1121
Exp Physiol 99.9 (2014) pp 1121–1127
Themed Issue Issue (Obesity) (Obesity) Review Review Themed
Brain responses to food and weight loss Preeshila Behary and Alexander D. Miras
Experimental Physiology
Division of Diabetes, Endocrinology and Metabolism, Hammersmith Hospital, Imperial College London, Du Cane Road, London, UK
New Findings r What is the topic of this review? This report provides an overview of how functional neuroimaging technology has accelerated our understanding of human eating behaviour. r What advances does it highlight? The human brain responds not only to hunger and satiation, but also to the rewarding value of food. The latter is encoded in the brain reward system, which promotes the consumption of energy-dense food and is dysregulated in the context of obesity. In the majority of obese patients, dieting-induced weight loss is resisted by compensatory activation of the homeostatic and hedonic brain systems, whereas pharmacological and bariatric surgery interventions might be more effective in overcoming this brain response.
In this symposium report, we examine how functional neuroimaging has revolutionized the study of human eating behaviour. In the last 20 years, functional magnetic resonance and positron emission tomography techniques have enabled researchers to understand how the human brain regions that control homeostatic and hedonic eating respond to food in physiological and pathological states. Hypothalamic, brainstem, limbic and cortical brain areas form part of a well-co-ordinated brain system that responds to central and peripheral neuronal, hormonal and nutrient signals. Even in physiological conditions, it promotes the consumption of energy-dense food, because this is advantageous in evolutionary terms. Its function is dysregulated in the context of obesity so as to promote weight gain and resist weight loss. Pharmacological and bariatric surgical interventions might be more successful than lifestyle interventions in inducing weight loss and maintenance because, unlike dieting, they reduce not only hunger but also the reward value of food through their actions in homeostatic and hedonic brain regions. Functional neuroimaging is a research tool that cannot be used in isolation; its findings become meaningful and useful only when combined with data from direct measures of eating behaviour. The neuroimaging technology is continuously improving and is expected to contribute further to the in-depth understanding of the obesity phenotype and accelerate the development of more effective and safer treatments for the condition. (Received 1 April 2014; accepted after revision 9 June 2014) Corresponding author A. D. Miras: Division of Diabetes, Endocrinology and Metabolism, 6th Floor Commonwealth Building, Hammersmith Hospital, Imperial College London, Du Cane Road, London W12 0NN, UK. Email: [email protected]
Human eating behaviour and brain responses to food
The term ‘eating behaviour’ describes all the facets of the relationship of animals or humans with food. This relationship is characterized by the physiological, C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
psychological and social forces that are integrated in the brain and determine not only how much we eat but also what, when and how we eat. The study of eating behaviour is not only interesting from an academic point of view but is also a translational research priority due to the DOI: 10.1113/expphysiol.2014.078303
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
1122
P. Behary and A. D. Miras
exorbitant rise in obesity in most parts of the world. Eating behaviour is affected not only by physiological signals but also by psychological traits, cultural and social influences. Figure 1 illustrates some of the numerous factors that determine human eating behaviour. For the purposes of this report, we will focus on physiology and, in particular, on the responses of the brain to food cues. Until relatively recent times, the concept of eating behaviour was limited to the recognition of hunger and fullness as the major driving forces. Hunger is the predominant physiological signal to seek for or initiate an eating episode or meal. During the consumption of a meal, hunger subsides and satiation increases up to the point at which a meal is terminated. The homeostatic control of food intake takes place predominantly in the hypothalamus and brainstem, responds to peripheral hormonal and neural signals and is activated to counterbalance any drift in adiposity below a certain ‘set point’. The set point theory suggests that body weight is regulated around an ‘ideal’ level, which is determined at any time through the interaction of internal signals (i.e. the adiposity hormone leptin) and the environment. The evolutionary protective mechanism resisting fat loss
Exp Physiol 99.9 (2014) pp 1121–1127
has protected the human species from extinction through famine and war. In the last century, food and, in particular, high-calorie food has become readily available in many parts of the world. Now that our hunger and fullness can be readily satisfied, food is no longer merely fuel for survival but also a source of pleasure. Indeed, the reward value of food has emerged as a major force affecting eating behaviour and food preferences. Not only are our brains ‘hardwired’ to choosing energy-dense unhealthy foods, but their consumption in the context of fullness also suggests that the hedonic control of food intake can override the homeostatic control. In the modern environment, appealing energy-dense food is widely available at relatively low cost. Its characteristics, including appearance, texture, taste, smell and nutrient content, have been deliberately designed to maximize reward responses and promote overconsumption in the absence of hunger. The pleasantness of food alongside the emotional and cognitive aspects of eating behaviour are determined in the ‘reward’ system of the brain; this is comprised of a number of limbic and cortical areas that communicate with each other and with the hypothalamus, predominantly
Figure 1. Illustration of some of the factors affecting eating behaviour
C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
Exp Physiol 99.9 (2014) pp 1121–1127
