Elsevier

Brain and Cognition

Volume 89, August 2014, Pages 3-14
Brain and Cognition

A cross-sectional and longitudinal analysis of reward-related brain activation: Effects of age, pubertal stage, and reward sensitivity

https://doi.org/10.1016/j.bandc.2013.10.005Get rights and content

Highlights

  • We studied adolescents’ reward-related neural responses and risky choice.

  • Cross-sectional reward-activation was positively related to risky choice.

  • Functional connectivity strength between striatum and insula related negatively to risky choice.

  • Reward regions such as striatum and medial PFC were consistently activated over time.

  • Longitudinal change in striatal reward-activation related to individual’s change in funseeking.

Abstract

Neurobiological models suggest that adolescents are driven by an overactive ventral striatum (VS) response to rewards that may lead to an adolescent increase in risk-taking behavior. However, empirical studies showed mixed findings of adolescents’ brain response to rewards. In this study, we aimed to elucidate the relationship between reward-related brain activation and risky decision-making. In addition, we examined effects of age, puberty, and individuals’ reward sensitivity. We collected two datasets: Experiment 1 reports cross-sectional brain data from 75 participants (ages 10–25) who played a risky decision task. Experiment 2 presents a longitudinal extension in which a subset of these adolescents (n = 33) was measured again 2 years later. Results showed that (1) a reward-related network including VS and medial PFC was consistently activated over time, (2) the propensity to choose the risky option was related to increased reward-related activation in VS and medial PFC, and (3) longitudinal comparisons indicated that self-reported reward sensitivity was specifically related to VS activation over time. Together, these results advance our insights in the brain circuitry underlying reward processing across adolescence.

Introduction

Adolescence is characterized as a period of hormonal changes and pronounced changes in social-affective engagement such as increases in sensation seeking and risk taking. Neurobiological models of adolescent development have suggested that adolescents are more sensitive to rewards due to a relatively increased limbic response in combination with reduced down-regulation by the prefrontal cortex and other cortical areas (Ernst and Fudge, 2009, Nelson et al., 2005, Somerville et al., 2010). Accordingly, these models suggest that such neurobiological changes may underlie typical adolescents’ risky behaviors such as substance abuse, unsafe sexual behavior, and reckless driving (Dahl, 2004; Steinberg et al., 2008).

A typically found ‘reward-network’ in the brain includes dopamine-rich areas in the midbrain and their targets: striatum and medial prefrontal cortex (Blakemore and Robbins, 2012, Clark et al., 2009, Tom et al., 2007). More specifically, ventral striatum (VS) has been implicated in anticipating and processing different types of rewards, as well as in producing learning signals known as prediction errors (Cohen et al., 2010, Delgado, 2007, Galvan et al., 2005, Knutson et al., 2001). Similarly, medial PFC – specifically the part that overlaps with the anterior cingulate cortex (ACC) – is also related to prediction-error coding (Van den Bos, Cohen, Kahnt, & Crone, 2012), but also to action-related reward associations (Kennerley and Walton, 2011, Rushworth et al., 2011), and detecting the need for increased control (Ridderinkhof, Ullsperger, Crone, & Nieuwenhuis, 2004). In contrast, a more ventral region of the medial prefrontal cortex, adjacent to medial orbital frontal cortex, has been implicated in coding rewards and is linked to representations of ‘value’ (Kuhnen and Knutson, 2005, McKell Carter et al., 2010). Moreover, research indicates strong interconnections between the VS and several parts of the medial PFC. These so-called striatal-cortical loops may be important for regulating reward-related responses and subsequent goal-directed behavior (Haber & Knutson, 2010). Together, these findings suggest that goal-directed behavior (e.g., risk taking) is driven by a reward-valuation system, in which VS encodes the more ‘basic’ aspects of reward and medial PFC integrates the different aspects of the reward to represent its subjective value and is important for selecting actions and controlling behavior.

Results of previous developmental functional MRI studies suggest that adolescent decision-making may be biased by a relatively hypersensitive VS response to rewards. That is, research has indicated that adolescents (ages 13–17 years) show a larger VS response to rewards compared to children and adults (Galvan et al., 2006, Padmanabhan et al., 2011, Van Leijenhorst et al., 2010, Van Leijenhorst et al., 2010). However, other studies have indicated striatal hypo-activation in adolescents during reward anticipation (Bjork et al., 2004, Bjork et al., 2010) or have shown little differences between adolescents and adults in VS response to rewards (May et al., 2004, Paulsen et al., 2012). Moreover, only some studies have found that the VS response to rewards correlates with risk-taking behavior in every-day life (Galvan, Hare, Voss, Glover, & Casey, 2007). Thus, several questions remain with respect to the specificity of the VS and medial PFC responses to rewards in adolescence and their relationship to risky behavior. For instance, it remains to be determined whether higher risk-taking in adolescence is associated with a higher VS response to rewards, a lower medial PFC response, or less functional connectivity between these areas (see also Cohen et al., 2012, Van den Bos et al., 2012).

