COMMENTS: It’s clear from this study that pathological gambling mirrors the neurobiology of substance addictions. They found diminished reward circuitry in wins and losses, unlike normal controls. Another finding is that important personal relevant stimuli did not activate the reward circuitry. This too is found in substance addictions. The new DSM will classify pathological gambling as an addiction.
Hum Brain Mapp. 2010 Nov;31(11):1802-12.
de Greck M, Enzi B, Prösch U, Gantman A, Tempelmann C, Northoff G.
Department of Psychiatry at Otto-von-Guericke University Magdeburg, Leipziger Straße 44, 39120 Magdeburg, Germany. [email protected]
Pathological gamblers impress by an increasing preoccupation with gambling, which leads to the neglect of stimuli, interests, and behaviors that were once of high personal relevance. Neurobiologically dysfunctions in reward circuitry underlay pathological gambling. To explore the association of both findings, we investigated 16 unmedicated pathological gamblers using an fMRI paradigm that included two different tasks: the evaluation of personal relevance and a reward task that served as a functional localizer. Pathological gamblers revealed diminished deactivation during monetary loss events in some of our core reward regions, the left nucleus accumbens and the left putamen. Moreover, while pathological gamblers viewed stimuli of high personal relevance, we found decreased neuronal activity in all of our core reward regions, including the bilateral nucleus accumbens and the left ventral putamen cortex as compared to healthy controls. We demonstrated for the first time altered neuronal activity in reward circuitry during personal relevance in pathological gamblers. Our findings may provide new insights into the neurobiological basis of pathological gamblers’ preoccupation by gambling.
‘‘You’ve become insensible,’’ he remarked. ‘‘You have not only renounced life, your own interests and those of your society, your duty as a man and a citizen, your friends (and you did have them all the same)—you’ve not only renounced every aim whatever in life, except winning at roulette—you have even renounced your memories.’’
Dostoyevsky, The Gambler, 1867
The Russian novelist Dostoyevsky describes two of the core symptoms of pathological gambling, which presentday psychiatrists would characterize as craving for gambling and increasing neglect of formerly self-relevant interests. Current diagnostic manuals [DSM-IV, American Psychiatric Association, 1994; ICD-10, World Health Organization, 1992] classify pathological gambling as an impulse-control disorder. However, resemblances to addictive disorders, such as alcoholism and cocaine addiction, allow a new perspective to be taken. Pathological gambling may be viewed as a nonsubstance related addictive disorder [Reuter et al., 2005].
The classification of pathological gambling as a nonsubstance- related addictive disorder suggests abnormalities in reward circuitry such as those in substance addiction. Such abnormalities have been found in the nucleus accumbens (NACC)/ventral striatum (VS), putamen, ventromedial prefrontal cortex (VMPFC), orbitofrontal cortex (OFC), ventral tegmental area (VTA) [for an overview see Knutson and Gibbs, 2007; McClure et al., 2004; O’Doherty, 2004; for the association of addictive disorders and reward circuitry see Martin-Soelch et al., 2001; Volkow et al., 2004, 2007a]. Reuter et al.  investigated the neuronal activity of pathological gamblers using a card-guessing task and fMRI. During the receipt of monetary reward, they found altered neuronal activity in the reward circuitry of pathological gamblers including the right VS and VMPF when compared to healthy control subjects. Moreover, the authors found a diminished difference in neuronal activity between monetary gains and losses in these subjects.
Potenza et al. , who investigated pathological gamblers performing a Stroop task, also found reduced VMPFC activity. However in a different study, the same region showed increased activity in pathological gamblers during a black jack task with monetary reward when compared to the same task without it [Hollander et al., 2005]. During the presentation of gambling scenes, decreased activity of other regions such as the OFC, the thalamus, and the basal ganglia, was also observed [Potenza et al., 2003]. These results may be supplemented with findings from substance-related addictive diseases such as alcoholism and cocaine addiction. Much like pathological gamblers, alcoholic patients showed decreased neuronal activity in VS during monetary gains [Wrase et al. 2007] and reduced striatal dopamine activity during the intake of methylphenidate as measured with PET using [11C]-raclopride [Volkow et al., 2007b]. Cocaine addicted patients showed decreased neuronal activity during monetary rewards in OFC, lateral prefrontal cortex, and mesencephalon among others [Goldstein et al., 2007]. Finally, Tanabe et al.  demonstrated altered neuronal activity during decision making in the ventromedial prefrontal cortex and other regions, showing the similarity of pathological gambling to other addictive disorders.
Taken together, these findings demonstrate the crucial importance of reward circuitry in pathological gambling as well as its resemblance to other addictive disorders. According to Reuter et al. , such reduced responsiveness to reward may symptomatically lead to the chronic impression of discontentment. This in turn may increase the risk to seek satisfaction by stronger reinforcers such as gambling, cocaine, or other drugs of abuse to obtain sufficient activation level in reward regions.
Another striking symptom of pathological gambling is a pronounced shift in personal relevance. Patients are increasingly preoccupied by gambling and so begin to neglect other formerly self-relevant stimuli and behaviors. Psychologically, the evaluation of personal relevance or self-relatedness, as previous studies have called it [de Greck et al., 2008, 2009; Kelley et al., 2002; Northoff and Bermpohl, 2004; Northoff et al., 2006; Phan et al., 2004], describes how important and how close to themselves subjects experience specific stimuli. Neurobiologically, tasks that engage the notion of self-relatedness, and hence personal relevance have implicated regions from reward circuitry such as the NACC, VTA, and the VMPFC [de Greck et al., 2008; Northoff et al., 2006; Northoff et al., 2007; Phan et al., 2004].
