P e r s p e c t i v e s

, are often used to guide the balance SCIENCE & SOCIETY between seeking and avoiding (BOX 2). The subjective — or ‘mean- A common neurobiology for pain ing’ — of pain or pleasure for the individual is determined by sensory, homeostatic, cultural and other factors that, when com- and pleasure bined, bias the hedonic experience of pain or pleasure. Siri Leknes and Irene Tracey The Motivation-Decision Model Abstract | are powerful motivators of behaviour and have The processes that underlie the subjective historically been considered opposites. Emerging evidence from the pain and interpretation of a sensory event can be reward research fields points to extensive similarities in the anatomical substrates understood as the manifestation of an of painful and pleasant sensations. Recent molecular-imaging and animal studies unconscious decision process4,14. The deci- have demonstrated the important role of the and systems sion process requires information about in modulating both pain and pleasure. Understanding the mutually inhibitory the homeostatic state of the individual (such as or ), sensory effects that pain and reward processing have on each other, and the neural input and knowledge about impending mechanisms that underpin such modulation, is important for alleviating threats and available rewards. According unnecessary and improving well-being. to the Motivation-Decision Model of pain, as put forward by Fields4,14, the basic premise for the decision process is that “Nature has placed mankind under the The utility of pain and pleasure anything that is potentially more important governance of two sovereign masters, pain The large variability between the strength of for survival than pain should exert anti­ and pleasure.” — a sensory and the resulting hedonic nociceptive effects. This allows the animal is of great medical and neuroscien- to ignore the pain and attend to the more Most of what is known about pain and tific . For instance, athletes can be important event. The Motivation-Decision pleasure derives from the study of each oblivious to pain in the heat of competition, Model predicts that pain–pleasure dilem- phenomenon in . Recently, however, in which winning is the reward. A key factor mas in which a large reward is gained neuroscientists investigating opioid and for the interpretation of pain and pleasure is at the price of a small pain are resolved analgesia1–3, addiction4 and subjective utility7. For example, the reward through the antinociceptive effects of learning5 have begun to bridge the gap value of a stimulus increases with the effec- the pleasurable reward (FIG. 1). In some between the pain and pleasure research tiveness of that stimulus in restoring bodily instances, threatening and pleasure-related fields. This development has been equilibrium (homeostasis)6,8. This effect, cues are more important for survival than strengthened by the increasing focus on known as alliesthesia6, is well-documented pain, and it is assumed that any antinocic- the subjective emotional for food rewards, which are more pleasur- eptive effects are mediated by the descend- (hedonics) that are elicited by rewards able when they relieve a hunger state9. ing pain modulatory system, which is and (BOX 1). As the experience of pain represents a located in the . This circuit, which Rewards and punishments are defined deviation from homeostatic balance10, the consists of excitatory and inhibitory cells, as something that an animal will work to same principle can be applied to pain and communicates with in the pre- achieve or avoid, respectively. Pleasure the pleasantness of its relief 11. Similarly, frontal cortex, the and the represents the subjective hedonic value of when a perceived threat to an organism to control the nociceptive affer- rewards. The term ‘pain’ encompasses both becomes greater, pain unpleasantness ent pathway in the spinal and trigeminal the hedonic (suffering) and motivational increases, enhancing defensive and dorsal horn4,14,15 (FIG. 2). and (avoidance) aspects of a painful experience. avoidance mechanisms12. endogenous act on this descending Clearly, seeking pleasure and avoiding pain Pain and pleasure encourage the constant system to produce pharmacological, pla- is important for survival, and these two optimization of our internal homeostatic cebo, -induced and pleasure-related motivations probably compete for prefer- balance. Although pleasure-seeking and analgesia1,2,4,14–20. ence in the brain. Put simply, which of two pain-avoidance generally increase our coinciding pain and pleasure events should chances for survival, it is easy to envisage Pain–pleasure interactions be processed and acted on first? Consistent scenarios in which these two motivations Evidence of pleasure-related analgesia with the idea that a common currency of are in competition. A simple case would has been reported in various and emotion6 enables the comparison of pain involve a large reward that is only accessible animal studies: pain is decreased by pleas- and pleasure in the brain, the evidence at the ‘price’ of a small pain. Sometimes it ant odours21, images22, pleasurable music23, reviewed here points to there being exten- seems that overcoming a small amount palatable food16,17 and sexual behaviour18,19. sive overlap in the neural circuitry and of pain might even enhance the pleasure, In addition, considerable evidence suggests chemistry of pain and pleasure processing as reflected perhaps by the common that expectation of treatment effect, which at the systems level. This article summarizes expression ‘no pain, no gain’ or the pleasure contributes to placebo analgesia, is a type of current research on pain–pleasure interac- of eating hot curries. Pain–pleasure dilem- reward expectation24,25. Interestingly, when tions and the consequences for human mas abound in social environments13, and subjects who were not expecting an injec- behaviour. culture-specific moral systems, such as tion of pain-relieving received

314 | april 2008 | volume 9 www.nature.com/reviews/neuro © 2008 Nature Publishing Group