Mixed findings in adolescents’ reward-related brain activation might have several causes, such as differences in task design and analyses (Galvan, 2010). In addition, prior contradictory findings may point toward individual differences in adolescence (Somerville et al., 2010). One important source of influences on subcortical and cortical responses could be pubertal development, which may serve as an important individual difference measure in adolescents’ brain activation in response to rewards and appetitive cues. That is, gonadal hormone levels significantly increase during adolescence and have both organizational and activating effects on brain functioning (Blakemore et al., 2010, Sisk and Zehr, 2005). For instance, higher testosterone levels have been associated with increased VS activation (Forbes et al., 2010, Op de Macks et al., 2011) and to adolescent typical risk-behavior such as experimentation with alcohol (De Water, Braams, Crone, & Peper, 2013).

Another possible source to explain individual differences in reward-related brain activation could be a persons’ sensitivity to rewards. For instance, prior studies reported that activation in the VS correlated positively with self-reported (1) reward sensitivity, as measured by the behavioral approach system (BAS) scale (Beaver et al., 2006), (2) sensation seeking, as measured by the brief sensation-seeking scale (Bjork, Knutson, & Hommer, 2008), (3) impulsivity, as measured by the psychopathic personality inventory (Buckholtz et al., 2010), and (4) real-life risk taking (Galvan et al., 2007). Possibly, these personality differences in reward-related response tendencies may explain why some adolescents are more responsive to rewards than others.

In the current study we examined reward processing in adolescence in more detail. Specifically, we aimed to elucidate the relationship between reward-related brain activation, frontostriatal connectivity strength, and behavior. In addition, we focused on examining effects of age, pubertal development, and individual’s self-reported reward sensitivity on reward-related brain activation. To these ends, we report two experiments using a risky decision task, in which participants could choose to take a gamble (and win or lose 10 Eurocents) or pass on this gamble (in which case nothing was gained or lost). We were specifically interested in the brain’s response to rewards and losses as a result of an active gamble, since prior studies have shown that outcome monitoring is more salient when the outcomes are the result of an active choice (Rao et al., 2008, Tricomi et al., 2004).

In the first experiment, we reanalyzed the adolescent sample (ages 10–16 years) previously reported by Op de Macks et al. (2011) and added a young–adult sample (18–25 years). The study by Op de Macks et al. (2011) primarily examined individual differences in the reward-related brain activation in relation to testosterone levels, but made no age comparisons. In the current study, we studied age, puberty, and individual differences in reward sensitivity in the same sample. The second experiment included a longitudinal extension of Experiment 1. That is, a subset of the adolescents from Experiment 1 was re-invited 2 years later, and completed the same risky decision task. This combined cross-sectional/longitudinal approach presents unique insights in the development of the reward system across adolescence and allows us to link changes in reward-related activation to individual’s changes in behavior, age, pubertal development, and reward sensitivity.

Replicating prior studies, we expected to observe activation in VS and medial PFC when processing rewards. Second, we predicted that risk-taking propensity would be positively correlated with VS activation, negatively correlated with medial PFC activation and/or the strength of connectivity in this reward network. Third, based on prior findings we expected VS activation to change with age (quadratic or linear). Finally, we tested whether the VS response to rewards was related to pubertal development, or to self-reported reward-sensitivity (as measured with the self-report BAS scale).

Section snippets

Participants

Seventy-eight right-handed participants (50 adolescents, 28 adults) were scanned while performing a risky decision task. All participants reported an absence of neurological or psychiatric impairments (on a brief screening module) and provided written informed consent for the study (parental consent and participant assent for minors). The cross-sectional adolescent data has been reported before in Op de Macks et al. (2011), but that study focused primarily on the association between individual

Behavior

The average proportion of ‘play’ decisions was .67 (range = .28–1, SD = .14). A linear regression with proportion of plays as a dependent and Age as an independent variable showed no significant effect of Age (p’s > .1). Similar analyses with PDS score, and the BAS scales (Drive, Fun-seeking, and Reward-responsiveness) as an independent variable, also showed no significant effects of PDS or BAS scores on proportion of plays (p’s > .1). Together these results reveal that the tendency to make a risky

Participants

A subset of the adolescents from Experiment 1 (n = 33) were scanned again approximately 2 years later, and were administered the same risky decision task. The goal of this study was to extend this dataset with a longitudinal sample. All participants signed informed consent (parental consent and participant assent for minors) and procedures were approved by the local Medical Ethical Committee.

