Elsevier

Neuropsychologia

Volume 50, Issue 5, April 2012, Pages 567-575
Neuropsychologia

Reviews and perspectives
Reward processing in anorexia nervosa

https://doi.org/10.1016/j.neuropsychologia.2012.01.036Get rights and content

Abstract

Individuals with anorexia nervosa (AN) demonstrate a relentless engagement in behaviors aimed to reduce their weight, which leads to severe underweight status, and occasionally death. Neurobiological abnormalities, as a consequence of starvation are controversial: evidence, however, demonstrates abnormalities in the reward system of patients, and recovered individuals. Despite this, a unifying explanation for reward abnormalities observed in AN and their relevance to symptoms of the illness, remains incompletely understood. Theories explaining reward dysfunction have conventionally focused on anhedonia, describing that patients have an impaired ability to experience reward or pleasure. We review taste reward literature and propose that patients’ reduced responses to conventional taste-reward tasks may reflect a fear of weight gain associated with the caloric nature of the tasks, rather than an impaired ability to experience reward. Consistent with this, we propose that patients are capable of ‘liking’ hedonic taste stimuli (e.g., identifying them), however, they do not ‘want’ or feel motivated for the stimuli in the same way that healthy controls report. Recent brain imaging data on more complex reward processing tasks provide insights into fronto-striatal neural circuit dysfunction related to altered reward processing in AN that challenges the relevance of anhedonia in explaining reward dysfunction in AN. In this way, altered activity of the anterior cingulate cortex and striatum could explain patients’ pathological engagement in behaviors they consider rewarding (e.g., self-starvation) that are otherwise aversive or punishing, to those without the eating disorder. Such evidence for altered patterns of brain activity associated with reward processing tasks in patients and recovered individuals may provide important information about mechanisms underlying symptoms of AN, their future investigation, and the development of treatment approaches.

Highlights

► First comprehensive review of neuroimaging of reward processing in anorexia nervosa. ► Challenges prevailing neurocognitive theories for reward processing in anorexia nervosa. ► Redefines the relationship between reward processing and symptoms of the illness.

Introduction

Anorexia nervosa (AN) is diagnosed in females in up to 95% of cases (DSM-IV-TR, 2000) with a prevalence of 0.9% (Hudson, Hiripi, Pope, & Kessler, 2007). AN is characterized behaviorally by self-starvation and excessive exercise in up to 80% of patients (Dalle Grave et al., 2008, Davis, 1997) and has the highest mortality of any psychiatric illness (Sullivan, 1995). There are two types of eating related-behaviors common to AN. Patients with restricting type anorexia primarily lose weight by dieting without binge eating or purging whereas, binge-eating/purging type patients also restrict their food intake to reduce their weight but also periodically engage in binge eating and/or purging as do individuals with bulimia nervosa (Kaye, Fudge, & Paulus, 2009). Although it is well accepted that some patients can experience both syndromes, consistent with shared risk and liability factors (Lilenfeld et al., 1998, Walters and Kendler, 1995), the current review focuses on restricting type AN.

Abnormalities in several neural systems have been identified in patients with AN including: serotonin, dopamine, hypothalamopituitary adrenal (HPA) and hypothalamopituitary gonadal (HPG) axes, appetite related neuropeptides and other neurochemical systems (Barbato et al., 2006, Kaye, 2008, Monteleone et al., 2001, Sodersten et al., 2008). Consistent with these observations, there are many frameworks that have been proposed to explain AN. Examples of these frameworks include: evolutionary explanations that excessive exercise observed in the illness may be ‘displaced food-foraging behavior’ akin to times of food shortage (Sodersten et al., 2008). Disturbances in appetite-related hypothalamic pathways (Kaye et al., 1990, Sodersten et al., 2008) which may contribute to disordered behavior toward food (Sodersten et al., 2008) which tend to be restored following recovery (Sodersten et al., 2008). In addition, serotonin concentrations are hypothesized to inhibit appetite (Simansky et al., 2004) and may lead to anxiety-avoidance behavior toward food (Kaye et al., 2009). Consistent with this, starvation induced reductions in extracellular serotonin concentrations have been proposed to decrease dysphoric symptoms (Kaye et al., 2009). Variations of the anhedonia hypothesis contend that patients engage in disordered behaviors to alleviate a dysphoric and anhedonic mood state (Davis and Woodside, 2002, Kaye, 2008, Kaye et al., 2009) with multiple neurobiological mechanisms (including dopamine (Frank et al., 2005) and serotonin (Kaye et al., 2009)) proposed in these hypotheses.