P e r s p e c t i v e s

Box 1 | The increasing focus on pain and reward hedonics access to our own hedonic and motivational processes, which are thought to be primarily Hedonic feelings — also known as — subconscious31. Importantly, however, the drive motivation and behaviour. Qualia motivation and hedonic subsystems seem to determine what it is like to be a human being87. No theory of the relationship between the be mediated by different neurotransmitters. brain and the is complete without Carefully controlled studies have found accounting for hedonic feelings. In recent specific effects for two neurotransmitter years, several exciting research directions have systems: dopamine increases motivation emerged in the pain and reward research fields for, but not the pleasure of, eating palatable that successfully combine the need for foods32,33, whereas the opioid system influ- carefully controlled, ‘objective’ research ences motivation indirectly by modulating methodologies with a focus on hedonics. One subjective emotional feelings of pain and example is a body of work on ‘liking’ and reward34. In summary, opioids are neces- ‘wanting’ — two subconscious reward sary for hedonic experience (‘liking’) but processes that are thought to underpin conscious pleasure and motivation31. Using reactivity as a primary outcome measure (see dopamine motivates you to get ready for it Nature Reviews | 31,35 figure), this research has used pharmacological and lesion techniques to determine (‘wanting’) . causal relationships between neuronal signalling and hedonic feelings. In the pain field there is µ-opioids have been shown to cause a growing recognition that the ‘subjective interpretation’ or ‘meaning’ of pain determines the positive shift in across the hedonic amount of pain-related suffering15,88. The definition of pain, according to the International spectrum: they enhance the pleasantness of Association for the Study of Pain, emphasizes the ‘unpleasant’ and ‘emotional’ aspects, and also sweet and decrease the aversiveness includes subjective feelings of pain, which are not caused by tissue damage. Other research of pain and bitter foods31. Both painful and 89 areas that are turning their to hedonic feelings include the fields of obesity research pleasant events are associated with the and decision making: the shift in focus from ‘cold’ rational consideration to ‘hot’ -based release of endogenous µ-opioids in decision making has influenced cognitive neuroscience for more than a decade90,91. Even the brain and, importantly, in the NAc19,36 economists are now looking to hedonic feelings to explain human behaviour such as the ‘warm (FIG. 1). Blocking of µ-opioid signalling glow’ that accompanies donations to charity92. Figure modified, with permission, from Ref. 31  (2003) Academic Press. with naloxone decreases the pleasantness of food rewards34 and sexual behaviour37 and reverses reward-related analgesia16,18,26. a hidden injection, its effects Opioids and hedonic feelings Interestingly, a recent conditional gene- were significantly reduced26. Although the Pain and reward are complex constructs knockout study showed a dissociation of placebo treatment might not be pleasurable that encompass motivational, hedonic and µ-opioid-mediated reward and analgesia: in itself, reduced pain represents the better learning signals30. As the motivation to seek only µ-opioid antinociception depends on of two alternative outcomes (the other being reward or avoid pain is generally correlated an intact central serotonergic system38. The unchanged pain levels), and therefore has a with the pleasantness or aversiveness of an κ-opioid system presents another example higher reward value. event (respectively), it is difficult to disen- of pleasure–analgesia dissociation: κ-opioids A related phenomenon predicted by tangle the of the hedonic reduce pain but also induce feelings of Fields’ Motivation-Decision Model is the and motivational components of pain and aversion39,40. Furthermore, the κ-opioid effect of pain on the ability to experience reward. In addition, we have only limited activity caused by tonic (sustained) pain has pleasure. By decreasing reward pleasant- ness, pain and other threatening events Box 2 | The pain–pleasure dilemma ensure that necessary action is taken to protect the individual, thus attenuating “Pleasure is the greatest incentive to evil” — Plato the normal reward-seeking behaviour. The increased neuroscientific interest in pleasure (BOX 1) perhaps reflects a greater general focus Correspondingly, decreased consumption on pleasure and positive affect () in the Western world85. Historically, however, a strong of palatable foods is considered to be a belief in and stoicism (in the case of pleasure and pain, respectively) has prevailed. Learning measure of pain suffering and is reversible to curb impulses for instant and to tolerate some pain ‘for the greater good’ is an with morphine treatment27 (FIG. 1). Similarly, important part of child development. Considering the unnecessary pain of childbirth and the sustained pain inhibits morphine reward in stress of child rearing, it is perhaps not surprising that , selflessness and stoicism are 13 ; this is likely to be due to sustained highly regarded traits in many cultures . In neuroscience, prominent researchers advocate a ‘hedonic Calvinistic’ approach to pleasure, in which the use of the reward system is activation of the κ-opioid system in the 93 28 restricted, as they believe that unregulated pleasure-seeking might lead to addiction . The (NAc) . In Calvinistic focus on moderation, or even abstinence, of pleasure has deep roots in Western culture there is extensive co-morbidity between and is powerfully connected with shame94. Whereas excessive reliance on shame and stoicism and , which often might cause unnecessary suffering86, extreme pleasure-seeking and pain avoidance () involves a reduction in the ability of chronic can have undesirable consequences such as drug addiction93,95 and obesity89. However, the pain sufferers to enjoy everyday inability to take pleasure in everyday rewards is also a form of suffering29. In fact, paradoxical and ()29. This reduction in pleasure risky human behaviours such as self-harm and skydiving have been related to a to alleviate might form part of a vicious cycle for the emotional ‘numbness’, possibly owing to a dysfunction in the opioid and/or dopamine 78,96,97 patients, in which both negative mood systems . The strong historical association between shame, and pleasure might help to and lack of pleasure result in exacerbated explain a number of paradoxical human behaviours, as well as the historical preference for formulating scientific research questions in terms of behaviour rather than pleasure and other pain, leading to more negative mood and hedonic feelings. anhedonia.

nature reviews | neuroscience volume 9 | april 2008 | 315 © 2008 Nature Publishing Group

