Before we summarize the findings, we need to acknowledge some general limitations. First, there are very few neuroimaging studies with patients with compulsive shopping or sexual behavior, so the evidence primarily builds on pathological gambling and to a lesser degree on Internet addiction and binge eating. Furthermore, there is a great lack of longitudinal studies of behavioral addictions. As a consequence, we do not know if the findings are triggers or consequences.
In summary, the data on behavioral addiction show a pattern similar to the neurobiology in SUDs. The findings indicate a lower dopamine receptor binding in the striatum [35
], reflecting either a reduced receptor density or a heightened dopamine level. The blunted response to predicted rewards might be a sign of reduced sensitivity to “normal” rewards, or might stem from an increased baseline activity [53
•]. Heightened activation of the mesocorticolimbic system following addiction-related cues [40
] speaks for a dopaminergic hypersensitivity. Reduced loss sensitivity and slower loss learning rates [55
] indicate a lack of a tonic dopamine level dip that usually appears during punishment. Additionally, subjects with behavioral addictions show impairments in inhibition and reversal learning tasks correlating with reduced activity in the ventrolateral and dorsolateral PFC [58
]. Altered reward sensitivity as well as impaired top-down control also correlate with heightened risk taking and delay discounting [68
All in all, the results mainly point to a pattern of heightened salience attribution and impaired inhibition as proposed by the I-RISA model of SUDs [28
]. The question why some people develop ICDs and some do not still remains open. Prevailing evidence suggests a specific dopaminergic at-risk endophenotype (see Fig. ): considering models of phasic and tonic dopamine functioning in the striatum and the PFC [92
], one could hypothesize that individual predisposition implicates heightened tonic dopamine levels in the striatum [33
]. Tonic dopamine predominantly activates D2
receptors, whereas phasic dopamine stimulates D1
]. Augmented tonic dopamine levels would explain prefrontal deficits in behavioral addictions, since an increasing tonic D2
stimulation has been shown to attenuate PFC inputs and was correlated with reduced PFC activity [5
]. Punishment, however, would not lead to sufficient reduction of tonic dopamine levels and hence hinder punishment learning. Suprathreshold phasic bursts following particularly strong reinforcers would thus promote habit formation. The results of studies in the PD population support the importance of an increased tonic dopamine level, since dopamine agonists primarily raise the tonic dopamine level.
Of course, this model is a gross simplification, not only with regard to dopaminergic transmission, but also because it disregards contributions of other neurotransmitters. Still, this model of a dopaminergic at-risk endophenotype is based on empirical neurobiological evidence and may inform future research and development of therapeutic strategies.