Global integration of the hot-state brain network of appetite predicts short term weight loss in older adult.

Obesity is a public health crisis in North America. While lifestyle interventions for weight loss (WL) remain popular, the rate of success is highly variable. Clearly, self-regulation of eating behavior is a challenge and patterns of activity across the brain may be an important determinant of success. The current study prospectively examined whether integration across the Hot-State Brain Network of Appetite (HBN-A) predicts WL after 6-months of treatment in older adults. Our metric for network integration was global efficiency (GE). The present work is a sub-study (n = 56) of an ongoing randomized clinical trial involving WL. Imaging involved a baseline food-cue visualization functional MRI (fMRI) scan following an overnight fast. Using graph theory to build functional brain networks, we demonstrated that regions of the HBN-A (insula, anterior cingulate cortex (ACC), superior temporal pole (STP), amygdala and the parahippocampal gyrus) were highly integrated as evidenced by the results of a principal component analysis (PCA). After accounting for known correlates of WL (baseline weight, age, sex, and self-regulatory efficacy) and treatment condition, which together contributed 36.9% of the variance in WL, greater GE in the HBN-A was associated with an additional 19% of the variance. The ACC of the HBN-A was the primary driver of this effect, accounting for 14.5% of the variance in WL when entered in a stepwise regression following the covariates, p = 0.0001. The HBN-A is comprised of limbic regions important in the processing of emotions and visceral sensations and the ACC is key for translating such processing into behavioral consequences. The improved integration of these regions may enhance awareness of body and emotional states leading to more successful self-regulation and to greater WL. This is the first study among older adults to prospectively demonstrate that, following an overnight fast, GE of the HBN-A during a food visualization task is predictive of WL.