Dynamic fMRI networks predict success in a behavioral weight loss program among older adults

Neuroimage, 2018.

The connectivity principal components revealed by HOSVD. The left side of each component is network-based representation and the right side shows the individual nodes in anatomical location. The connections between two regions do not represent correlation between the fMRI time series of those nodes. Rather, each network component is a collection of nodes and edges that capture the greatest amount of variance across time and within individuals. The size of each node is directly related with its number of connections. The normalized variance explained by each component is noted on the top of the same component’s graph. For each component, the variance was computed as the square of the corresponding singular value normalized by the first singular value, and finally averaged across 100 validation folds.

Abstract. More than one-third of adults in the United States are obese, with a higher prevalence among older adults. Obesity among older adults is a major cause of physical dysfunction, hypertension, diabetes, and coronary heart diseases. Many people who engage in lifestyle weight loss interventions fail to reach targeted goals for weight loss and most will regain what was lost within 1-2 years following cessation of treatment. There is a need to identify biomarkers that are predictive of weight loss success in the hope that this would lead to more individually tailored treatments, an idea that is consistent with the concept of precision-based medicine. Although biochemical and metabolic markers hold great promise, neurobiological measures may prove useful as substantial behavioral change is necessary to achieve success in a lifestyle intervention. In the present study, we proposed to use dynamic brain networks from functional magnetic resonance imaging (fMRI) data to prospectively identify those individuals most likely to succeed in a behavioral weight loss intervention. Brain imaging was performed in overweight or obese older adults (age: 65-79 years) who participated in an 18-month lifestyle weight loss intervention. Machine learning and functional brain networks were combined to produce multivariate prediction models. The average prediction accuracy of random subsampling cross validation with 100 permutations of the participants was above 95%, suggesting that there exists a consistent pattern of connectivity which correctly predicts success with weight loss at the individual level. Connectivity patterns that contributed to accurate predictions consisted of complex multivariate circuits that substantially overlapped with known brain networks; e.g. default mode network, motor circuitry, cerebellum, insular cortex, dorsal striatum and anterior cingulate cortex which are associated with behavior emergence, self-regulation, body awareness, and the sensory features of food. Future work on independent datasets will be an important step toward corroborating our findings and the development of innovative clinical tools that target the complex nature of intentional weight loss.