Food reward from a behavioural and (neuro)physiological perspective

Abstract

Food reward is an important driver of food intake and triggers consumption of foods for pleasure, so-called hedonic eating, even in the absence of any energy deficits. Hedonic eating can trigger overeating and may therefore lead to obesity. Given the rise in obesity rates and the health risks associated with being obese, hedonic eating and food reward are important phenomena to study. This thesis aimed to add on to the existing knowledge on food reward. The phenomenon was approached from a behavioural, sensory and (neuro)physiological perspective in healthy, lean and in obese gastric bypass populations. For the behavioural perspective, the main outcome measure used in this thesis was food preferences. To be able to study food preferences for four macronutrient and two taste categories, a new food preference task was developed. In chapter 2, the development and validation of the Macronutrient and Taste Preference Ranking Task (MTPRT) were described. The MTPRT uses a ranking method to determine preferences for four macronutrient (high-carbohydrate, high-fat, high-protein, low-energy) and two taste (sweet and savoury) categories. For the sensory and physiological perspective, focus was put on the endocannabinoid system (ECS): a neuromodulatory system that plays a role in food reward. To gain more insight into this role, the effect of ECS modulation with pharmacological challenges on sensory perception of sweet taste and on food preferences were studied, as well as endocannabinoid responses to food intake. In chapter 3 it was shown that inhaling Cannabis with low doses of Δ9-tetrahydrocannabinol (THC) and cannabidiol (CBD) does not alter sweet taste intensity perception and liking in humans, nor does it affect food preferences. Vice versa, in chapter 4 it was found that liking of a food taste does not affect endocannabinoid responses to food intake, after controlling for expectations. When palatability of the food is unknown until the first bite, response of endocannabinoids, ghrelin and pancreatic polypeptide did not differ between a palatable and a neutral food across anticipatory, consummatory and post-ingestive phases of food intake. Endocannabinoid and ghrelin plasma concentrations decreased after food intake, which suggests an orexigenic function for endocannabinoids. In chapters 5, 6 and 7, studies with patients who underwent Roux-en-Y gastric bypass surgery were described. These studies were intended to gain more insight into alterations in food reward in relation to (morbid) obesity and in response to surgical treatment by RYGB surgery. First, in chapter 5 food preferences were assessed before, and at two months and one year after RYGB. It was shown that patients have decreased preference for high-carbohydrate and high-fat foods, and increased preference for low-energy foods after compared with before surgery. In addition, liking ratings for the high-carbohydrate and high-fat foods were decreased after RYGB surgery, whereas liking of low-energy products changed minimally. Potential mechanisms behind these alterations in food preferences include changes in neural processing of food cues and changes in appetite-related gut hormones. In chapter 6, it was shown that alterations in food preferences after RYGB surgery are indeed related to changes in neural activation in response to food cues. With regards to the appetite-related hormones it was shown that plasma concentrations of the endocannabinoid anandamide were increased after compared with before surgery. Plasma concentrations of other endocannabinoids and ghrelin did not change. Moreover, changes in endocannabinoid or ghrelin concentrations did not correlate with changes in food preferences or neural response to food cues. Together, these results suggest that changes in neural processing of food cues contribute to changes in food preferences towards low-energy foods, and provide a first indication that the endocannabinoid system does not seem to play a role in this process. To gain more insight into behavioural responses to food cues, a response-inhibition paradigm was used in chapter 7, in which response-inhibition to high-energy and low-energy food cues was assessed during brain imaging. The behavioural data did not show differences in performance when comparing before and two months after RYGB surgery. The brain imaging data showed that activation in reward-related brain areas was decreased in response to both high- and low-energy food pictures after RYGB surgery. Also, prefrontal brain areas were more activated in response to the high-energy pictures, which suggests improved response inhibition. In conclusion, the findings in this thesis show that modulating the ECS with low doses of THC and CBD does not influence sweet taste perception and liking and food preferences, and vice versa, food taste liking in the absence of expectations does not affect endocannabinoid responses to food intake. With regards to RYGB surgery it was uncovered that changes in food preferences after RYGB surgery are related to altered brain reward processing, but no relation with changes in endocannabinoid tone was found. The success of RYGB surgery and the changes in food choice might in part be caused by an improved inhibitory response to high-energy foods.