About 50% of pathological gamblers who try to quit experience a relapse with seriously negative consequences (Hodgins & el Guebaly 2004
), and other studies indicate frequent relapses in treatment-seeking pathological gamblers (Ledgerwood & Petry 2006
). Because cue reactivity is a key mechanism in the development of addictive disorders, and because it has been associated with a higher risk of relapse in substance dependence (Cooney et al. 1997
; Kosten et al. 2006
; Marissen et al. 2006
), investigating the neurobiological mechanisms of cue reactivity in this population is highly relevant. So far, only two functional magnetic resonance imaging (fMRI) studies on exposure to gambling-related cues in pathological gamblers have been published (Potenza et al. 2003
; Crockford et al. 2005
). Both studies employed video fragments of gambling-related and various control scenes, but yielded inconsistent results. In the first study among 10 pathological gamblers and 11 normal controls, PG subjects revealed decreased, rather than increased activation in the ventral anterior cingulate cortex, orbitofrontal cortex, basal ganglia and thalamus during gambling-associated versus control epochs. Increased activation during viewing of gambling-related material was found in the occipital lobe only (Potenza et al. 2003
). In the second study in 10 pathological gamblers and 10 healthy controls (HC) (Crockford et al. 2005
), PG subjects showed higher brain activation in response to gambling stimuli in the left occipital cortex, left fusiform gyrus, right parahippocampal gyrus and right prefrontal areas, compared with HC.
Thus, whereas these PG studies indicate increased activation of brain regions involved in attention, memory and visual processing, no evidence for abnormally increased activity in limbic structures during processing of gambling cues was found (e.g. increased activation in amygdala), unlike neuroimaging studies on cue reactivity in substance dependence (Kilts et al. 2004
; Tapert et al. 2004
; Kosten et al. 2006
; McBride et al. 2006
; Franklin et al. 2007
). Possible reasons for this discrepancy are the use of videos instead of pictures and lack of power because of small sample sizes. Furthermore, both studies enrolled gamblers recruited through advertisements, and neither study investigated whether treatment-seeking problem gamblers (PRGs) would differ in cue reactivity to gambling cues from normal controls. In an fMRI study focusing on the processing of rewards in pathological gamblers (Reuter et al. 2005
), a blunted response to wins versus losses was found in the limbic reward areas in pathological gamblers versus HC. When presenting pathological gamblers with gambling videos, the limbic system may thus be relatively underactivated because of a diminished response to gambling situations in which money is gained. Given this blunted response to monetary gains, the investigation of limbic activation to gambling cues versus neutral cues that do not include monetary gain may provide insight in cue reactivity to general gambling cues.
In the present study, we wanted to address these issues by investigating brain activation patterns to gambling or smoking cues in chronic PRGs seeking treatment, heavy smokers (HSM) and non-smoking non-gambling healthy controls (HC). We employed an event-related picture paradigm (George et al. 2001
; Myrick et al. 2004
; Smolka et al. 2006
) because this provides optimal flexibility with regard to stimulus timing and avoids modeling problems which may arise when analyzing video paradigm fMRI data. In order to compare cue reactivity in PRG to cue reactivity of a substance-dependent group, a comparison group of HSM was included as well. An HSM control group was chosen because the neurotoxic effects of nicotine are limited compared with those of other drugs of abuse, such as alcohol (Sullivan 2003
; Mudo, Belluardo & Fuxe 2007
). Based on previous cue-reactivity studies in substance dependence, we hypothesized that gambling cues in PRG and smoking cues in HSM would elicit higher brain reactivity compared with brain reactivity in healthy non-smoking controls in brain regions associated with emotion processing and motivational behavior such as the amygdala, ventral striatum and ventral prefrontal cortex, and in attention and cognitive control-related brain areas such as the dorsal prefrontal cortex and anterior cingulate cortex (ACC). In addition, the relation between cue-related brain activity and subjective craving in PRG and HSM was studied. We hypothesized that subjective craving would be associated with increased activation in emotion and motivation-related brain areas in PRG and HSM.