«by Amy Lynn Byrd, Ph.D. B.S. in Psychology, College of Charleston, 2006 M.S. in Clinical Psychology, University of Pittsburgh, 2010 Submitted to the ...»
A primary objective of this dissertation was to characterize the BOLD response to the receipt of reward and punishment among subgroups of boys with early-onset CP. It was hypothesized that youth with CP, regardless of their level of CU traits (or psychopathic features), would exhibit hyper-reactivity to reward in the VS relative to HC. Contrary to hypotheses, boys with CP and low CU traits exhibited reduced reactivity to big reward in the DS (i.e., caudate) relative to both HC and youth with CP and CU, who evidenced increased activation to reward in this same region. Moreover, youth with CP and low levels of CU traits exhibited a similar pattern of hypoactivation to reward in the amygdala and mPFC relative to HC and these differences were not evident when subgroups were characterized by the presence of psychopathic features. Taken together, results provide some initial evidence for potential differences in the mechanisms underlying reward processing in subgroups of youth with and without CU traits.
Abnormalities in punishment processing were expected in all ROIs examined and were hypothesized to be most pronounced in a subgroup of CP youth with elevated levels of CU traits (and psychopathic features). Results provided partial support of hypotheses and, consistent with the behavioral literature, were indicative of a primary deficit within this domain. Results were most reliable and robust with regard to responsivity to punishment in the amygdala. As predicted, boys with CP demonstrated reduced reactivity to punishment in this region relative to HC who evidenced significant activation. Additionally, CP youth evidenced reduced activation
analyses and in the opposite direction of hypotheses. Also contrary to prediction, there was no evidence to support that punishment insensitivity was most pronounced in CP youth with CU traits or psychopathic features. Moreover, no consistent differences emerged within regulatory regions (e.g., ACC, OFC) suggesting abnormalities during initial encoding are specific to subcortical regions. Lastly, differences in neural response were specific to big reward and big punishment and this may be a reflection of the task design rather that the magnitude of the monetary gain/loss. Specifically, small reward and punishment may be less salient given the possibility of larger reward/punishment. Results are discussed within the context of the broader literature below.
The second aim of the current dissertation was to evaluate the extent to which individual differences in reward and/or punishment processing were associated with responsiveness to intervention. This objective was notably exploratory in nature and sought to test the notion that abnormalities in reward and/or punishment processing may influence the impact of a multimodal intervention that focuses on behavioral principles associated with reward and punishment contingencies. Contrary to hypotheses, reactivity to reward and punishment among CP youth was unrelated to level of CP following treatment and abnormalities in neural processing failed to moderate the effectiveness of SNAP intervention. Though small sample sizes and reduced power may have contributed to null effects, results provide support for the success of this multi-modal, empirically supported intervention among boys with early-onset CP and aberrant reward/punishment processing. These findings are discussed in greater detail below.
Group differences in reward processing emerged within the striatum, providing some evidence for abnormalities in response to reward among youth with CP. However, contrary to prediction, differences were specific to the DS, namely the caudate, and were unrelated to VS reactivity.
This suggests group variation in reward processing may be less related to the emotional experiences of reward (Cardinal, et al., 2002) and instead specific to responsivity to uncertain reward and the acquisition of reward-action associations (Delgado, Miller, Inati, & Phelps, 2005;
Elliott, Newman, Longe, & William Deakin, 2004). Moreover, group differences emerged in the unexpected direction and between subgroups of CP youth, specific to the presence or absence of CU traits. Boys with CP and CU demonstrated increased activation in the caudate, no different than that of HC, but significantly greater than boys with CP and low levels of CU traits. Results also provide additional evidence for hypo-activity to reward among boys with CP and low levels of CU, who demonstrated reduced activation to reward in the amygdala and mPFC relative to HC. These deficits may have important implications for reward-based learning (Cardinal, et al., 2002; Fareri, et al., 2008; Matthys, Vanderschuren, & Schutter, 2012a) and suggest that CP youth with low levels of CU traits could experience difficulty in this regard.
At the same time, it is important to note that the aforementioned group differences were reduced to non-significance after controlling for potential confounds, with some suggestion that deficits in responsivity to reward were uniquely associated with clinically significant internalizing problems. This is particularly interesting given emerging literature that documents associations between internalizing problems, namely depression, and reduced responsivity throughout reward-related circuitry (Forbes, et al., 2006; Forbes & Dahl, 2005; Forbes, Shaw, & Dahl, 2007). In light of research documenting high rates of comorbidity between CP and
to disentangle whether potential deficits in reward processing represent a shared or unique etiology. As such, the utilization of alternative comparison groups (e.g., youth with internalizing problems only) may help to elucidate these findings.
It is also important to highlight that the hypo-activity to reward seen among CP youth with low CU stands in direct contrast to youth with CP and CU, who demonstrated more normative responsivity to reward in the caudate. Moreover, in continuous analyses CU traits were uniquely associated with increased activation in the caudate after controlling for CP, though this fell just below the cluster threshold. Additionally, whole-brain analyses found youth with CP and CU to exhibit hyper-activation to reward within the cingulate and postcentral gyrus. While these analyses were notably exploratory in nature and identify regions outside of the rewardrelated circuitry that could be less functionally significant, this pattern of over-activation in this subgroup may warrant further inquiry. Along these lines, recent neuroimaging work reports similar inconsistencies among samples of CP youth with regard to hyper- versus hypo-activation within reward related circuitry (e.g., Bjork, et al., 2010; Rubia, et al., 2009), which may be related to a failure to examine CU traits. Thus, while there continues to be debate about whether CP is driven by over- versus under-reactivity to reward (Quay, 1993; Zuckerman, 1996), it may be important to consider potential differences in the mechanisms underlying reward function between subgroups of CP youth with and without CU traits and/or with and without co-occurring internalizing problems.
