4. Discussion
In this study, we found that compared with controls, individuals with PWS had significantly lower rCBF in the right thalamus, left insular cortex, bilateral lingual gyrus, and bilateral cerebellum. They had significantly higher rCBF in the right inferior frontal gyrus, left middle/inferior frontal gyrus (anterior and posterior clusters), and bilateral angular gyrus. Additionally, rCBF in the left insula, which was significantly lower in PWS individuals, was negatively correlated with the eating behavior severity score.
The insula is known to be a critical neural substrate for conscious interoception, which is the neural mapping of bodily states that have special relevance to the maintenance of homeostasis [24]. Because the awareness of physical sensation is an important component of emotion, the representation of visceral sensation can guide motivational behavior. For example, the perception of gastric distention is an important contributor to meal termination and short-term satiety; it is one of the mechanisms that modulate feeding behavior. An fMRI study of non-painful gastric distention showed that a subjective sense of fullness was associated with activation in the amygdala and insula in the left hemisphere [25]. Moreover, the insula may also mediate the memory of the rewarding effects of food. A study by Naqvi et al. [26] provides strong evidence that interoceptive processes within the insula underlie the craving associated with nicotine addiction; they found that insular damage, either to the left or right side, is related to conscious urges to smoke. Although there is no direct evidence that alterations in eating behavior are derived from insular dysfunction, during food craving, fMRI recordings showed activation of the left insula, right caudate nucleus, and left hippocampus [27]. The insula is also regarded as a primary taste cortex and integrates the gustatory, olfactory, and visceral sensory inputs required for taste representation and the perception of flavor [28]. Cereda et al. [29] reported that a patient with a selective left insular infarction showed a taste recognition deficit. The patient, who had normal responses to bitter and sweet, did not recognize saline and acidic solutions. Other lesion studies have shown that patients with either a left or right insular lesion have significantly decreased rating scores for taste intensity perception [30] and [31]. In line with these lesion studies, some imaging studies have demonstrated that the right or bilateral insula is activated when healthy subjects view pictures of disgusting foods or are exposed to disgusting tastes [32] and [33]. These studies indicate that the insula distinguishes tastants that differ in hedonic value. With regard to food-related behavior associated with PWS, recent neuroimaging studies have reported that individuals with PWS show delayed signal reduction [8] and significantly different responses to food pictures [12] in the bilateral insula compared with healthy controls. These data support the findings in the present study that insular hypoperfusion may be associated with abnormal eating behavior in individuals with PWS, although the functional roles of the left and right insula should be studied further.
We also found significantly lower rCBF in the right thalamus in individuals with PWS compared with controls, although there was no significant correlation between rCBF and the severity of abnormal eating behavior. The thalamus is closely connected with the insula, and its dysfunction may cause abnormal eating behavior. Cerrato et al. [34] reported a patient with compulsive hyperphagia after a medial thalamic ischemic stroke. The authors proposed that thalamo-cortical dysfunction in the patient may have been the cause of the symptom.
Additionally, we found that PWS individuals had significantly higher rCBF in the right inferior frontal gyrus, left middle/inferior frontal gyrus, and bilateral angular gyrus and significantly lower rCBF in the bilateral lingual gyrus and bilateral vermis cerebelli. These findings are consistent with previous fMRI results showing that PWS individuals with the deletion subtype exhibited greater activation in the insula, middle/inferior frontal gyrus, and some regions in the frontal, temporal and parietal lobes in response to food images after a meal compared with healthy controls (there was no description of the laterality of these areas) [13]. As previously mentioned, a PET study reported that cerebral glucose metabolism at rest increased in the bilateral middle frontal gyrus, right inferior frontal gyrus and some regions in the frontal and temporal lobes and decreased in the left cerebellar vermis and the right superior temporal gyrus in PWS subjects [16]. Cornier et al. [35] used fMRI to identify the effects of short-term overeating on the neuronal responses to food-related visual stimuli. They showed that overeating resulted in diminished activation in cortical re