Fasting may increase incentive signaling for nonfood rewards

Abstract

During acute energy deprivation, hunger signaling mechanisms support homeostasis by enhancing incentive for food. There is some evidence (primarily based on nonhuman experiments) that fasting heightens incentive signaling for nonfood reward as well. We hypothesized that, consistent with results from research in rodent and nonhuman primates, human participants would evidence increased incentive-related brain activity for nonfood rewards during fast (relative to satiety) and that this increase would be heightened when available rewards were immediate. To assess these possibilities, healthy participants with body mass index between 18 and 29 kg/m² completed a task which engaged participants in opportunities to win immediate and delayed money (Monetary Incentive Delay Task) during 2 neuroimaging sessions (1 postprandial, 1 fasted). Analyses of participants (N = 18 included, body mass index 22.12± 2.72, age 21.39± 3.52) focused on brain activity during the incentive window of the task. Region of interest, as well as whole-brain analyses, supported the hypothesized increase in incentive signaling during fasting in regions that included caudate and putamen. No evidence of interaction was observed between fasting and the effect of reward immediacy or reward magnitude. Although provisional given the modest sample size, these results suggest that acute fasting can heighten incentive signaling for nonfood rewards.

Keywords: Fasting; Hunger; Incentive tracking; Ingestive state; Reward; fMRI.

Figure 1:

Selection of subjects for the study.

Figure 2:

Visualization of a large-immediate reward trial of the intertemporal MID-Task. Each trial began with a fixation cross, which remained on the screen for a mean of 2 seconds (exponential distribution of duration of fixation cross). Fixation was followed by 1 of 5 reward cues for a randomized duration between 3.5–5.5 seconds (“incentive period”). Next a large response target ‘X’ appeared over the reward cue, and participants responded with a button press as quickly as they were able, followed immediately by presentation of feedback indicating success or failure. Winning was kept at approximately 60% to promote engagement, in line with typical task parameters [21].

Figure 3:

Three measurements of hunger scores on fast and fed days. In plot, boxes identify medians and inter quartile range, with whiskers extending from the hinge to the largest / smallest value no further than 1.5 * the inter-quartile range (IQR) from the hinge, and outliers included as small filled circles. Time1: Upon arrival (prior to eating during the fed day); Time2: Before scan (post-eating if it was the fed day); Time3: After scan.

Figure 4:

Reaction time data on fast and fed days. In plot, boxes identify medians and inter quartile range, with whiskers extending from the hinge to the largest / smallest value no further than 1.5 * IQR from the hinge, and outliers included as small filled circles.

Figure 5:

Red areas denote a priori regions expected to track incentive value based on prior work with the task. This mask included midbrain, insula, supplementary motor area (SMA), putamen, caudate, thalamus and accumbens (all bilateral).

Figure 6 :

Beta values extracted from the ROI mask, on fast and fed days. In plot, boxes identify medians and inter quartile range, with whiskers extending from the hinge to the largest / smallest value no further than 1.5 * IQR from the hinge, and outliers included as small filled circles.

Figure 7 :

Whole brain contrast of incentive period (all rewarded trials) during fast - fed conditions based on whole-brain paired-t test. Greater signal increase was observed during the fast condition in several regions including caudate, putamen, anterior cingulate, and left precentral gyrus. The results are based on repeated measures permutation-based nonparametric test (randomize, FSL tool) approach, correcting for multiple comparisons by threshold-free cluster enhancement. p- values <0.05 were considered significant.