P e r s p e c t i v e s

been shown to disrupt the positive interac- motivation. Interestingly, in patients it seems Just as the colocalization of opioid and tion between µ-opioids and mesolimbic that this normal interaction between the dopamine pathways highlights the impor- dopamine28. dopamine system and pain is disrupted. One tance of interactions between these two study showed that, compared with controls, systems, the striking overlap in regions that Dopamine, motivation and analgesia patients suffering from generalized pain are involved in pain and pleasure processing Considering the close association that exists () released less dopamine in the (FIG. 2) might explain the modulatory effects between motivation, learning and hedonic yet found the stimulus (hypertonic- feelings, it is not surprising that dopamine saline-induced deep muscle pain) more signalling has been consistently reported painful48. This result is consistent with a a to correlate with stimulus reward value41,42. normal role for dopamine in endogenous Striatal dopamine neurons also respond to antinociception51. Any analgesic effects OFC aversive events43,44 but, in contrast to the of dopamine seem to rely on a reactive firing bursts that signal pleasant events or phasic dopamine system, and this might be VTA their cues, aversive stimulation causes a disrupted in chronic pain conditions — per- NAc 45,46 VP brief inhibition of baseline firing . The haps through increased tonic dopamine Amy many time-courses of the dopamine signal levels that inhibit phasic release48,51 (BOX 3). are often measured by different techniques, In line with this evidence, and based on b Pain making the literature on the precise role of interactions between the descending pain + Pleasure dopamine in pain and reward complicated system and the mesostriatal dopamine and somewhat inconsistent43,45 (BOX 3). circuit for drug and food reward, the Pain For instance, on the one hand, positron Motivation-Decision Model proposes + Pleasure emission tomography (PET) studies of base- that phasic dopamine has a key role in + µ-opioid Pain line dopamine receptor availability provide endogenous analgesia in situations in which receptor antagonists a measure of tonic dopamine levels43,47. On reward is expected4. Evidence from human c the other hand, PET studies comparing studies perhaps supports this concept, as low + Pain Pleasure receptor availability between two stimulus tonic dopamine levels, present in individu- conditions measure dopamine signalling als with the catechol‑O-methyltransferase Pleasure at the temporal mid-range between brief (COMT) Val/Val polymorphism, produce + Pain 52 Pleasure phasic activation and constant tonic high phasic dopamine and concomitantly + µ-opioid firing24,43,45,48. Despite the complex effects high endogenous-opioid release during receptor agonists and interactions of the various dopamine tonic pain53. Val/Val subjects also reported Figure 1 | Schematic illustration of pain– time-courses, it is clear that endogenous significantly lower pain compared with Met/ pleasure inhibition.Natur Thee Re Motivation-Decisionviews | Neuroscience 4,14 dopamine is involved in the process- Met subjects with higher tonic dopamine Model of pain posits that anything of poten- ing of both pain and pleasure3,30,41,43–46,48. levels. A recent molecular-imaging study tially greater importance than pain should have antinociceptive effects (be it a greater threat or Pharmacological manipulation of dopamine investigated the link between reward the possibility of a reward). By the same evolution- levels has also been shown to modulate both expectancy, dopamine and analgesia more ary-psychology rationale, it is clear that anything 20,30,45,49,50 pain and reward behaviours . directly, and showed that inter-individual that is potentially more important than a reward The precise role of dopamine in pain variation in NAc dopamine release during (such as an even greater reward or a threat for and reward processing is hotly debated. In a placebo manipulation correlated with which action is needed) should similarly decrease the reward literature, one main question subsequent variability in placebo analgesia3. its pleasantness, thus allowing for the appropriate has been whether the dopamine signal is Furthermore, NAc activation during antici- avoidance or approach behaviours. The µ-opioid necessary for reward learning, salience, pation of a monetary reward accounted and mesolimbic dopamine systems are the prime motivation or hedonics30,35,45. For pain, for 28% of the variance in the formation of candidates for systems that transmit signals relat- dopamine agonists, such as amphetamine, placebo analgesia in the same individuals. ing to motivational and hedonic aspects of both pain and pleasure and, in particular, their interac- reduce tonic pain but do not change phasic This study therefore supports a direct link 49 tions, as illustrated here. a | Both pain and pleasure pain behaviours . Similarly, tonic but not between dopamine and endogenous-opioid have been shown to elicit opioid release in the phasic pain events are thought to induce release with regards to reward and analgesia orbitofrontal cortex (OFC), the amygdala (Amy), endogenous analgesia through dopamine in humans. the nucleus accumbens (NAc) and the ventral pal- release in the NAc43. Dopamine receptor lidum (VP)2,65,68. Pleasure and reward expectation availability studies have shown that endog- Common regions for pain and pleasure are also associated with increased phasic dopa­ enous striatal dopamine release correlates Although the opioid and dopamine systems mine signalling from the ventral tegmental area positively with sensory and affective com- are closely related neuroanatomically54, they (VTA) to the NAc and VP42, which in turn causes 55 ponents of tonic pain in healthy subjects43,48. interact in complex ways. Phasic dopamine increased µ-opioid release in the NAc . Pain has Although these studies in healthy volunteers has been shown to increase opioid levels55, been associated with both increases and decreases in mesolimbic dopamine signalling, provide clear demonstrations of dopamine’s whereas tonic dopamine decreases opioid 53,56 depending on the type of measurement and pain involvement in pain processing, they can- levels . Conversely, opioids upregulate model that have been used42,43,46,48,49. b | µ-opioid not unequivocally answer the question of phasic dopamine in the striatum (by inhibit- receptor antagonists, such as naloxone, reverse directionality. The dopamine signal could ing local GABAergic interneurons in the pleasure-related analgesia16,18,19. c | µ-opioid recep- reflect a sustained increase in dopamine ventral tegmental area)57,58 and downregu- tor agonists, such as morphine, have been shown that might exacerbate pain, but it could also late slower striatal dopamine signalling, as to re-enable pleasure that has been previously reflect brief signals related to pain-avoidance measured by PET59. reduced by concomitant pain27.

316 | april 2008 | volume 9 www.nature.com/reviews/neuro © 2008 Nature Publishing Group

P e r s p e c t i v e s

Region Pleasure/reward Pain

101 102 Lateral prefrontal • Humans, fMRI, taste reward • Humans, H2O PET, hyperalgesic pain cortex • Humans, fMRI, pain103 Anterior insula • Humans, fMRI, food cravings104 • Humans, fMRI, pain105 75 106 PFC • Humans, H2O PET, chocolate reward • Humans, fMRI, placebo analgesia INS Posterior insula • Humans, fMRI, hypothetical reward107 • Humans, direct brain stimulation108 • Humans, fMRI, pain105 OFC Orbitofrontal • Humans, fMRI, pleasant touch74 • Humans, fMRI, pain74 cortex • Humans, fMRI, chocolate reward75 • Humans, fMRI, placebo analgesia106

64 110,111 Medial prefrontal • Humans, H2O PET, pleasurable music • Humans, fMRI, pain cortex • Humans, fMRI, monetary reward109 Anterior cingulate • Monkeys, electrophysiology112 • Humans, fMRI, pain113 75 83 gyrus • Humans, H2O PET, chocolate reward • Humans, opioid PET Dorsal striatum • Humans, fMRI, fruit juice114 • Humans, dopamine ligand PET, pain43 • Humans, fMRI, monetary reward115 • Humans, fMRI, pain116 Nucleus accumbens/ • Humans, fMRI and dopamine ligand PET 3 , • Humans, dopamine ligand PET43 ventral striatum monetary reward 65 • Humans, fMRI, expectation of pain44 • Rodents, hedonic hotspot, taste reactivity• Rodents, pain-induced analgesia20 • Humans, dopamine ligand PET 41 , drug reward Ventral pallidum • Rodents, taste reactivity 62,65 • Rodents, tracing, pain a ect72 • Humans, µ-opioid PET, sustained pain 2 75 106 • Humans, H2O PET, chocolate reward • Humans, fMRI, placebo analgesia 117 72 Hypothalamus • Humans, H2O PET, pleasurable music • Rodents, tracing of nociceptive pathway • Humans, direct brain stimulation118

