Edited by:
Reviewed by:
*Correspondence:
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
The study of food aversion in humans by the induction of illness is ethically unthinkable, and it is difficult to propose a type of food that is disgusting for everybody. However, although cheese is considered edible by most people, it can also be perceived as particularly disgusting to some individuals. As such, the perception of cheese constitutes a good model to study the cerebral processes of food disgust and aversion. In this study, we show that a higher percentage of people are disgusted by cheese than by other types of food. Functional magnetic resonance imaging then reveals that the internal and external globus pallidus and the substantia nigra belonging to the basal ganglia are more activated in participants who dislike or diswant to eat cheese (Anti) than in other participants who like to eat cheese, as revealed following stimulation with cheese odors and pictures. We suggest that the aforementioned basal ganglia structures commonly involved in reward are also involved in the aversive motivated behaviors. Our results further show that the ventral pallidum, a core structure of the reward circuit, is deactivated in Anti subjects stimulated by cheese in the wanting task, highlighting the suppression of motivation-related activation in subjects disgusted by cheese.
Disgust, described for the first time as a basic emotion by
What happens in the brain when food produces such disgust or aversive reaction due to learning? Several studies of patients with brain injury (
Pleasure and reward mechanisms have been established to play central roles in the control of human food intake (
The purpose of the present study was twofold. First, we aimed to estimate the proportion of individuals who are disgusted by cheese. We conducted a survey of the French population to evaluate individual preferences for 75 foods distributed into eight categories. Second, we aimed to compare the brain activities of individuals with aversion to cheese and those who enjoy and commonly eat it using functional magnetic resonance imaging (fMRI); we will refer to these subjects as Anti and Pro, respectively. To grasp the food disgust and aversion processes, we focused our investigation on the food reward components of affective liking and motivational wanting. Participants successively performed two tasks, and we were mainly focused on the brain responses associated with the liking and wanting scores for cheese when these scores were significantly lower in Anti than in Pro subjects, that is when Anti participants, respectively, disliked and diswanted cheese. Animal as well human studies show that the reward circuit includes areas of the basal ganglia that comprise the dorsal and ventral striatum, the globus pallidus, the ventral pallidum, and the substantia nigra (SN), but also areas of midbrain such as the ventral tegmental area (VTA) (
One hundred and forty-five men (38.58 ± 15.36 years; range 16.03–72.45 years) and 187 women (36.07 ± 13.74 years, range 18.63–73.78) matched in age (
The participants completed a questionnaire in which they had to judge how much they liked 75 food items distributed into eight categories: fruit, cheese, charcuterie, fish, vegetable, meat, dessert, and various foods (
List of 75 foods that were distributed into eight categories.
Fruit | Cheese | Charcuterie | Fish | Vegetable | Meat | Dessert | Various | |
---|---|---|---|---|---|---|---|---|
1 | Strawberry | Boiled ham | Sardine | Pea | Beef Bourguignon | Apple pie | Sauerkraut | |
2 | Cherry | Camembert | Cured ham | Trout | Pork chop | Rum baba | Potée | |
3 | Coconut | Gruyère | Dry sausage | Smoked salmon | Courgette | Roast veal | Strawb. char. | Couscous |
4 | Grapefruit | Bouillabaisse | Coq au vin | Chocolat tart | B. Reine | |||
5 | Orange | Liver pâté | Tuna | Bean | Rabbit | Chesnut cream | Pasta | |
6 | Passion fruit | Foie gras | Leek | Steak | Mille feuille | |||
7 | Apple | Salami | Cauliflower | Boiled chicken | Chocolat cake | Quiche | ||
8 | Kiwi | Picodon | Smoked sausage | Spinach | Brownies | Chips | ||
9 | Pear | St Félicien | Merguez | Beet | Floating island | |||
10 | Mango | Munster | Rillettes | Potato | Tiramisu | Green olive | ||
11 | Red currant | |||||||
12 | Pineapple |
The study was conducted in accordance with the Declaration of Helsinki. Participants were informed about the procedures used in the tasks and provided informed written consent as required by the local Institutional Review Board according to French regulations on biomedical experiments with healthy volunteers [Ethical Committee of CPP-Sud Est II (n° CPP 07-043), DGS2007-0554, December 17, 2007]. Handedness was checked by the Edinburgh Handedness Inventory (
Fifteen healthy right-handed subjects liking cheese (11 women; mean age ± SD: 27.5 ± 4.9 years; range: 22.0–36.9 years) and 15 healthy right-handed subjects hating cheese (10 women; mean age ± SD: 30.8 ± 7.6 years; range: 18.5–42.2 years) were assigned to Pro and Anti groups, respectively. Both groups of participants were matched in age (
The participants were further checked as being without known olfactory impairments, rhinal disorders (colds, active allergies, a history of nasal-sinus surgery, or asthma), pregnancy, neurological diseases, ferrous implants (e.g., pacemakers and cochlear implants), or claustrophobia. In addition, they were screened for their olfactory detection ability (odor vs. no odor) and mean breathing cycle duration. Here, included subjects achieved at least 86.7% correct responses (Pro = 96.0 ± 5.2; Anti = 99.0 ± 3.2;
Forty odorants were used: 28 for training purposes and 12 for the fMRI scanning session. For fMRI, stimuli included six cheese varieties (blue cheese, Cheddar, goat cheese, Gruyère, Parmesan, and tomme) and six OFoods (cucumber, fennel, mushroom, pâté, peanut, and pizza) odorants whose names were also included in the questionnaire used for the survey. They were graciously supplied by Mane (Bar-sur-Loup, France), René Laurent (Le Cannet, France), and Givaudan-Roure (Lyon, France) and purchased from Sigma-Aldrich (Saint Quentin-Fallavier, France). The odorants were diluted in odorless mineral oil (Sigma Aldrich, Saint-Quentin-Fallavier, France) to a concentration of 10% in volume. For stimuli presentation, 5 ml of this solution was absorbed into compressed polypropylene filaments inside of a 100-ml white polyethylene squeeze-bottle equipped with a dropper (Fisher Scientific, Illkirch, France).