This review focuses on the framework proposing that behaviors linked to AN (self-starvation and excessive exercise) upregulate the stress hypothalamopituitary adrenal (HPA) axis, the consequence of which is increased dopamine secretion from the ventral striatal terminals of mesolimbic neurons in the brain (Bergh & Sodersten, 1996). The experience of reward associated with these behaviors is proposed to reinforce the illness, plunging patients into a severely emaciated state. Involvement of the HPA axis and dopamine systems have been proposed in the onset of AN (Bergh & Sodersten, 1996). Whether a sensitive HPA axis or dopamine system, or their interaction renders some patients more vulnerable to develop behaviors that ultimately become pathological is not known. The majority of research tends to show that elevated concentrations of cortisol (Boyar et al., 1977, Casper et al., 1979, Gerner and Wilkins, 1983, Gold et al., 1986, Misra et al., 2004, Monteleone et al., 2011) normalize following weight-recovery (Gold et al., 1986, Gwirtsman et al., 1989, Walsh et al., 1981). Nevertheless, supporting a contribution of stress and dopamine systems to the illness the majority of patients with AN show abnormally elevated HPA axis activity (Boyar et al., 1977, Hotta et al., 1986, Licinio et al., 1996, Monteleone et al., 2001) in addition to which abnormalities in the dopamine system and relevant brain regions linked to reward processing have been revealed (Barbato et al., 2006, Fladung et al., 2010, Frank et al., 2005). The majority of research suggests that dopamine is abnormal in patients with AN as assessed via eye-blink rate (Barbato et al., 2006) as well as evidence for D2 receptor linked genetic polymorphisms (Bergen et al., 2005). In addition, in recovered individuals, abnormalities in the dopamine system have been revealed via reduced D2 and D3 receptor binding in the reward linked ventral striatum (Frank et al., 2005) (Kaye, Frank, & McConaha, 1999). Despite this, research has not yet directly investigated the dopamine system during reward processing tasks in patients with AN, or recovered individuals.

Evidence that dopamine abnormalities are present in the recovered state suggests that changes in reward linked system may be a trait marker for AN. Indeed, changes in reward processing generally have recently been proposed to be a neural biomarker for AN (Cowdrey, Park, Harmer, & McCabe, 2011). There are however, few frameworks in which to explain the clinical implications of reward-linked abnormalities that have been revealed in the literature. Recently, the concept of stress induced reward linked to illness behaviors (Bergh & Sodersten, 1996) has been developed to provide an explanation for why patients are able to engage in these behaviors despite them being highly pathological. It has been termed reward contamination theory (Keating, 2010). Reward contamination theory proposes to explain the observation that many patients ultimately experience otherwise rewarding stimuli as punishing, for example food, and vice versa otherwise punishing (or aversive) stimuli as rewarding, for example self-starvation, excessive exercise and emaciated body image (Keating, 2011). Reward contamination theory suggests that while patients’ pathological behaviors are in the first instance rewarding, when relentlessly engaged in, they become reinforced in a manner that becomes pathological (Keating, 2010). Patients, however, may not recognize that they are contaminating aspects of reward with punishment, due to overlapping neurocircuits that process reward and punishment (e.g., dopamine), which may facilitate neural and behavioral reinforcement, thus impairing patients’ ability to regulate their behaviors (Keating, 2010). The anterior cingulate cortex, within a broader fronto-striatal neurocircuit is proposed to represent a key locus for reward-contamination (Keating, 2010).