Amygdala Group differences were expected be most pronounced and diffuse with regard to punishment processing. Results provide confirmation for reduced reactivity to punishment among boys with early-onset CP. Both group and continuous analyses support hypotheses and consistently found boys with CP to demonstrate reduced activation in the amygdala following the receipt of punishment. Moreover, these findings remained significant after accounting for important confounds (i.e., income, IQ, etc.). While other neuroimaging studies in this area have provided some support for amygdala dysfunction in response to punishment (e.g., Finger, et al., 2011), these studies have utilized relatively complex tasks that incorporate multiple phases of learning (i.e., encoding, acquisition, extinction) and often examine the neural response to removal of reward as opposed to the introduction of punishment (Bjork, et al., 2010; Rubia, et al., 2009). To our knowledge, this dissertation represents the first neuroimaging study to examine basic responsivity to the receipt of punishment in the form of monetary loss among CP youth in late childhood. Current findings build upon behavioral work in this area which consistently finds children with CP have lower reactivity to inherently aversive stimuli or positive punishment (e.g., lound tones; Herpertz, et al., 2001; van Goozen, et al., 2004), by demonstrating a similar insensitivity to negative punishment (i.e., loss of money). This reduced sensitivity to punishment has been well-documented across childhood and adolescence, with recent work suggesting that these deficits are present as early as 3 years of age and serve to predict criminal offending in adulthood (Gao, Raine, Venables, Dawson, & Mednick, 2010). Taken together, this provides evidence that reduced emotional arousal to both positive (i.e., introduction of something aversive) and negative (i.e., removal of something pleasurable) punishment may hinder the
ultimately increasing the likelihood of the development and persistence of CP (Kochanska, 1994).
In line with findings from the current dissertation, the amygdala has been a recent focus of the neuroimaging literature within this population, with studies consistently demonstrating links between amygdala dysfunction and heightened levels of CP in youth. This is consistent with theory proposing deficient processing within this region, particularly among youth with elevated CP and CU (Blair, 2007; Kiehl, 2006). For example, studies have shown structural differences in gray matter volume of the amygdala among youth with CP (De Brito et al. 2009;
Huebner et al. 2008; Sterzer et al. 2007) and adults with a history of early-onset CP (Pardini et al. 2013). Other work in this area has focused on amygdala reactivity during emotion processing tasks, specifically reactivity to emotional faces, and find youth with CP and high levels of CU traits or psychopathic features to evidence reduced amygdala reactivity relative to controls (for review see Hyde, Shaw, & Hariri, 2013; Jones, Laurens, Herba, Barker, & Viding, 2009; Marsh, et al., 2008). However, contrary to prediction, the current dissertation found no differences between subgroups of CP youth, whether defined by the presence of CU traits or psychopathic features. Noteworthy, the majority of previous research in this area compares youth with CP and CU traits (or psychopathic features) to HC, making it impossible to discern whether significant group differences are driven by the presence of CP or CU. The current dissertation is one of the only investigations to examine potential differences in neural processing between subgroups of CP youth (see also Hyde, 2012 for studies examining emotion processing among subgroups;
Viding, et al., 2012 ). Moreover, CP subgroups within the current study were indistinguishable in terms of demographic and other clinically relevant variables (e.g., internalizing problems,
of the rigorous study design, failure to detect significant within group differences is indicative of similar punishment processing among these subgroups of CP youth and suggests that previous findings may be attributable to the presence of CP as opposed to CU or psychopathic features.
Striatum In addition to robust associations between CP and reduced activation in the amygdala, the current dissertation also provided some evidence for reduced reactivity to punishment within the striatum, though this varied by location and subgroup. Specifically, CPCU- evidenced reduced activation in the caudate relative to HC (no differences with CPCU+) while CP PSY+ exhibited decreased activation in the putamen compared to HC (no differences with CP PSY-).
Inconsistency of results may reflect the lack of complete overlap between CU and PSY group classifications and certainly warrant caution with regard to interpretation. However, it is also important to note that the direction of these findings are contrary to previous work in the area (Finger, et al., 2008; Gatzke-Kopp, et al., 2009) and emerging theory (Glenn & Yang, 2012).
Specifically, past research has documented increased activation within the DS following punishment and this has been interpreted as an inability to appropriately process the absence of reward. In other words, CP youth may be processing punishment as if it was (or should be) reward due to ineffective error monitoring or a failure to process contingency change. This is thought to increase their propensity to continuously engage in perseverative, reward-focused action and may impair the ability to flexibly respond to the environment (Newman & Lorenz, 2002). Results in the current dissertation may reflect the unpredictable nature of the task design (i.e., no learning or contingency change) and thus, findings can only speak to an inherently
the presence of competing reward and punishment.