Social anhedonia and medial prefrontal response to mutual liking in late adolescents

Abstract

Anhedonia, a cardinal symptom of depression defined as difficulty experiencing pleasure, is also a possible endophenotype and prognostic factor for the development of depression. The onset of depression typically occurs during adolescence, a period in which social status and affiliation are especially salient. The medial prefrontal cortex (mPFC), a region implicated in reward, self-relevant processing, and social cognition, exhibits altered function in adults with anhedonia, but its association with adolescent anhedonia has yet to be investigated. We examined neural response to social reward in 27 late adolescents, 18–21 years old, who varied in social anhedonia. Participants reported their social anhedonia, completed ratings of photos of unfamiliar peers, and underwent a functional magnetic resonance imaging task involving feedback about being liked. Adolescents with higher social anhedonia exhibited greater mPFC activation in response to mutual liking (i.e., being liked by someone they also liked) relative to received liking (i.e., being liked by someone whom they did not like). This association held after controlling for severity of current depressive symptoms, although depressive severity was also associated with greater mPFC response. Adolescents with higher levels of social anhedonia also had stronger positive connectivity between the nucleus accumbens and the mPFC during mutual versus received liking. These results, the first on the pathophysiology of adolescent anhedonia, support altered neural reward-circuit response to social reward in young people with social anhedonia.

Introduction

Anhedonia, or difficulty experiencing pleasure in anticipation or response to rewarding stimuli, is a cardinal feature of depression. Across the lifespan, depression is a leading cause of disability and suffering (World Health Organization, 2011). In addition to characterizing depression, anhedonia is associated with the onset, course, and outcome of the disease. Low levels of anhedonia in people with depression are protective, presaging better outcomes (Joiner, Lewinsohn, & Seeley, 2002) and a lower chance of recurrence (Kasch, Rottenberg, Arnow, & Gotlib, 2002). The clinical relevance of anhedonia is particularly important in young people, as anhedonia predicts the onset of clinical-level depression in children and adolescents (Pine et al., 1999, Wilcox and Anthony, 2004) and resistance to standard depression treatment in adolescents (McMakin et al., 2012). Despite anhedonia’s associations with the development of depression, a disorder that most commonly begins during adolescence (Lewinsohn, Clarke, Seeley, & Rohde, 1994), relatively little work has been done to examine the neural correlates of anhedonia in adolescence.

Anhedonia is relevant to appetitive and consummatory aspects of reward function, which reflect the motivation to pursue rewarding experiences and the enjoyment of rewarding experiences once obtained, respectively. Frontostriatal circuits, including dopamine-targeted regions such as the ventral striatum (VS) and medial prefrontal cortex (mPFC), play critical roles in reward function. Specifically, the striatum contributes to the motivation to approach rewarding stimuli and the enjoyment of those stimuli, and a subregion of the mPFC (i.e., Brodmann Areas 24, 32, and medial 10) contributes to the regulation of reward function, partly through its input to the striatum (Haber & Knutson, 2010). Broadly, meta-analytic evidence indicates that depression involves low striatal response to monetary reward for youth and adults (Zhang, Chang, Guo, Zhang, & Wang, 2013). More specifically, Wacker, Dillon, and Pizzagalli (2009) found that self-reported anhedonia—but not other symptoms of depression—was associated with nucleus accumbens activation to reward receipt during a monetary incentive delay task. In that study, anhedonia was negatively correlated with self-reported positive affect, indicating that affective aspects of anhedonia are related to neural response to reward.

Consistent with anhedonia’s subjective, behavioral, and neural aspects, adolescents with depression experience lower subjective positive affect (Lonigan et al., 2003, Silk et al., 2011), express shorter-duration, lower-intensity positive affect (Sheeber et al., 2009), and exhibit low striatal response to reward (see Forbes & Dahl, 2012). Specific associations between anhedonia and neural response to reward have yet to be reported in adolescents, but the literature on clinical-developmental neuroscience suggest that adolescence is an important developmental period for investigating anhedonia. Reward-related behavior and affect change markedly during adolescence, with increased risky, reward-seeking activities (see Somerville, Jones, & Casey, 2010), more sensation-seeking (Steinberg et al., 2008), and stronger experience of rewards (Ernst et al., 2005, Steinberg et al., 2008). Strikingly, these changes occur in tandem with apparent decreases in reward responding, including low levels of subjective positive affect (e.g., Larson, Moneta, Richards, & Wilson, 2002), and increasing levels of depressive symptoms (Sawyer, Pfeiffer, & Spence, 2009). Developmental neuroimaging studies indicate that adolescents show altered response to reward in the striatum (Bjork et al., 2004, Ernst et al., 2005, Forbes et al., 2010, Galván et al., 2006) and mPFC (Bjork et al., 2004, Forbes et al., 2010).