Two participants showed head motion exceeding 3 mm during scanning at time point 2 (T2) and were therefore

Behavior

The proportion of plays in the adolescent longitudinal sample was .62 (SD = .13) for T1 and .63 (SD = .11) for T2. A correlational analysis between T1 and T2 showed that proportion of plays was significantly correlated across sessions (r = .41, p < .02), however, this correlation also indicates there was a fair amount of within-individual differences in choice behavior across time.

A set of linear regressions with proportion of plays at each time point as a dependent and Age (continuous) at each time

General discussion

The goal of this study was to examine stability, change, and individual differences in reward processing in adolescence. We first examined the relation between brain and behavior in the context of reward processing and risky decision-making. Second, we examined the effects of age, pubertal development, and reward sensitivity on reward-related brain activation in a cross-sectional and longitudinal comparison. To these ends, Experiment 1 utilized a risky decision task in a cross-sectional sample

Acknowledgement

The authors would like to thank Wouter van den Bos for his help with functional connectivity analyses.

References (65)

  • A. Padmanabhan et al.

    Developmental changes in brain function underlying the influence of reward processing on inhibitory control

    Developmental Cognitive Neuroscience

    (2011)
  • M.P. Paulus et al.

    Increased activation in the right insula during risk-taking decision making is related to harm avoidance and neuroticism

    NeuroImage

    (2003)
  • H. Rao et al.

    Neural correlates of voluntary and involuntary risk taking in the human brain: An fMRI study of the balloon analog risk task (BART)

    NeuroImage

    (2008)
  • J. Richards et al.

    A systematic review of fMRI reward paradigms used in studies of adolescents vs. adults: The impact of task design and implications for understanding neurodevelopment

    Neuroscience and Biobehavioral Reviews

    (2013)
  • M. Rushworth et al.

    Frontal cortex and reward-guided learning and decision-making

    Neuron

    (2011)
  • C.L. Sisk et al.

    Pubertal hormones organize the adolescent brain and behavior

    Frontiers in Neuroendocrinology

    (2005)
  • L.H. Somerville et al.

    A time of change: Behavioral and neural correlates of adolescent sensitivity to appetitive and aversive environmental cues

    Brain and Cognition

    (2010)
  • E.M. Tricomi et al.

    Modulation of caudate activity by action contingency

    Neuron

    (2004)
  • L. Van Leijenhorst et al.

    Adolescent risky decision-making: Neurocognitive development of reward and control regions

    NeuroImage

    (2010)
  • J.D. Beaver et al.

    Individual differences in reward drive predict neural responses to images of food

    Journal of Neuroscience

    (2006)
  • G.S. Berns et al.

    Adolescent engagement in dangerous behaviors is associated with increased white matter maturity of frontal cortex

    PLoS ONE

    (2009)
  • J.M. Bjork et al.

    Incentive-elicited brain activation in adolescents: Similarities and differences from young adults

    Journal of Neuroscience

    (2004)
  • J.M. Bjork et al.

    Incentive-elicited striatal activation in adolescent children of alcoholics

    Addiction

    (2008)
  • J.M. Bjork et al.

    Adolescents, adults and rewards: Comparing motivational neurocircuitry recruitment using fMRI

    PLoS ONE

    (2010)
  • M.C. Brett et al.

    Region of interest analysis using an SPM toolbox

    NeuroImage

    (2002)
  • S.J. Blakemore et al.

    The role of puberty in the developing adolescent brain

    Human Brain Mapping

    (2010)
  • S.J. Blakemore et al.

    Decision-making in the adolescent brain

    Nature Reviews Neuroscience

    (2012)
  • J.W. Buckholtz et al.

    Mesolimbic dopamine reward system hypersensitivity in individuals with psychopathic traits

    Nature Neuroscience

    (2010)
  • Y.T. Cho et al.

    Nucleus accumbens, thalamus and insula connectivity during incentive anticipation in typical adults and adolescents

    NeuroImage

    (2012)
  • C.S. Carver et al.

    Behavioral inhibition, behavioral activation, and affective responses to impending reward and punishment: The BIS/BAS scales

    Journal of Personality and Social Psychology

    (1994)
  • C.A. Cocosco et al.

    Brain web: Online interface to a 3D MRI simulated brain database

    NeuroImage

    (1997)
  • J.R. Cohen et al.

    A unique adolescent response to reward prediction errors

    Nature Neuroscience

    (2010)
  • Cited by (80)

    • Puberty and risky decision-making in male adolescents

      2023, Developmental Cognitive Neuroscience
    View all citing articles on Scopus

    This research was supported by a VIDI grant and an ERC grant awarded to Eveline A. Crone.

    1

    Both authors contributed equally.

    View full text