Impact of Intensive Lifestyle Intervention on Neural Food Cue Reactivity: Action for Health in Diabetes Brain Ancillary Study

Abstract

Objective: The Action for Health in Diabetes (Look AHEAD) research study was a randomized trial comparing the effects of an intensive lifestyle intervention (ILI) versus a diabetes support and education (DSE) control group in adults with type 2 diabetes and overweight or obesity. Functional magnetic resonance imaging was used to determine whether neural food cue reactivity differed for these groups 10 years after randomization.

Methods: A total of 232 participants (ILI, n = 125, 72% female; DSE, n = 107, 64% female) were recruited at three of the Look AHEAD sites for functional magnetic resonance imaging. Neural response to high-calorie foods compared with nonfoods was assessed in DSE versus ILI. Exploratory correlations were conducted within ILI to identify regions in which activity was associated with degree of weight loss.

Results: Voxel-wise whole-brain comparisons revealed greater reward-processing activity in left caudate for DSE compared with ILI and greater activity in attention- and visual-processing regions for ILI than DSE (P < 0.05, family-wise error corrected). Exploratory analyses revealed that greater weight loss among ILI participants from baseline was associated with brain activation indicative of increased cognitive control and attention and visual processing in response to high-calorie food cues (P < 0.001, uncorrected).

Conclusions: These findings suggest there may be legacy effects of participation in a behavioral weight loss intervention, with reduced reward-related activity and enhanced attention or visual processing in response to high-calorie foods.

Trial registration: ClinicalTrials.gov NCT00017953.

Conflict of interest statement

Conflict of Interest: The authors declare that they have no conflict of interest

© 2019 The Obesity Society.

Figures

Fig. 1.

Mean percent weight loss from baseline for each year by group (ILI vs. DSE). Error bars represent standard error of the mean. Asterisk (*) denotes significant difference (p

Fig. 2.

Statistical maps representing differences between groups in response to high-calorie food cues compared to neutral images. A whole-brain voxel-wise independent samples t-test revealed greater activation for DSE relative to ILI in regions of caudate (A) and greater activation for ILI relative to DSE in angular gyrus (C). For visualization purposes bar graphs display mean Beta weights for each group in the caudate (B) and angular gyrus (D). Error bars represent standard error of the mean.

Fig. 3.

For visualization purposes, scatterplots depicting the relationship between food cue-reactivity and percent weight loss in each of the areas noted in Table 3 are displayed. Regions in which greater activity is associated with greater percent weight loss are presented (A, B, C, D, and E) and the one region in which greater activity is associated with less weight loss is plotted (F). The strength of the association between percent weight loss and activity in right middle frontal gyrus (E) and right middle temporal lobe (F) is attenuated when a potential outlier is removed.