For fMRI, 12 visual stimuli (landscape mode, 720 × 467) were selected for matching with the odor stimuli listed above (
The odorants were presented to the participants using an airflow olfactometer, which allows the stimuli to be synchronized with breathing (
Participants’ responses were acquired with a five key-press button box that provided logic signals. The five buttons were placed in a configuration similar to the five fingers (thumb, forefinger, middle finger, ring finger, and pinkie) of the right hand, simulating the five levels of a Likert-type scale, respectively. Breathing was recorded using polyvinyl-chloride foot bellows (Herga Electric Limited, Suffolk, UK) secured to the subject’s abdomen with a cotton belt. The participants’ behavioral responses, breathing data, stimulation onset, and trigger signals from the MRI scanner were recorded online (100 Hz sampling rate) on a laptop equipped with a digital acquisition board I/O card (PCI-6527) (National Instruments®, Austin, TX, USA) using LabVIEW software package (National Instruments®). The data were further analyzed using custom routines created with Matlab (The Mathworks, Natick, MA, USA).
Two sessions were planned for each participant on two consecutive days (
During the liking run, the participants were asked to press one of five buttons with the corresponding finger depending on their judgment (thumb: very unpleasant; forefinger: unpleasant; middle finger: neutral; ring finger: pleasant; or pinkie: very pleasant). During the wanting run, the participants were asked to press one of five buttons depending on their desire to eat the food evoked by the stimulus (not at all, not desired, just a little, much desired, or urge) at the present time. If the subject did not smell food during the first session, then they did not press a button. For each stimulus, the subjects had to provide a response as soon as they had performed their liking or wanting judgment. Subjective reward responses of liking and wanting were measured in terms of scores and response times (RT).
General instructions were provided outside the scanner. The day before receiving fMRI scans, the participants were trained outside the MR facility to breathe naturally and regularly without sniffing or holding their breath, to detect odors during inspiration while avoiding sniffing and to provide rapid finger responses using the 5-button box. They were asked to rate the intensities of 28 odorants in a first session and their familiarity in a second session by pressing one of five buttons with the corresponding finger. Control pictures (color rectangles) were synchronously presented with the odor stimuli for 3 s. On the day of fMRI scans, the subjects were specifically instructed to correctly perform the two tasks (liking, wanting) and to avoid confounding them. These instructions have been detailed in a previous study (
All participants were scanned in the hunger state (between 09:45 am and 1:30 pm) and were instructed to have a light breakfast (tea or coffee, plus a slice of bread) no later than 7:00 or 9:00 am, depending on the time at which the scan began. As the metabolic state has been shown to influence liking and wanting performances (
ANOVAs with repeated measurements (
As breathing variations are known to impact brain activation (
Images were acquired using a 1.5-Tesla MAGNETOM Sonata whole-body imager (Siemens Medical®, Erlangen, Germany) equipped with a 4-channel circularly polarized head coil. For functional imaging, we obtained 26 interleaved, 4-mm-thick axial slices using a T2∗-weighted echo-planar sequence with the following parameters: repetition time (TR) = 2500 ms, echo time (TE) = 50 ms, flip angle (FA) = 80°, field-of-view (FOV) = 240 mm × 240 mm, and imaging matrix = 64 × 64 (voxel size: 3.75 mm × 3.75 mm × 4 mm). In total, 390 scans were collected for each functional run. A high-resolution structural T1-weighted anatomical image (inversion-recovery 3D Gradient-Echo sequence, 1 mm × 1 mm × 1 mm) parallel to the bicommissural plane and covering the entire brain was acquired over ∼10 min. Foam wedges were used to restrict head motion. An oil-filled capsule was fixed on the right temple to subsequently locate the right side of the images.