75 119 • Humans, H2O PET, chocolate reward • Humans, fMRI, of pain 64 120 • Humans, H2O PET, pleasurable music • Humans, fMRI, pain 64 70,120 Amygdala • Humans, H2O PET, pleasurable music • Humans, fMRI, pain • Primates, reward anticipation/learning63 Hippocampus • Humans, fMRI, unexpected reward121 • Humans, fMRI, pain122 64 119 • Humans, H2O PET, pleasurable music • Humans, fMRI, anticipation of pain Cerebellum • Humans, fMRI, unexpected reward121 • Humans, fMRI, pain123 Brainstem • Rodents, taste reactivity124 • Humans, fMRI, pain123 • Rodents, conditioned place preference40 • Rodents, pain40 Figure 2 | Brains regions implicated in pain and pleasure processing. in animals, show striking overlap. The studies included as examples in this At the systems level, the major regions that have been implicated in pain figure unequivocally demonstrate the involvementNature Reofvie eachws |region Neuroscienc in bothe and reward processing by functional imaging studies and direct brain pain and pleasure processing. fMRI, functional MRI; PET, positron emission stimulation in humans, as well as by electrophysiology and tracing studies tomography.

of one over the other. Whether one or cues63. Evidence from human patient studies proportion of the signals that are generated two neural systems (at any spatial scale) also highlights the importance of the NAc by the unmyelinated primary afferent underpin aversive and appetitive process- and the pallidum for reward processing, as pathway72. These pallidal ‘pain ing in the brain60 is still subject to debate5. dysregulation or lesion of these regions is affect’ neurons seem to be located laterally Regions that are particularly well situated associated with anhedonia66,67. to the pallidal pleasure hotspot65,72. Thus, to mediate interactions between pain and In addition to their participation in it seems that two distinct subregions of the pleasure include the NAc, the pallidum pleasure processing, the amygdala, the pallidum are involved in appetitive and and the amygdala. These regions receive NAc and the pallidum have distinct but aversive processing. A similar finding has direct or indirect reward-related signals from important roles for pain. All three regions been reported for the NAc. Whereas neurons dopamine neurons in the midbrain and are have been shown to release endogenous located in the rostral part of the NAc shell thought to signal either reward-prediction µ-opioids during painful stimulation in mediate pleasure, stimulation of more caudal error (discrepancy between the expected humans2,68. The amygdala modulates pain regions of the NAc causes a negative shift and the received reward; NAc42,61 and amyg­ through direct connections with in affect73. A similar rostrocaudal ‘hedonic dala61) or hedonic reward value (pallidum62 the descending pain inhibitory system69,70. gradient’ in the ventral striatum was recently and amygdala63,64). The NAc and pallidum The amygdala and the NAc mediate both reported for economic gains and losses in each contain a ‘hedonic hotspot’ in which reward- and stress-induced analgesia4,69, and humans5. In the amygdala, adjacent neuronal µ-opioid stimulation increases the liking of these two regions show alterered endog- populations represent positive and negative rewards65. In fact, these two ~1mm³ regions enous-opioid analgesic activity in fibromy- hedonic value63. are necessary for the opioid-mediated algia patients71. Stress-induced analgesia can The close adjacency of such pain and enhancement of food palatability65. Different be blocked by intra-accumbens injection pleasure hotspots suggests that functional populations in the amygdala have of dopamine and opioid antagonists20. The interactions between them are involved in the been found to encode the negative and posi- pallidum contains a population of encepha- mechanism by which pain decreases pleasure tive hedonic value of reward and lin-containing neurons that receive a large and rewards induce analgesia. Evidence

nature reviews | neuroscience volume 9 | april 2008 | 317 © 2008 Nature Publishing Group

P e r s p e c t i v e s

Box 3 | ’s ‘Golden Mean’ and phasic dopamine signalling between competing pleasant and aversive events. As the Motivation-Decision Model To maintain homeostasis, animals must aim for Normal tonic dopamine level suggests, being able to ‘switch off’ pain in the ‘Golden Mean’ — that is, the right balance order to gain a reward could increase sur- between pleasure-seeking and pain-avoidance. The responsiveness of the phasic dopamine vival, if the pain–pleasure (or cost–benefit) system (a system which is caused by brief bursts ratio is right. Similarly, aversive cues of neuronal firing and relates to reward Normal phasic dopamine signal must be able to disrupt pleasure-seeking motivation and prediction error) is important if the potential danger outweighs the for the regulation of appetitive and aversive potential gain. behaviours. Impulsive behaviour and An important and as yet unanswered have been linked to an question concerns the effects of chronic excessively responsive phasic dopamine Increased tonic dopamine level pain on the ability to enjoy rewards29. 