression of goal-directed behaviors (Haber & Calzavara, 2009
; Krebs, Boehler, Roberts, Song, & Woldorff, 2011
). Thus, outcome phase data revealed hypoactivation in multiple nodes of the reward network, although we note that outcome phase results were evident only when between-group activations were not masked by grand average task-based activations, and thus we consider these findings exploratory in nature.
Analysis of in-scanner task-related behavior revealed a trend towards relatively slower responses in the rMDD group during only potential reward trials, suggesting a possible behavioral component to altered reward circuitry brain function in rMDD. Exploratory analyses indicated that greater frontal pole activation during monetary anticipation was associated with a greater number of lifetime depressive episodes within the rMDD sample. Given the role of this brain region for monitoring and evaluating decisions after the presentation of rewards (Tsujimoto et al., 2010
), it may be the case that greater decision monitoring predicts greater vulnerability for MDD, perhaps due to linkages with rumination during reward outcomes (Koster, De Lissnyder, Derakshan, & De Raedt, 2011
). It should be noted that there were no significant correlations between brain activation and profiles on the Ruminative Responses Scale (Nolen-Hoeksema, Morrow, & Fredrickson, 1993
) in this rMDD sample, such effects may be evident only during active task conditions, though this interpretation is highly speculative.
Study limitations include a small sample size, the fact that two rMDD participants were receiving psychotherapy at the time of participation, and the fact that five rMDD participants had previously used psychotropic medications. Additionally, we recently reported that reward network dysfunction in MDD may be more pronounced in response to pleasant images rather than monetary rewards, and thus future studies should examine the effects of positive image reward in individuals with rMDD.
In summary, results are suggestive of reward network dysfunction in currently euthymic individuals with a history of MDD. Specifically, results indicate reward network hyperactivation during the anticipation of rewards and reward network hypoactivation during reward network outcomes. These results imply a double dissociation between reward network activity and the temporal phase of the reward response in rMDD, highlighting the importance of considering the chronometry of the reward response when evaluating reward network function related to MDD. More broadly, these findings suggest that aberrant frontostriatal response to rewards may represent a trait endophenotype for MDD, although future studies in other high risk groups, particularly first-degree relatives of MDD patients with no history of MDD, will be critical to establish this potential marker of MDD vulnerability.