For each subject, the first five volumes of each functional run were discarded to avoid T2∗ non-equilibration effects. We then processed all functional images using a pipeline in Nipype workspace (
Preprocessed data were statistically analyzed on a subject-by-subject basis using the General Linear Model implemented in SPM8. For each subject, activation associated with three factors of interest [food (cheese, OFood), task (liking, wanting), and modality (Odor, Od-Pic)] was modeled as events (corresponding to the onset times for each condition) convolved with both the canonical hrf and its time derivative (
As a dichotomy has been suggested to exist between the ventral and dorsal striatum (
Brain linear regression analyses were further performed to evaluate whether the activation data were correlated with the self-rated liking/wanting data. This assessment was performed for the structures of striatum and midbrain that, according to our hypotheses, were differentially activated between Anti and Pro: the internal and external segments of the globus pallidus (GPi/GPe), the SN and VTA.
Percentages of individuals as a function of the rating scale scores (from 0 to 10) were computed for the eight food categories and are illustrated in Supplementary Figure
Mean liking and wanting scores were determined for cheese and OFood (food factor) in both groups of subjects exposed to 12 odor or Od-Pic (modality factor) stimuli (
Comparing rating scores for cheese between the liking and wanting tasks, a multivariate ANOVA (MANOVA) then revealed a marginally significant group × task interaction (Roy’s
Analysis of RTs for cheese and OFood during both tasks in both groups of subjects further revealed a significant food × group interaction (
We next examined whether differences in liking and wanting rating scores between Anti and Pro could be attributed to varying hunger states. A two-way (time × group) ANOVA with repeated measurements showed a significant increase in the hunger state from the onset to the end of the session (time factor:
Finally, we investigated the impact of the stimuli on the amplitude of inspiratory (inspi) volumes. No significant effect of the group factor was noted (
As mentioned above, no significant differences were observed between the rating scores of the participants in the two groups stimulated with odors only. Therefore, we examined the brain imaging data obtained with the Od-Pic stimuli. We first investigated whether neural networks were differentially activated between Anti and Pro participants ([Anti – Pro] and [Pro – Anti] contrasts) exposed to cheese or OFood stimuli during the liking and wanting tasks (
Brain areas differentially activated in Pro and Anti subjects exposed to cheese Od-Pic stimuli.
Task | Contrast | Brain areas | |||||
---|---|---|---|---|---|---|---|
LIKING | Anti > Pro | Postcentral gyrus | 136 | 5.35 | 64 | -34 | 40 |
Superior frontal gyrus | 78 | 4.21 | 8 | -4 | 66 | ||
Supramarginal gyrus | 43 | 4.11 | -60 | -52 | 40 | ||
Middle frontal gyrus | 30 | 4.08 | -44 | 38 | 34 | ||
Middle frontal gyrus | 36 | 3.83 | 42 | -4 | 62 | ||
Middle frontal gyrus | 27 | 3.56 | -36 | -6 | 66 | ||
GPi/GPe | 19 | 3.18 | -18 | -2 | -4 | ||
Pro > Anti | Anterior cingulate gyrus | 32 | 4.60 | 20 | 44 | 4 | |
Cuneus | 34 | 3.99 | 22 | -80 | 2 | ||
Lingual gyrus | 56 | 3.67 | 32 | -66 | -4 | ||
WANTING | Anti > Pro | Cerebellum | 35 | 4.26 | -10 | -84 | -18 |
Supramarginal gyrus | 38 | 4.02 | -58 | -40 | 50 | ||
Superior frontal gyrus | 62 | 3.93 | 12 | -2 | 66 | ||
VTA | 37 | 3.93 | 6 | -14 | -12 | ||
GPi/GPe | 35 | 3.63 | 26 | -8 | -10 | ||
SN | 3.35 | 18 | -18 | -8 | |||
Pro > Anti | Ant./Lat. orbital gyrus | 38 | 4.79 | -28 | 48 | -6 | |
Anterior cingulate gyrus | 32 | 3.81 | -4 | 38 | 24 |
Second, we investigated whether the brain areas with increased activation by cheese in Anti compared with Pro (
We further observed that in Pro, the liking and wanting scores for OFood were negatively correlated with activation of the GPi/GPe (-18 -2 -4) (liking:
The insula is commonly investigated in studies on disgust (
As the striatum has been involved in aversive learning (
In humans, it is difficult to find disgust reactions to the same type of food because each of us has acquired idiosyncratic reactions. Moreover, food aversion has been under-researched because it is ethically inconceivable to experimentally induce illness in humans. In this study, we showed that a substantial proportion of people in France are disgusted by cheese and that this situation is experimentally favorable for studying the cerebral processes of food disgust and aversion. In these individuals, cheese odors and pictures induce stronger activation of the GPi/GPe, and SN than in people who like and eat cheese. This finding suggests that the GPi/GPe and SN code the hedonic (not only positive but also negative) and motivational components of food reward. Further, we observed that the lack of desire to eat cheese (diswanting) is associated with lack of activation of the VP, a core structure in incentive motivation (