Response in the mPFC may be particularly relevant to adolescent anhedonia. In addition to its role in reward processing as the destination of the mesocortical dopamine pathway, the mPFC is a key region in the brain’s default-mode network and is implicated in affect regulation and in processing social and self-relevant stimuli (Amodio and Frith, 2006, Denny et al., 2012). In responding to rewarding events, mPFC response could therefore indicate how relevant the reward is to one’s preferences and whether the reward enhances one’s status among others. Depression is associated with greater mPFC response to reward, which could reflect processes such as difficulty shifting out of a negative, self-focused pattern of thought or over-regulation of more basic, striatal response to reward (Forbes & Dahl, 2012). Notably, anhedonia itself is associated with greater mPFC response to positive autobiographical memories but less mPFC response to negative stimuli (Keedwell, Andrew, Williams, Brammer, & Phillips, 2005b). This pattern is similar to that observed in adults with depression but the opposite of that exhibited by healthy adults (Keedwell, Andrew, Williams, Brammer, & Phillips, 2005a). This suggests that neural function underlying anhedonia, especially in the context of depression, is disrupted in response to what are generally highly valued, self-defining experiences. Furthermore, this disruption in the mPFC could reflect a pattern of responding to such typically pleasant, poignant stimuli as somehow aversive.

In addition to specific regional responding, anhedonia is postulated to involve altered connectivity between the mPFC and VS, such that the mPFC might serve to dampen VS responses to reward (Forbes & Dahl, 2012). The disruption of functional connectivity between the mPFC and VS during response to monetary reward in adolescents with a history of depression (Morgan et al., submitted for publication) suggests that such functional connectivity—likely via altered input to the VS from the ventral tegmental area—could be a mechanism for the development of anhedonia. Neural response to reward and functional connectivity between regions in rewarding contexts answer different questions: the first, about response in specific regions in isolation, and the second, about coordination between regions during specific contexts. Both of these can address function in reward circuitry, but they provide different information. Given that previous conceptual models have postulated that both neural response and frontostriatal connectivity are disrupted in anhedonia and in those who develop depression, investigations with anhedonia in adolescents are crucial to critically test developmental hypotheses.

Although functional connectivity techniques such as psychophysiological interaction (PPI) have been traditionally used to examine how regions coordinate in response to task context, there is a burgeoning mental health literature in which investigations of brain circuitry in psychopathology include examining whether psychopathology is related to functional connectivity within the circuitry of interest. For example, studies of affective disorders have reported differences between subgroups of patients or between patients and healthy comparison participants in functional connectivity between the amygdala and orbitofrontal or prefrontal cortex during the processing of affective faces (e.g., Almeida et al., 2009, Almeida et al., 2011, Kong et al., 2013, Versace et al., 2010, Wang et al., 2012). Similarly, in schizophrenia, another disorder that includes the symptom of anhedonia, group differences in functional connectivity have been reported during experiences such as working memory and social processing (e.g., Eack et al., 2013, Mukherjee et al., 2013, Straube et al., 2013). Aside from categorical group differences, studies of mental health have also employed functional connectivity to examine function in neural circuitry that varies with a dimensional characteristic, such as neuroticism or symptom severity (Cremers et al., 2010, Davey et al., 2012, Doucet et al., 2013, Servaas et al., 2013, Yue et al., 2013). Several studies have now used both approaches, describing functional connectivity as a correlate of both between-group differences and continuous, within-sample variability (e.g., Davey et al., 2012). Most relevant to our questions in the current study, a study of adults recently reported that trait anhedonia is associated with altered response and altered functional connectivity in contexts involving pleasant musical stimuli (Keller et al., 2013).