98 system , whereas depression, chronic pain and Anhedonia is a major symptom of depres- anhedonia have been associated with low Pain and stress syndromes sion, and several recent papers have sug- responsiveness to reward cues78,79. gested that it might be related to reductions Tonic dopamine activity refers to the level of Reduced phasic dopamine signal extrasynaptic dopamine that is present at a in dopaminergic neurotransmission that steady-state concentration in the extracellular are similar to those that are seen during space45,99. The baseline dopamine concentration abstinence of addictive drugs78,79. By is thought to enable a number of behavioural contrast, positive mood and cognitive flex- Decreased tonic dopamine level processes, many of which are affected in ibility are thought to arise from a highly Parkinson’s disease42. Importantly, tonic responsive phasic dopamine system80. The dopamine levels regulate the responsiveness of significant co-morbidity between chronic Impulsive behaviour the phasic dopamine system to salient pain and depression suggests that these environmental cues: high tonic dopamine patients might also lose-out on the poten- attenuates phasic dopamine release45,99 whereas tial analgesic effects of the rewarding every­ low tonic dopamine facilitates phasic dopamine Increased phasic dopamine signal firing98. The level of tonic dopamine in the limbic day events that they are no longer able to striatum is in turn modulated by corticostriatal savour. A lack of reward-induced analgesia 98,99 and hippocampal afferents and homeostasis . Nature Reviews | Neuroscience has been reported in ‘anhedonic’ stressed Increased tonic dopamine is known to result from prolonged stress or pain51 (see figure), a rats81. Indeed, endogenous-opioid activity mechanism that might have evolved to ensure rest and low activity levels during injury. is disrupted both during sad mood82 and in Unfortunately, the same mechanism is thought to cause increased pain sensitivity in certain pain chronic pain patients83. 48,51 syndromes through its inhibition of endogenous phasic dopamine antinociception . Abstinence For Jeremy Bentham, a ‘good life’ consisted from addictive drugs has also been associated with and increased tonic signalling. The of the presence of pleasures combined with resulting inhibition of phasic signalling is thought to underpin reduced responsiveness to pleasure the absence of pains7. As we have seen, the (anhedonia) during abstinence, and can be reversed by re-administering the addictive drug79. At the other extreme, decreased tonic dopamine, causing hyper-responsiveness of phasic dopamine, inability to feel pleasure is associated with has been related to positive symptoms in schizophrenia98. Impulsivity in schizophrenia is associated negative mood and depression. By contrast, with excessive pleasure-seeking and substance abuse100. positive affect is considered the hallmark of well-being80 and might actually improve health84. Bentham’s view might nevertheless that separate neuronal populations encode µ-opioid signalling truly reflects pleasurable be too simplistic. As stated in the beginning, aversive and appetitive processing in the and analgesic effects in the brain. Similarly, closely related to the subjective interpretation amygdala, the NAc and the pallidum supports although dopamine firing patterns differ in of a sensory stimulus is the concept of mean- the existence of two neural systems for pain response to reward prediction, uncertainty ing. Meaning allows for many alternative and pleasure at the within-region spatial scale. and aversive events, the mutual reinforce- paths to well-being85. Consideration of this A similar finding has also been reported for ment of phasic dopamine and opioid release factor might help to explain the abundance higher cortical regions: different subregions in is consistent with the idea that dopaminer- of paradoxical aversive or life-threatening the orbitofrontal cortex represent the hedonic gic motivation signalling takes place during human behaviours found across society that value of reward and punishment74–77. preparation for, or consummation of, a are considered ‘pleasurable’. Even suffering pleasurable reward. The finding that high can be rewarding if it has meaning to the suf- A common currency for hedonic experience tonic dopamine activity is associated with ferer86. Continued study of the commonalities The robust evidence for opioid and both increased pain and decreased pleasure, and differences between pain and pleasure dopamine involvement in the processing of and that tonic over-activity of the dopa­ is therefore necessary if we are to advance pain and pleasure makes these two neuro- mine system is known to reduce phasic our understanding of human suffering and transmitter systems the prime candidates dopamine and µ-opioid release, further well-being. for mediating the mutually inhibitory corroborates the idea that interactions Siri Leknes and Irene Tracey are at the Oxford Centre effects of pain and reward. Although both between µ-opioids and phasic dopamine for Functional MRI of the Brain, Department of Clinical pleasurable and painful events are often signalling mediate pleasure and analgesic , Nuffield Department of Anaesthetics, Oxford University, John Radcliffe Hospital, Oxford, accompanied by endogenous-opioid effects in the brain. These two neurotrans- OX3 9DU, UK. neurotransmission, the pleasure-enhancing mitter systems are thus likely to mediate Correspondence to I.T. and antinociceptive effects of µ-opioid the brain’s common currency, allowing for e-mail: [email protected] agonist treatment suggest that endogenous action selection based on the comparison doi:10.1038/nrn2333