To assess whether disgust for cheese is widespread among individuals, we performed a survey of the French population. It revealed that among the individuals showing disgust for a given food, those disliking cheese represented a higher proportion (6% with a score of 0 to 1 on an 11-point scale) than those disliking the other food categories. This finding is rather surprising because France is the country with the greatest variety of cheeses (Sperat-Czar and Boulenger
Lactose intolerance involves dairy products that contain lactose, but semi-soft and hard cheeses (e.g., cantal, cheddar, and raclette) no longer contain lactose after processing. Therefore, disgust reactions observed for these cheeses in individuals with lactose intolerance are likely caused by a generalization effect. In the absence of identified intolerance (e.g., due to genetic predisposition), disgust for cheese shared by several members of the same family begs the question of whether it results from a simple social transmission, is in sum vicariously acquired, or is the consequence of an epigenetic determinism.
We observed that the liking and wanting scores were globally lower for cheese than for OFood in both Anti and Pro. This result was surprising and suggested that the appetitive properties of the odors and pictures were decreased for cheese compared with OFood. However, focusing on cheese data, we found that the liking and wanting scores were lower in Anti than in Pro subjects stimulated with both odors and pictures, but not with odors only. The responses of the subjects to the odors alone could have been partly influenced by an inability to easily identify cheese odors, as they are fundamentally difficult to name/identify (
During the liking task, marginally higher activation of the GPi/GPe was detected in Anti than in Pro subjects stimulated with cheese but not with OFood. During the wanting task, differential activation of the GPi/GPe, SN and VTA between the groups was also observed for cheese. However, as the VTA was also differentially activated between the groups for OFood, only the SN and GP were considered critical areas with more specific involvement in Anti than in Pro for cheese. We also found that the VP was significantly less activated in Anti subjects stimulated with cheese than with OFood.
The aforementioned structures belong to the reward circuit and are commonly involved in appetitive situations. For instance,
Although neuroimaging studies on the neural substrates of disgust have mainly focused on the insula, caudate and putamen (
How can we explain the exacerbated response of GPi/GPe and SN to cheese in Anti subjects? The projections from the GPe to the GPi and SN pars reticularis, two major output structures of the basal ganglia, are well established (
An important result of our study is the lack of activation of the VP in Anti subjects exposed to cheese stimuli during the wanting task. Several studies have suggested that the VP could represent “an essential convergent point for hedonic and motivational signaling pathways in the brain” (
The VP occupies a prominent place in the circuit mediating the integration of reward perception and adaptive behavioral responses (
The study of food disgust and aversion in humans is difficult because it is rare to find individuals who present disgust for the same type of food, and it is not conceivable to experimentally induce a food aversion by provoking gastro-intestinal symptoms. Our findings show that a higher-than-expected proportion of individuals are disgusted by cheese than by other food categories. These individuals dislike cheese to the point that they cannot eat it, an experimental context quite adapted to studying the brain mechanisms of food disgust. Odor and sight of cheese activate the GPi/GPe and SN, indicating that in addition to encoding reward, these structures may also encode disgust and thus the aversive properties of food. We also report that motivation-related activation of the VP in response to food is suppressed in individuals disgusted by the smell and/or sight of cheese. In brief, our findings show that disgust for cheese, which may be the result from an initial physiological discomfort, is associated with modified activation of the mesocorticolimbic circuitry of reward.
J-PR planned and designed research. NT and J-PR performed the experiments. DM and J-PR analyzed data. J-PR drafted manuscript and prepared figures. J-PR, DM, NT, A-MM and TJ edited and approved final version of manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We thank P. Charrié, L. Morin-Audebrand, and the members of CERMEP (e.g., D. Ibarrola, C. Vighi, and F. Vey) for their invaluable assistance. We are greatly indebted to S. Garcia for designing the software (with Python) to analyze the behavioral and breathing data and to R. Soussignan and R. Margraff for helpful comments on the manuscript. The authors gratefully thank the companies Mane, René Laurent, and Givaudan-Roure, who provided the odorants used as stimuli.
The Supplementary Material for this article can be found online at: