Using positron emission scanning and brain MRI, neuroanatomical correlates of hunger and satiety have been investigated. Importantly, recent analyses of this collected data has indicated that the left dorsolateral prefrontal cortex, an area of the brain important in reward processing, may be a satiety center. Neuronal activation in the left dorsolateral prefrontal cortex following a meal is consistently lower in this area in obese versus lean individuals, in both men and women. Furthermore, we have found that patients with Prader Willi Syndrome have deactivation in this region compared to matched control subjects. To investigate the effect of stimulation of the left dorsolateral prefrontal cortex on food intake, a randomized study using trans-cranial direct current stimulation (TDCS) is ongoing. Obese volunteers are randomized to TDCS versus sham therapy. Volunteers will receive treatment 3 days in a row as inpatients on the clinical research unit, while eating ad-libitum from computerized vending machines. Volunteers will continue to receive TDCS or sham for an additional 4 weeks to investigate the effects of this treatment on weight loss. The placement of the electrodes for the TDCS has been optimized during the ongoing study. As the previous placement may not have delivered the needed stimulation, we invited study participants who had completed the study with the initial electrode placement back to participate in the study again with the new electrode placement. In those who returned, we found that those who received the active stimulation consumed fewer overall calories overall and fewer calories from fat; during the inpatient portion of the study these individuals also lost a greater percentage of body weight. We are going confirm these results in our ongoing randomized study, and recruitment is currently ongoing for this protocol. To understand how TDCS activates or deactivates brain regions, we will be including functional MRIs prior to and following the first TDCS session. Previous studies investigating the association of gray matter density with adiposity have not differentiated between fat mass (FM, adipose tissue only) and fat free mass (FFM), and both increase with increasing adiposity. We found that fat free mass indexed to height (FFMI) was associated with reduced gray matter volume (GMV) in the bilateral temporal medial and inferior gyri, the bilateral ventromedial prefrontal cortex extending to the anterior cingulate, and the bilateral orbitofrontal cortex with extension to the insula on the left. Similar overlapping associations were seen with fat mass indexed to height (FMI). Percent body fat was associated only with reduced GMV in left temporal lobe and left cerebellum. Most importantly, in models adjusting for both FFM and percent body fat or FFM and FM, only FFM remained associated with the above brain regions. These regions are part of important brain networks which monitor reward related behavior and are involved in homeostatic regulation. We have followed up this analysis by investigating FFM as a determinant of cerebral blood flow (rCBF) in individuals who had whole brain PET scans. We have demonstrated that FFM is associated with rCBF in specific mid-brain regions which connect the hypothalamus and higher brain regions. Moreover there is substantial overlap between these associations between FFM and hunger scores in this study. Further analyses indicated that the rCBF in these midbrain regions mediated the affect of FFM on hunger scores. Our results indicate that differences in brain regions with increased adiposity are due to the associated increases in fat free rather than fat mass, and that there are regions that mediate the associated between FFM and hunger. In a separate histologic study we also investigated whether neuronal and astrocyte density differed in striatal regions of lean versus obese individuals. After staining and using stereology to count the number of neurons and astrocytes, we did not find any difference in neuronal or astrocyte number between lean and obese individuals. Obese individuals did have greater variance in neuronal number however. Thus, differences in gray matter volume as assessed by analyses of brain MRI may not be due to cellular density in these areas.