Since reward contamination theory was first published (Keating, 2010) there have been a number of neuroimaging studies undertaken and published, the results of which, we hypothesize, are consistent with the principles of reward contamination theory. Although the clinical literature base of dopamine-related investigations in patients with AN are relatively few, this system may represent a candidate substrate for dysfunctional reward processing in AN. Theories regarding mechanisms that underlie the expression of AN symptoms, which include reward contamination theory, could have important implications for the development of novel interventions. This review systematically assesses reward processing in AN. We include studies involving food-reward and non-food reward tasks, as well as recent advances involving neuroimaging. We discuss whether theories including anhedonia, a fear of weight gain, or reward contamination (reward conflict) comprehensively account for patients’ responses to reward processing tasks. Future directions for research are considered consistent with elucidating mechanisms underlying symptoms of AN. Understanding such mechanisms may provide insight for the development of novel therapies.

Section snippets

Taste-reward processing

Rewards are associated with a subjective feeling of pleasure (Wise, 2004) or hedonism. Anhedonia is the inability to experience pleasure or reward (Wise, Spindler, deWit, & Gerberg, 1978). Anhedonia is an explanatory model for reward-linked abnormalities reported in AN, suggesting that patients have an impaired ability to experience pleasure or reward. Observation (Kaye, 2008) and empirical evidence (Davis & Woodside, 2002) show that some patients with AN experience greater anhedonia (Davis and

Complex taste-reward tasks: reward is not always what it seems

More recently, research has extended our understanding of reward processing to more complex cognitive tasks, including responses to reward and aversive/conflicted conditions. For example, investigators used disorder-relevant stimuli, chocolate and moldy strawberries, during fMRI to study 15 recovered individuals relative to 16 health controls’ matched for age and body mass index. The rewarding condition was the sight and flavor of chocolate and their combination, and the aversive condition

Beyond taste-reward processing

To more comprehensively understand reward processing in AN, we have considered the scope of patients’ abnormal experiences of reward beyond food hedonic-stimuli. In the context of reward linked to AN, researchers have investigated reward-processing specific to stimuli that translate to symptoms of the illness. For example, patients find emaciated body images and self-starvation to be inherently reinforcing and rewarding, consistent with clinical features of the illness. Recent neuroimaging

Dopamine and reward processing

Neuroimaging investigations have consistently demonstrated differences in patients and recovered individuals compared to HCs in areas of the striatum and more executive regions (e.g., ACC) during reward processing. The precise neurochemistry, however, providing a substrate for the differences reported in fMRI investigations, remains unclear. Dopamine has been the most widely studied reward candidate in AN. Increased D2 and D3 receptor binding in the striatum has been revealed in recovered

Summary of reward processing in AN

In light of the literature reviewed, taste-reward tasks reveal abnormalities in the way in which patients with AN experience simple stimuli (Drewnowski et al., 1987, Simon et al., 1993, Stoner et al., 1996). Although ‘liking’ or appropriately identifying/judging a taste-reward remains intact, patients experience reduced preferences (‘wants’) to the stimuli, relative to HCs (Drewnowski et al., 1987, Simon et al., 1993, Stoner et al., 1996). Whereas anhedonia has been proposed as an explanation

Limitations of reward contamination theory

There have not been any studies specifically designed to investigate reward contamination theory, despite clear evidence for altered reward processing, which is consistent with reward contamination as an explanation. Although the outcomes of reward processing investigations can be explained by reward contamination, it is also not yet clear from the evidence whether reward contamination begins as a result of an aversion to food, or increasing social/cognitive goals of thinness. Whether a

Conclusions and future directions

The outcomes of our assessment of reward processing literature in AN (and recovered individuals) in the context of competing theories for reward processing in AN, demonstrates that anhedonia is limited in its capacity to explain patients’ experience of simple taste-reward tasks. The results of these conventional taste-reward tasks, when interpreted in the context of modern theories for reward processing (e.g., liking and wanting) reflect that patients and HCs similarly rate the hedonic

Financial disclosures

The authors, C.K., A.J.T., S.L.R., P.G.E. of this manuscript report no biomedical financial interests or potential conflicts of interest, in association with the content of this manuscript. P.G.E. is supported by an NHMRC Clinical Research Fellowship. P.B.F. is supported by a NHMRC Practitioner Fellowship and has received equipment for research from Magventure A/S and Brainsway.

