Conditioning by a Previous Experience Impairs the Rewarding Value of a Comfort Meal

Abstract

Background: Meal ingestion induces a postprandial experience that involves homeostatic and hedonic sensations. Our aim was to determine the effect of aversive conditioning on the postprandial reward of a comfort meal.

Methods: A sham-controlled, randomised, parallel, single-blind study was performed on 12 healthy women (6 per group). A comfort meal was tested before and after coupling the meal with an aversive sensation (conditioning intervention), induced by infusion of lipids via a thin naso-duodenal catheter; in the pre- and post-conditioning tests and in the control group, a sham infusion was performed. Participants were instructed that two recipes of a tasty humus would be tested; however, the same meal was administered with a colour additive in the conditioning and post-conditioning tests. Digestive well-being (primary outcome) was measured every 10 min before and 60 min after ingestion using graded scales.

Results: In the aversive conditioning group, the comfort meal in the pre-conditioning test induced a pleasant postprandial experience, which was significantly lower in the post-conditioning test; the effect of aversive conditioning (change from pre- to post-conditioning) was significant as compared to sham conditioning in the control group, which showed no differences between study days.

Conclusion: The hedonic postprandial response to a comfort meal in healthy women is impaired by aversive conditioning.

Clinicaltrials: gov ID: NCT04938934.

Keywords: Pavlovian conditioning; aversive conditioning; digestive sensations; digestive well-being; eating behaviour; food valence; postprandial symptoms.

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1

Experimental design and procedure. In a sham-controlled, parallel, randomised, blind study, a comfort meal was paired with duodenal lipid infusion to induce aversive conditioning (DAY 2) and the responses to the meal were compared before (DAY 1) and after conditioning (DAY 3).

Figure 2

Homeostatic sensations. Concomitant duodenal lipid infusion on day 2 impaired the postprandial response and the effect was significant for abdominal discomfort and nausea (effect measured as the change in the area under the curve on day 2 minus day 1; * p = 0.002 vs. sham infusion). However, data on day 3 show that aversive conditioning in the test group (duodenal lipid infusion on the previous day) did not induce significant effects. Values represent mean ± SE.

Figure 3

Hedonic sensations. Concomitant duodenal lipid infusion on day 2 impaired the postprandial response and the effect was significant for digestive well-being (effect measured as the change in the area under the curve on day 2 minus day 1; * p < 0.001 vs. sham infusion). Data on day 3 show that aversive conditioning in the test group (previous exposure to the aversive stimulus) significantly impaired postprandial well-being (effect measured as the change in the area under the curve on day 3 minus day 1; ** p = 0.004 vs. sham conditioning). Values represent mean ± SE.

Figure 4

Reward value during meal ingestion. The comfort meal was served in 3 portions; meal wanting was measured before each serving and at the end of ingestion; meal liking was measured after each serving. On day 2, concomitant duodenal lipid infusion impaired the ingestive response (effect measured as the change in the area under the curve on day 2 minus day 1 vs. sham infusion; * p = 0.041 for meal wanting; ** p = 0.013 for meal liking). Data on Day 3 show that aversive conditioning in the test group (previous exposure to the aversive stimulus) significantly reduced the valence of the comfort meal (effect measured as the change in the area under the curve on day 3 minus day 1 vs. sham conditioning; †p = 0.023 for meal wanting; ††p = 0.030 for meal liking). Values represent mean ± SE.

Figure 5

Digestive response to meal ingestion. On day 2, duodenal lipid infusion in the test group was associated with a sustained increase in antral cross-sectional area (delayed gastric emptying; * p < 0.001) and diaphragmatic ascent (prolonged abdominal accommodation; ** p = 0.002); effects measured as the changes in the area under the curve on day 2 minus day 1 vs. sham infusion. However, data for day 3 show that aversive conditioning in the test group (previous exposure to the aversive stimulus) did not induce significant effects. Values represent mean ± SE.