Neuroimaging of cigarette cravings: individual differences matter: commentary.

Abstract

Due to the role of cigarette cravings in smoking relapse, a fair body of work has examined the neural bases underlying the appetitive response to cigarette cues. Unfortunately, a majority of studies performed in the field have only triggered moderate craving levels in their smokers that might not have been potent enough to capture the over-riding effects of cravings on behavior. Arguably, one of the main reasons for this failure is that too many studies were performed in smokers who were allowed to smoke ad libitum before being scanned, instead of being in a state of withdrawal. In our opinion, the intensity of cravings is a critical factor when studying the neural bases of tobacco cravings. What is more challenging about Sayette & Wilson's 1 reasoning is the assertion that moderate craving levels may trigger a brain response that has little to do with that associated with uncontrollable cravings. This stimulating idea is based on the results of Engelmann et al.'s 2 meta-analysis of the neural correlates of tobacco cravings and an updated meta-analysis performed by the authors. Consistent with the authors' idea, the meta-analysis from Engelmann et al. 2 found that studies performed in deprived smokers (who typically reported more intense cravings) showed greater activation in the right superior frontal and the left lingual gyrus than studies performed in non-deprived smokers. However, the Engelman et al. meta-analysis also showed a sizeable overlap between both sets of studies, which produced activations in 44 widespread clusters 2. These results do not suggest strongly that moderate and intense cravings elicit largely different brain responses. However, it must be acknowledged that this meta-analysis included only 11 studies. The updated meta-analysis of Sayette & Wilson 1, which included 24 studies, showed that studies of deprived smokers elicited stronger activations in the (rostral) anterior cingulate cortex (ACC) than studies of non-deprived smokers. However, the paper does not mention if there were common brain regions activated in both sets of studies. Without this crucial information, it is difficult to judge if moderate and intense cravings elicit unrelated brain networks. Their meta-analysis also did not examine the relationship between brain activations and subjective cravings. It therefore remains unclear whether the increased ACC activations reflect specific differences in cravings or differences in other withdrawal-related symptoms. A review of the studies included in the meta-analysis from Sayette & Wilson 1 reveals that only four of 11 studies that assessed the relationship showed significant associations between subjective cravings and ACC activations 3-6. Although these results do not rule out a direct relationship between ACC activations and cravings, they do not offer a strong support for this assumption. Recently, Tang et al. 7 performed a meta-analysis of 13 functional magnetic resonance imaging (fMRI) studies on tobacco cravings that aggregated analyses of correlations between cravings and activations in these studies. The results showed that cravings were correlated positively with activations in the left middle frontal gyrus, the bilateral anterior insula, the right inferior parietal gyrus and the right para-hippocampal gyrus, but not in the ACC. The search for craving-specific brain responses has been compromised significantly by the lack of attention in the literature to the difference between appetitive cigarette cues and pleasant cues unrelated to smoking. Instead, studies have focused almost exclusively upon the craving minus neutral contrast. Overall, the idea that uncontrollable and moderate cravings may trigger fundamentally different brain responses is an inspiring one. We are simply unsure that it is supported strongly by the evidence at the moment. As acknowledged by the authors, factors other than deprivation may affect urge intensities, such as the methods used to trigger cravings. We would also point out that cigarette cravings elicit brain responses that vary considerably from one smoker to another, and that the sources of this heterogeneity remain unknown. Although clinical evidence shows clearly that psychological and demographic variables influence tobacco cravings 8-10, brain imaging studies have largely ignored the importance of these factors 11, for reasons that are difficult to fathom. None.