318 | april 2008 | volume 9 www.nature.com/reviews/neuro © 2008 Nature Publishing Group

P e r s p e c t i v e s

1. Petrovic, P., Kalso, E., Petersson, K. M. & Ingvar, M. 28. Narita, M. et al. Direct evidence for the involvement of 54. Khachaturian, H. & Watson, S. J. Some perspectives Placebo and opioid analgesia— imaging a shared the mesolimbic κ-opioid system in the morphine- on monoamine- interaction in central neuronal network. Science 295, 1737–1740 (2002). induced rewarding effect under an inflammatory pain- . Brain Res. Bull. 9, 441–462 (1982). 2. Zubieta, J.‑K. et al. Placebo effects mediated by like state. Neuropsychopharmacology 30, 111–118 55. Roth-Deri, I. et al. Effect of experimenter-delivered endogenous opioid activity on µ-opioid receptors. (2004). and self-administered cocaine on extracellular J. Neurosci. 25, 7754–7762 (2005). 29. Marbach, J. J. & Lund, P. Depression, anhedonia and β-endorphin levels in the nucleus accumbens. 3. Scott, D. J. et al. Individual differences in reward in temporomandibular joint and other facial J. Neurochem. 84, 930–938 (2003). responding explain placebo-induced expectations and pain syndromes. Pain 11, 73–84 (1981). 56. King, M. A., Bradshaw, S., Chang, A. H., Pintar, J. E. & effects. Neuron 55, 325–336 (2007). 30. Berridge, K. The debate over dopamine’s role in Pasternak, G. W. Potentiation of opioid analgesia in 4. Fields, H. L. Understanding how opioids contribute to reward: the case for incentive salience. dopamine2 receptor knock-out mice: evidence for a reward and analgesia. Reg. Anesth. Pain Med. 32, Psychopharmacology 191, 391–431 (2007). tonically active anti-opioid system. J. Neurosci. 21, 242–246 (2007). 31. Berridge, K. C. Pleasures of the brain. Brain Cogn. 52, 7788–7792 (2001). 5. Seymour, B., Daw, N., Dayan, P., Singer, T. & Dolan, R. 106–128 (2003). 57. Johnson, S. W. & North, R. A. Opioids excite Differential encoding of losses and gains in the human 32. Pecina, S., Cagniard, B., Berridge, K. C., Aldridge, dopamine neurons by hyperpolarization of local striatum. J. Neurosci. 27, 4826–4831 (2007). J. W. & Zhuang, X. Hyperdopaminergic mutant mice interneurons. J. Neurosci. 12, 483–488 (1992). 6. Cabanac, M. Sensory pleasure. Q. Rev. Biol. 54, 1–29 have higher “wanting” but not “liking” for sweet 58. Nugent, F. S., Penick, E. C. & Kauer, J. A. Opioids (1979). rewards. J. Neurosci. 23, 9395–9402 (2003). block long-term potentiation of inhibitory . 7. Bentham, J. An introduction to the principles of 33. Hnasko, T. S., Sotak, B. N. & Palmiter, R. D. Morphine Nature 446, 1086–1090 (2007). morals and legislation (Clarendon 1907) Accessible reward in dopamine-deficient mice. Nature 438, 59. Hagelberg, N. et al. µ-Receptor agonism with online at http://www.econlib.org/library/Bentham/ 854–857 (2005). alfentanil increases striatal dopamine D2 receptor bnthPML1.html. 34. Barbano, M. F. & Cador, M. Differential regulation of binding in man. 45, 25–30 (2002). 8. Cannon, W. B. The wisdom of the body (Norton and the consummatory, motivational and anticipatory 60. Gray, J. A. & McNaughton, N. The of Co., New York, 1932). aspects of feeding behavior by dopaminergic and anxiety (Oxford Univ. Press, Oxford, 2000). 9. Kringelbach, M. L., O’Doherty, J., Rolls, E. T. & opioidergic drugs. Neuropsychopharmacology 31, 61. Seymour, B. et al. Opponent appetitive-aversive neural Andrews, C. Activation of the human orbitofrontal 1371–1381 (2006). processes underlie predictive learning of pain relief. cortex to a liquid food stimulus is correlated with its 35. Barbano, M. & Cador, M. Opioids for hedonic Nature Neurosci. 8, 1234–1240 (2005). subjective pleasantness. Cereb. Cortex 13, experience and dopamine to get ready for it. 62. Tindell, A. J., Smith, K. S., Pecina, S., Berridge, K. C. & 1064–1071 (2003). Psychopharmacology 191, 497–506 (2007). Aldridge, J. W. Ventral pallidum firing codes hedonic 10. Craig, A. D. A new view of pain as a homeostatic 36. Zubieta, J.‑K. et al. Regional mu reward: when a bad taste turns good. J. Neurophysiol. emotion. Trends Neurosci. 26, 303–307 (2003). regulation of sensory and affective dimensions of pain. 96, 2399–2409 (2006). 11. Leknes, S., Wiech, K., Brooks, J. C. W. & Tracey, I. Is Science 293, 311–315 (2001). 63. Paton, J. J., Belova, M. A., Morrison, S. E. & there more to pain relief than a reduction in pain 37. Murphy, M. R., Checkley, S. A., Seckl, J. R. & Salzman, C. D. The primate amygdala represents the intensity? A psychophysical investigation. Eur. J. Pain Lightman, S. L. Naloxone inhibits oxytocin release at positive and negative value of visual stimuli during 10, S78 (2006). orgasm in man. J. Clin. Endocrinol. Metab. 71, learning. Nature 439, 865–870 (2006). 12. Price, D. D., Harkins, S. W. & Baker, C. Sensory- 1056–1058 (1990). 64. Blood, A. J. & Zatorre, R. J. Intensely pleasurable affective relationships among different types of 38. Zhao, Z.‑Q. et al. Central serotonergic neurons are responses to music correlate with activity in brain clinical and experimental pain. Pain 28, 297–307 differentially required for opioid analgesia but not for regions implicated in reward and emotion. (1987). morphine tolerance or morphine reward. Proc. Natl Proc. Natl Acad. Sci. USA 98, 11818–11823 (2001). 13. Harper, P. No pain, no gain: pain behaviour in the Acad. Sci. USA 104, 14519–14524 (2007). 65. Smith, K. S. & Berridge, K. C. Opioid limbic circuit for armed forces. Br. J. Nurs. 15, 548–551 (2006). 39. Mucha, R. F. & Herz, A. Motivational properties of reward: interaction between hedonic hotspots of 14. Fields, H. L. in Proceedings of the 11th World kappa and mu opioid receptor agonists studied with nucleus accumbens and ventral pallidum. J. Neurosci. Congress on Pain (eds Flor, H., Kalso, E. & Dostrovsky, place and taste preference conditioning. 27, 1594–1605 (2007). J. O.) 449–459 (IASP Press, Seattle, 2006). Psychopharmacology 86, 274–280 (1985). 66. Miller, J. M. et al. Anhedonia after a selective bilateral 15. Tracey, I. & Mantyh, P. W. The cerebral signature for 40. Hirakawa, N., Tershner, S. A., Fields, H. L. & Manning, lesion of the . Am. J. 163, pain perception and its modulation. Neuron 55, B. H. Bi-directional changes in affective state elicited 786–788 (2006). 377–391 (2007). by manipulation of medullary pain-modulatory 67. Schlaepfer, T. E. et al. to 16. Dum, J. & Herz, A. Endorphinergic modulation of circuitry. Neuroscience 100, 861–871 (2000). reward circuitry alleviates anhedonia in refractory neural reward systems indicated by behavioral 41. Drevets, W. C. et al. Amphetamine-induced dopamine major depression. Neuropsychopharmacology 33, changes. Pharmacol. Biochem. Behav. 21, 259–266 release in human ventral striatum correlates with 368–377 (2007). (1984). . Biol. Psychiatry 49, 81–96 (2001). 68. Wager, T. D., Scott, D. J. & Zubieta, J. K. Placebo 17. Reboucas, E. C. et al. Effect of the blockade of 42. Schultz, W. Behavioral dopamine signals. Trends effects on human µ-opioid activity during pain. Proc.