Research with adolescents introduces important considerations. One is the definition of adolescence itself, which has been debated in the fields of psychology, anthropology, and pediatrics, among others. Adolescence is defined as the period between the end of puberty and the attainment of adult-level status and competence. Specifying an age range for this developmental period requires consideration of a variety of factors (e.g., psychological and biological processes), as well as consideration of the ongoing developmental tasks. For research purposes, it is particularly important to consider the ongoing developmental tasks that are relevant to a research question when defining an adolescent population. Given that the developmental tasks of adolescence are postulated to include impulse control, accurate assessment of risk vs. reward, and affect regulation during challenges (Hazen, Schlozman, & Beresin, 2008) and that the neural circuitry underlying these cognitive and behavioral functions continues to develop throughout the teen years and into the 20s (Lenroot & Giedd, 2006), we defined adolescence as occurring through the early 20s. That is, our deliberate focus on a population in which the processes of interest (i.e., function in reward circuitry, social processing) have not yet reached adult levels led us to consider this a study of adolescent development. Other approaches might classify participants over age 18 as adults based on legal or cultural changes in status at that age (e.g., attaining the right to vote or perform military service). It is also notable that many psychology studies conducted with undergraduate samples describe their participants as adults rather than late adolescents, and this practice has been identified as a key limitation for interpreting findings on constructs that involve self and social cognition (Sears, 1987). Other terms have also been used for the late adolescent developmental period, such as “emerging adulthood” (Conger & Little, 2010). Based upon our focus on neural reward circuitry and the emergence of a symptom relevant to several forms of psychopathology that have onset during the later stages of brain development, we studied a population from age 18–21 years, termed hereafter as late adolescents.

In addition, investigating function in reward circuitry during adolescence requires sufficient context. First, reward circuitry is undergoing substantial development during adolescence (Spear, 2000). The function of the striatum and mPFC in response to reward changes with puberty (Forbes, 2009) and with adolescent development (Bjork et al., 2004, Ernst et al., 2005, Galván et al., 2007). Thus, these brain regions are especially malleable and, consequently, potentially vulnerable to developing dysfunction during this period (Davey, Yücel, & Allen, 2008). Second, the value and salience of social contexts increases as adolescents spend more time with peers, experience more intense emotions in peer contexts, and pursue newly valued abstract social rewards (e.g., initiating a romantic relationship; Furman and Simon, 1999, Larson et al., 1999, Steinberg, 2008). Hence, a class of rewards that is perhaps most important to anhedonia, to the function of frontostriatal reward circuitry, and to adolescent development is social reward. In particular, adolescents’ reward circuitry and decision-making behavior are sensitive to peer context (Albert et al., 2013, Chein et al., 2011), and peer experiences are postulated to influence the development of reward circuitry (Albert et al., 2013). Thus, peer social reward is likely to be salient during this developmental period.

The model presented by Davey et al. (2008) may be further elaborated to emphasize the role of being accepted by peers when peers’ opinions are highly valued, as opposed to being minimally valued or ambiguous. Accordingly, Davey, Allen, Harrison, Dwyer, and Yücel (2010) developed a social reward paradigm in which participants view images of unfamiliar peers and rate how much they think they would like the peer. Participants are led to believe that their own images will also be evaluated by these unfamiliar peers. In a separate fMRI assessment session, participants receive feedback about whether these unfamiliar peers liked them (i.e., social reward) or whether unfamiliar peers were unable to provide ratings (i.e., ambiguous outcome). A key feature of this paradigm is the participants’ own ratings of the unfamiliar peers, which allows the neural response to both acceptance in general and to mutual acceptance to be investigated. Indeed, Davey et al. (2010) found that being liked activates primary reward regions, such as the nucleus accumbens and vmPFC, for 15- to 24-year-old adolescents and young adults. Furthermore, mutual acceptance, or being liked by highly regarded peers, activated the mPFC more strongly in that study than being liked by less-liked peers. Response to social reward could be particularly relevant to social anhedonia, or difficulty experiencing pleasure from social contact (Blanchard, Gangestad, Brown, & Horan, 2000). Whereas previous studies have focused on anhedonia as broadly related to hedonic capacity—a construct that can include physical, affective, and social components—the importance of social reward to adolescents and the disruption of social functioning in depression together point to the importance of understanding neural response to social reward in relation to social anhedonia.

Using an fMRI paradigm involving feedback about peer social acceptance—specifically, presenting whether unfamiliar peers the participant had rated for their likeability also rated the participant as likeable—we examined the association of social anhedonia with adolescents’ mPFC response and functional connectivity during social reward contexts. Specifically, we examined functional connectivity between the ventral striatum and the mPFC, two key regions in reward circuitry that are also relevant to processing social reward and likely to respond in tandem to social reward. We hypothesized that late adolescents’ social anhedonia would be associated with higher mPFC response during the social reward of mutual acceptance, as opposed to acceptance from peers whom participants liked less. In addition, to confirm that mutual liking—rather than all positive feedback—was particularly associated with altered mPFC response, we examined whether mPFC response to social reward in general was associated with social anhedonia.