The recruitment of reward circuitry by stimuli of high personal relevance raises the question of the exact relationship between processing of reward and processing of personal relevance stimuli. In a preliminary study by our group, tasks of high personal relevance induced neuronal activity in exactly the regions implicated in reward function in healthy subjects [de Greck et al., 2008]. Recently, our group also found that alcoholic patients showed decreased neuronal activity in reward circuitry (i.e., left and right NACC/VS, VTA, VMPFC) during the evaluation of stimuli with a high personal relevance as compared to healthy controls [de Greck et al., 2009] showing that the apparent changes in behavior stem from a lack of activation in reward circuitry during the evaluation of stimuli of high personal relevance.
The general aim of our study was to explore the neural basis of the abnormal shift of perceived personal relevance in reward circuitry in unmedicated pathological gamblers. Specifically, we employed a paradigm to investigate the neuronal activity in reward circuitry of pathological gamblers during both a reward task consisting of monetary wins and losses, and during a task requiring the evaluation of self-relatedness, in which subjects rated different pictures containing gambling scenes, food or alcohol, as of high or low personal relevance. Our hypothesis were twofold. First, we expected to replicate the findings of Reuter et al.  by demonstrating that pathological gamblers exhibit decreased neuronal activity in reward regions during the reward task. Moreover, we expected to extend these findings by differentiating between gains and losses. We predict less activation during monetary gains and less deactivation during monetary losses. Second, based on the clinical symptoms and our own findings in alcoholism [de Greck et al., 2009], we hypothesized disturbed activity in reward circuitry during the evaluation specifically of high personal relevance in pathological gamblers when compared to healthy controls.
We investigated reward circuitry during the evaluation of personal relevance in pathological gamblers. Replicating the findings of Reuter et al. , pathological gamblers showed reduced neuronal activity in bilateral NACC and left ventral putamen during a reward task. Extending these findings, we demonstrated that pathological gamblers showed reduced signal changes in the same reward regions during the evaluation of personal relevance when compared to healthy subjects. Taken together, we, for the first time, demonstrate neuronal abnormalities in reward circuitry of pathological gamblers during evaluation of personal relevance.
Alterations of Reward Circuitry in Pathological Gamblers During Monetary Wins and Losses
Our data are in accordance with the findings of Reuter et al.  who found a decreased difference in neuronal activity during monetary wins and losses. In addition to this we were able to extend their results in two ways. First, we demonstrated that the diminished difference of neuronal activity between wins and losses stems from weaker deactivation in the left NACC and the left ventral putamen during lose-events rather than from smaller activation during win-events.
Alterations in Reward Circuitry of Pathological Gamblers During the Evaluation of Personal Relevance
The striking findings of our study concerns the alteration of brain activity during the evaluation of personal relevance in pathological gamblers. As expected we found a significant lack of neuronal activity in our three reward regions (left and right NACC, left putamen) during the evaluation of stimuli with high personal relevance. These findings are in line with our hypothesis and implicate diminished neuronal reactivity in reward circuitry of gambling- addicted patients during tasks of specifically high personal relevance. Our present findings complement previous ones from our group in which alcoholic patients also exhibited reduced neuronal activity in reward circuitry while viewing stimuli of high personal relevance [de Greck et al., 2009]. Also as in alcoholic patients, this reduced neuronal activity during self-relatedness in pathological gamblers is well in accordance with the clinical observation of a severe shift of personal relevance from formerly personally important habits to gambling as the only personally relevant activity. This assumption is supported by our behavioral finding that pathological gamblers classified gambling stimuli significantly more often as highly self-related when compared to healthy subjects.
Most importantly, our findings demonstrate for the first time that these clinical and behavioral alterations in the perception of personal relevance may correspond to disturbed neuronal activity in reward circuitry on a neurobiological level. Moreover, stimuli classified as highly personally relevant eventually fail to induce neuronal activity in reward circuitry. Therefore, in line with previous postulations [Reuter et al., 2005], one might hypothesize that due to the apparent inability to stimulate their reward circuitry by even highly self-related stimuli, these patients might be forced to seek situations that provide stronger reinforcement such as gambling or drugs to create sufficient baseline activity in their reward circuitry.
Finally, we must consider the methodological limitations of our study. First and foremost, the concept of personal relevance or self-relatedness may seem problematically vague empirically and/or conceptually. We used the concept from previous studies on personal relevance and self relatedness [de Greck et al., 2008, 2009; Northoff and Bermpohl, 2004; Northoff et al., 2006, 2007] that allowed subjects to explicitly indicate whether a presented stimulus was of high or low personal relevance. Although this concept of personal relevance is a rather broad approach, we nevertheless decided to implement it in our paradigm.
In this study, we have demonstrated the important underlying role of reward circuitry in pathological gambling. Pathological gamblers not only show diminished neuronal activity in reward circuitry (left and right NACC, left ventral putamen) during monetary wins and losses, but also—and more considerably—during the evaluation of stimuli with high personal relevance. While healthy subjects show high activity in reward circuitry during the evaluation of highly personally relevant stimuli, pathological gamblers lack this increase in neuronal activity. These findings may, in time, be found to correspond to the clinical observation of an increasing neglect of other (formerly relevant) activities and the total preoccupation with gambling.