µ1-opioid and 5HT2A-serotonergic/α1-noradrenergic Neurosci. 30, 203–210 (2007). Natl Acad. Sci. USA 104, 11056–11061 (2007). receptors on sweet‑substance‑induced analgesia. 43. Scott, D. J., Heitzeg, M. M., Koeppe, R. A., Stohler, 69. Fields, H. State-dependent opioid control of pain. Psychopharmacology (Berl.) 179, 349–55 (2005). C. S. & Zubieta, J.‑K. Variations in the human pain Nature Rev. Neurosci. 5, 565–575 (2004). 18. Forsberg, G., Wiesenfeld-Hallin, Z., Eneroth, P. & stress experience mediated by ventral and dorsal 70. Bingel, U., Lorenz, J., Schoell, E., Weiller, C. & Buchel, C. Sodersten, P. Sexual behavior induces naloxone- basal ganglia dopamine activity. J. Neurosci. 26, Mechanisms of placebo analgesia: rACC recruitment reversible in male . Neurosci. Lett. 81, 10789–10795 (2006). of a subcortical antinociceptive network. Pain 120, 151–154 (1987). 44. Jensen, J. et al. Direct activation of the ventral 8–15 (2006). 19. Szechtman, H., Hershkowitz, M. & Simantov, R. Sexual striatum in anticipation of aversive stimuli. Neuron 40, 71. Harris, R. E. et al. Decreased central µ-opioid receptor behavior decreases pain sensitivity and stimulated 1251–1257 (2003). availability in fibromyalgia. J. Neurosci. 27, endogenous opioids in male rats. Eur. J. Pharmacol. 45. Schultz, W. Multiple dopamine functions at different 10000–10006 (2007). 70, 279–285 (1981). time courses. Annu. Rev. Neurosci. 30, 259–288 72. Braz, J. M., Nassar, M. A., Wood, J. N. & Basbaum, A. I. 20. Gear, R. W., Aley, K. O. & Levine, J. D. Pain-induced (2007). Parallel “pain” pathways arise from subpopulations of analgesia mediated by mesolimbic reward circuits. 46. Ungless, M. A., Magill, P. J. & Bolam, J. P. Uniform primary afferent nociceptor. Neuron 47, 787–793 J. Neurosci. 19, 7175–7181 (1999). inhibition of dopamine neurons in the ventral (2005). 21. Villemure, C., Slotnick, B. M. & Bushnell, M. C. tegmental area by aversive stimuli. Science 303, 73. Reynolds, S. M. & Berridge, K. C. Positive and negative Effects of odors on pain perception: deciphering the 2040–2042 (2004). motivation in nucleus accumbens shell: bivalent roles of emotion and attention. Pain 106, 101–108 47. Hagelberg, N. et al. Alfentanil increases cortical rostrocaudal gradients for GABA-elicited eating, taste (2003). dopamine D2/D3 receptor binding in healthy subjects. “liking”/”disliking” reactions, place preference/avoidance, 22. Kenntner-Mabiala, R. & Pauli, P. Affective modulation Pain 109, 86–93 (2004). and . J. Neurosci. 22, 7308–7320 (2002). of brain potentials to painful and nonpainful stimuli. 48. Wood, P. B. et al. Fibromyalgia patients show an 74. Rolls, E. T. et al. Representations of pleasant and Psychophysiology 42, 559–567 (2005). abnormal dopamine response to pain. Eur. J. painful touch in the human orbitofrontal and cingulate 23. Roy., M., Peretz, I. & Rainville, P. Emotional valence Neurosci. 25, 3576–3582 (2007). cortices. Cereb. Cortex 13, 308–317 (2003). contributes to music-induced analgesia. Pain 49. Altier, N. & Stewart, J. The role of dopamine in the 75. Small, D. M., Zatorre, R. J., Dagher, A., Evans, A. C. & 134,140–147 (2008). nucleus accumbens in analgesia. Life Sci. 65, Jones-Gotman, M. Changes in brain activity related to 24. de la Fuente-Fernandez, R. et al. Expectation and 2269–2287 (1999). eating chocolate: from pleasure to aversion. Brain dopamine release: mechanism of the placebo effect in 50. Gerdelat-Mas, A. et al. Levodopa raises objective pain 124, 1720–1733 (2001). Parkinson’s . Science 293, 1164–1166 (2001). threshold in Parkinson’s disease: a RIII reflex study. 76. Tremblay, L. & Schultz, W. Relative reward preference 25. Petrovic, P. et al. Placebo in emotional processing— J. Neurol. Neurosurg. Psychiatry 78, 1140–1142 in primate orbitofrontal cortex. Nature 398, induced expectations of anxiety relief activate a (2007). 704–708 (1999). generalized modulatory network. Neuron 46, 51. Wood, P. B. Mesolimbic dopaminergic mechanisms 77. Kringelbach, M. L. The human orbitofrontal cortex: 957–969 (2005). and pain control. Pain 120, 230–234 (2006). linking reward to hedonic experience. Nature Rev. 26. Benedetti, F., Mayberg, H. S., Wager, T. D., Stohler, 52. Bilder, R. M., Volavka, J., Lachman, H. M. & Grace, Neurosci. 6, 691–702 (2005). C. S. & Zubieta, J.‑K. Neurobiological mechanisms of A. A. The catechol‑O‑methyltransferase 78. Nutt, D. et al. The other face of depression, reduced the placebo effect. J. Neurosci. 25, 10390–10402 polymorphism: relations to the tonic-phasic positive affect: the role of catecholamines in causation (2005). dopamine hypothesis and neuropsychiatric and cure. J. Psychopharmacol. 21, 461–471 (2007). 27. Stevenson, G. W., Bilsky, E. J. & Negus, S. S. Targeting phenotypes. Neuropsychopharmacology 29, 79. Dawkins, L., Powell, J. H., West, R., Powell, J. & pain-suppressed behaviors in preclinical assays of pain 1943–1961 (2004). Pickering, A. A double-blind placebo controlled and analgesia: effects of morphine on acetic acid- 53. Zubieta, J.‑K. et al. COMT val158met genotype experimental study of nicotine: I—effects on incentive suppressed feeding in C57BL/6J mice. J. Pain 7, affects µ-opioid neurotransmitter responses to a motivation. Psychopharmacology 189, 355–367 408–416 (2006). pain stressor. Science 299, 1240–1243 (2003). (2006).

nature reviews | neuroscience volume 9 | april 2008 | 319 © 2008 Nature Publishing Group