References (67)

  • H.C. Friederich et al.

    I’m not as slim as that girl: Neural bases of body shape self-comparison to media images

    Neuroimage

    (2007)
  • J.I. Hudson et al.

    The prevalence and correlates of eating disorders in the National Comorbidity Survey Replication

    Biological Psychiatry

    (2007)
  • W. Kaye

    Neurobiology of anorexia and bulimia nervosa

    Physiology and Behavior

    (2008)
  • W.H. Kaye et al.

    Altered dopamine activity after recovery from restricting-type anorexia nervosa

    Neuropsychopharmacology

    (1999)
  • C. Keating

    Theoretical perspective on anorexia nervosa: The conflict of reward

    Neuroscience and Biobehavioral Reviews

    (2010)
  • S. Kojima et al.

    Comparison of regional cerebral blood flow in patients with anorexia nervosa before and after weight gain

    Psychiatry Research: Neuroimaging

    (2005)
  • J. Licinio et al.

    The hypothalamic-pituitary-adrenal axis in anorexia nervosa

    Psychiatry Research

    (1996)
  • T.E. Robinson et al.

    The neural basis of drug craving: An incentive-sensitization theory of addiction

    Brain Research. Brain Research Reviews

    (1993)
  • K.J. Simansky et al.

    A 5-HT2C agonist elicits hyperactivity and oral dyskinesia with hypophagia in rabbits

    Physiology and Behavior

    (2004)
  • P. Sodersten et al.

    Behavioral neuroendocrinology and treatment of anorexia nervosa

    Frontiers in Neuroendocrinology

    (2008)
  • S.R. Sunday et al.

    Taste perceptions and hedonics in eating disorders

    Physiology and Behavior

    (1990)
  • A. Takano et al.

    Abnormal neuronal network in anorexia nervosa studied with 1-123-IMP SPECT

    Psychiatry Research: NeuroImaging

    (2001)
  • J. Wrase et al.

    Different neural systems adjust motor behavior in response to reward and punishment

    Neuroimage

    (2007)
  • A. Beck et al.

    Ventral striatal activation during reward anticipation correlates with impulsivity in alcoholics

    Biological Psychiatry

    (2006)
  • A.W. Bergen et al.

    Association of multiple DRD2 polymorphisms with anorexia nervosa

    Neuropsychopharmacology

    (2005)
  • C. Bergh et al.

    Anorexia nervosa, self-starvation and the reward of stress

    Nature Medicine

    (1996)
  • K.C. Berridge et al.

    Taste reactivity analysis of 6-hydroxydopamine-induced aphagia: Implications for arousal and anhedonia hypotheses of dopamine function

    Behavioral Neuroscience

    (1989)
  • R.M. Boyar et al.

    Cortisol secretion and metabolism in anorexia nervosa

    New England Journal of Medicine

    (1977)
  • S.J. Brooks et al.

    Differential neural responses to food images in women with bulimia versus anorexia nervosa

    PLoS One

    (2011)
  • R.C. Casper et al.

    Alterations in serum cortisol and its binding characteristics in anorexia nervosa

    Journal of Clinical Endocrinology and Metabolism

    (1979)
  • L.J. Chapman et al.

    Scales for physical and social anhedonia

    Journal of Abnormal Psychology

    (1976)
  • Cowdrey, F. A., Harmer, C. J., Park, R. J., & McCabe, C. Neural responses to emotional faces in women recovered from...
  • F.A. Cowdrey et al.

    Increased neural processing of rewarding and aversive food stimuli in recovered anorexia nervosa

    Biological Psychiatry

    (2011)
  • Cited by (107)

    View all citing articles on Scopus
    View full text