P e r s p e c t i v e s

80. Ashby, F. G., Isen, A. M. & Turken, A. U. A 98. Grace, A. A. Phasic versus tonic dopamine release and cingulate but not somatosensory cortex. Science 277, neuropsychological theory of positive affect and its the modulation of dopamine system responsivity: a 968–971 (1997). influence on . Psychol. Rev. 106, 529–550 hypothesis for the etiology of schizophrenia. 114. O’Doherty, J. et al. Dissociable roles of ventral and (1999). Neuroscience 41, 1–24 (1991). dorsal striatum in instrumental conditioning. Science 81. de Vasconcellos, A. P. S., Nieto, F. B., Fontella, F. U., 99. Floresco, S. B., West, A. R., Ash, B., Moore, H. & 304, 452–454 (2004). da Rocha, E. R. & Dalmaz, C. The nociceptive response Grace, A. A. Afferent modulation of dopamine neuron 115. Delgado, M. R., Stenger, V. A. & Fiez, J. A. Motivation- of stressed and lithium-treated rats is differently firing differentially regulates tonic and phasic dependent responses in the human caudate nucleus. modulated by different . Physiol. Behav. 88, dopamine transmission. Nature Neurosci. 6, Cereb. Cortex 14, 1022–1030 (2004). 382–388 (2006). 968–973 (2003). 116. Bingel, U., Glascher, J., Weiller, C. & Buchel, C. 82. Zubieta, J.‑K. et al. Regulation of human affective 100. Gut-Fayand, A. et al. Substance and suicidality Somatotopic representation of nociceptive information responses by anterior cingulate and limbic µ-opioid in schizophrenia: a common risk factor linked to in the putamen: an event-related fMRI study. Cereb. neurotransmission. Arch. Gen. Psychiatry 60, impulsivity. Psychiatry Res. 102, 65–72 (2001). Cortex 14, 1340–1345 (2004). 1145–1153 (2003). 101. Kringelbach, M. L., de Araujo, I. E. T. & Rolls, E. T. 117. Menon, V. & Levitin, D. J. The rewards of music 83. Willoch, F. et al. Central poststroke pain and reduced Taste-related activity in the human dorsolateral listening: response and physiological connectivity of opioid receptor binding within pain processing prefrontal cortex. Neuroimage 21, 781–788 the mesolimbic system. Neuroimage 28, 175–184 circuitries: a [11C]diprenorphine PET study. Pain 108, (2004). (2005). 213–220 (2004). 102. Lorenz, J., Minoshima, S. & Casey, K. L. Keeping pain 118. Bussone, G. et al. Deep brain stimulation in 84. Pressman, S. D. & Cohen, S. Does positive affect out of mind: the role of the dorsolateral prefrontal craniofacial pain: seven years’ experience. Neurol. Sci. influence health? Psychol. Bull. 131, 925–71 cortex in pain modulation. Brain 126, 1079–1091 28, S146–S149 (2007). (2005). (2003). 119. Fairhurst, M., Wiech, K., Dunckley, P. & Tracey, I. 85. Seligman, M. in Handbook of (eds 103. Wiech, K. et al. Anterolateral prefrontal cortex Anticipatory brainstem activity predicts neural Snyder, C. R. & Lopez, S. J.) (Oxford Univ. Press, New mediates the analgesic effect of expected and processing of pain in humans. Pain 128, 101–110 York, 2005). perceived control over pain. J. Neurosci. 26, (2007). 86. Driscoll, R. & Edwards, L. The misconception of 11501–11509 (2006). 120. Becerra, L., Breiter, H. C., Wise, R., Gonzalez, R. G. & Christian suffering. Pastoral Psychol. 32, 34–48 104. Pelchat, M. L., Johnson, A., Chan, R., Valdez, J. & Borsook, D. Reward circuitry activation by noxious (1983). Ragland, J. D. Images of desire: food-craving activation thermal stimuli. Neuron 32, 927–946 (2001). 87. Nagel, T. What is it like to be a bat? Philos. Rev. 83, during fMRI. Neuroimage 23, 1486–1493 (2004). 121. Ramnani, N., Elliott, R., Athwal, B. S. & Passingham, 435–450 (1974). 105. Brooks, J. C. W., Zambreanu, L., Godinez, A., Craig, R. E. Prediction error for free monetary reward in the 88. Price, D. D., Barrell, J. J. & Gracely, R. H. A A. D. & Tracey, I. Somatotopic organisation of the human prefrontal cortex. Neuroimage 23, 777–786 psychophysical analysis of experiential factors that human insula to painful heat studied with high (2004). selectively influence the affective dimension of pain. resolution functional imaging. Neuroimage 27, 122. Ploghaus, A. et al. Exacerbation of pain by anxiety is Pain 8, 137–149 (1980). 201–209 (2005). associated with activity in a hippocampal network. 89. Lowe, M. R. & Butryn, M. L. Hedonic hunger: a new 106. Wager, T. D. et al. Placebo-induced changes in fMRI in J. Neurosci. 21, 9896–9903 (2001). dimension of appetite? Physiol. Behav. 91, 432–439 the anticipation and experience of pain. Science 303, 123. Zambreanu, L., Wise, R. G., Brooks, J. C. W., Iannetti, (2007). 1162–1167 (2004). G. D. & Tracey, I. A role for the brainstem in central 90. Damasio, A. Descartes’ Error: emotion, reason, and 107. Wittmann, M., Leland, D. & Paulus, M. Time and sensitisation in humans. Evidence from functional the (Grosset/Putnam, New York, 1994). decision making: differential contribution of the magnetic resonance imaging. Pain 114, 397–407 91. Peters, E., Västfjäll, D., Gärling, T. & Slovic, P. Affect posterior and the striatum during a (2005). and decision making: a ‘hot’ topic. J. Behav. Decision delay discounting task. Exp. Brain Res. 179, 124. Pecina, S. & Berridge, K. C. Brainstem mediates Making 19, 79–85 (2006). 643–653 (2007). diazepam enhancement of palatability and feeding: 92. Harbaugh, W. T., Mayr, U. & Burghart, D. R. Neural 108. Ostrowsky, K. et al. Representation of pain and microinjections into fourth ventricle versus lateral responses to taxation and voluntary giving reveal somatic sensation in the human insula: a study of ventricle. Brain Res. 727, 22–30 (1996). motives for charitable donations. Science 316, responses to direct electrical cortical stimulation. 1622–1625 (2007). Cereb. Cortex 12, 376–385 (2002). Acknowledgements 93. Koob, G. F. & Moal, M. L. Drug abuse: hedonic 109. Knutson, B., Fong, G. W., Bennett, S. M., Adams, The authors wish to thank L. Moseley and M. Kringelbach for homeostatic dysregulation. Science 278, 52–58 C. M. & Hommer, D. A region of mesial prefrontal their helpful advice on the figures, the Wellcome and the (1997). cortex tracks monetarily rewarding outcomes: Medical Research Council (Functional Magnetic Resonance 94. Moss, J. in Oxford Studies in Ancient Philosophy (ed. characterization with rapid event-related fMRI. Imaging of the Brain Centre). Sedley, D.) (Oxford Univ. Press, USA, 2005). Neuroimage 18, 263–272 (2003). 95. Ballantyne, J. C. & LaForge, K. S. Opioid dependence 110. Porro, C. A., Cettolo, V., Francescato, M. P. & Baraldi, P. and addiction during opioid treatment of chronic pain. Functional activity mapping of the mesial hemispheric DATABASES Pain 129, 235–255 (2007). wall during anticipation of pain. Neuroimage 19, Entrez Gene: 96. Franken, I. H. A., Zijlstra, C. & Muris, P. Are 1738–1747 (2003). http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=gene nonpharmacological induced rewards related to 111. Ochsner, K. N. et al. Neural correlates of individual COMT anhedonia? A study among skydivers. Prog. differences in pain-related fear and anxiety. Pain 120, Neuropsychopharmacol. Biol. Psychiatry 30, 297–300 69–77 (2006). FURTHER INFORMATION (2006). 112. Shidara, M. & Richmond, B. J. Anterior cingulate: Irene Tracey’s homepage: http://www.fmrib.ox.ac.uk/pain 97. Symons, F., Thompson, A. & Rodriguez, M. Self- single neuronal signals related to degree of reward International Association for the Study of Pain: injurious behavior and the efficacy of naltrexone expectancy. Science 296, 1709–1711 (2002). http://www.iasp-pain.org treatment: a quantitative synthesis. Ment. Retard. 113. Rainville, P., Duncan, G. H., Price, D. D., Carrier, B. & All links are active in the online pdf Dev. Disabil. Res. Rev. 10, 193–200 (2004). Bushnell, M. C. Pain affect encoded in human anterior

320 | april 2008 | volume 9 www.nature.com/reviews/neuro © 2008 Nature Publishing Group