Brain responses to food and weight loss
1123
Table 1. The main regions and functions of the brain system involved with the hedonic, cognitive and emotional evaluation of food Region of the brain
Proposed function
Orbitofrontal cortex Amygdala Insula Dorsal striatum (caudate and putamen) Ventral striatum (nucleus accumbens) Hippocampus Anterior cingulate cortex
Dorsolateral prefrontal cortex
Receives inputs from the five basic senses (taste, olfaction, touch, hearing and vision), but also visceral sensory signals. Their integration leads to decision making, goal-directed behaviour and prediction of the anticipated reward value of specific actions Processes positive and negative emotions; plays a role in stimulus–reward association, learning and aversion Incorporates the primary taste cortex and is involved in the recognition of taste, memory and conditioned taste aversion, risk taking, anticipation, decision making and addiction Involved in learning and goal-directed behaviour through the association of a specific action with its expected reward value, including both food and non-food rewards Responds to both positive and negative stimuli from the environment and determines goal-directed behaviour. Novel stimuli activate the nucleus accumbens, which in turn initiates seeking behaviour and motivation to the stimulus. Responds directly to both the appetitive and consummatory reward of food and taste Involved in memory function; lesions of the hippocampus increase food intake but also appetitive behaviour Ventral subdivision is involved in assigning emotional valence and determining motivation in response to a stimulus, and also has autonomic and endocrine functions. Dorsal subdivision is mostly involved with cognitive control, including the processing of cognitively demanding information. Involved both with the interpretation and regulation of emotional responses but also with motivation for a specific cue and the assessment of reward and risk Inhibits actions which aim to obtain an immediate reward in response to its potential negative longer term effects
through dopamine, opioid and endocannabinoid neurotransmission. These areas include the orbitofrontal cortex (OFC), amygdala, insula, dorsal and ventral striatum, hippocampus, anterior cingulate cortex and dorsolateral prefrontal cortex, amongst others. The function of these brain regions has been determined mostly from animal studies using targeted stimulation and electrophysiological recordings. The allocation of specific roles to brain regions is an oversimplification considering the complexity of the human brain and behaviour, but is nevertheless useful for the basic understanding of the study findings in the next sections (Table 1 and Fig. 2). The study of human eating behaviour is exceptionally challenging due to the variability of its nature and of the tools used to assess it. Functional neuroimaging has been the most recent and exciting development in the study of eating behaviour in the last two decades because it allows the indirect measurement of brain activation in the living human. Functional magnetic resonance imaging (fMRI) and positron emission tomography (PET) are the mainstay investigative tools for appetite research. Functional magnetic resonance imaging provides both structural and functional information on brain activation by detecting changes in blood flow. It can also be useful for the anatomical identification of white matter tracks connecting regions of the brain and determining the functional connectivity between these regions. Positron emission tomography uses radioactive tracers that allow
the detection and quantification of surrogate markers of neuronal activation, including cerebral blood flow and glucose uptake, but also provides information on neurotransmission and neuroreceptor availability. Visual, gustatory, olfactory and tactile food cues have been used to study brain responses to food during neuroimaging protocols.
Figure 2. Illustration of the anatomical location of some of the regions of the brain system involved with the hedonic, cognitive and emotional evaluation of food Three-dimensional top-down view created using FMRIB Software Library (http://fsl.fmrib.ox.ac.uk/fsl/fslwiki).
C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
1124
P. Behary and A. D. Miras
Brain activation measurements are of limited value on their own; they need to be accompanied by data obtained from the more traditional eating behaviour methodologies (e.g. test meals, behavioural tasks, visual analog scale ratings), because this enables ‘blobs’ in the brain to become more useful and meaningful neural correlates of observed behaviour. The following are examples of studies demonstrating this principle; in healthy adults, the OFC, fusiform gyrus, inferior frontal gyrus and nucleus accumbens are activated preferentially to any visual food cues and also to high-calorie food cues (versus non-food or low-calorie food cues) in fMRI studies (van der Laan et al. 2011), and the increase in associated activation of the OFC when switching from the fed to the fasted state correlates with the increase in appeal bias towards high-calorie food (Goldstone et al. 2009). The cross-talk between the homeostatic and reward brain systems is illustrated in fMRI studies that have shown that the parahippocampal gyrus, amygdala, lateral OFC and inferior frontal gyrus are more active in response to food cues in the hungry than the fed state (van der Laan et al. 2011). In a seminal PET study, the release of dopamine in the dorsal striatum in healthy volunteers correlated with the subjective pleasantness of food (Small et al. 2003), whilst fMRI activation of the insula correlated with cravings for food when healthy volunteers were asked to imagine the sensory properties of their favourite food (Pelchat et al. 2004). Brain responses to food in obesity
In the last 10 years, there has been a marked rise in the number of functional neuroimaging studies in the context of obesity and eating behaviour in humans. In terms of differences in the activation of the homeostatic system, PET studies have showed that the activation of the hypothalamus in response to glucose or a meal is reduced in subjects of normal weight, whereas the pattern is much more ‘sluggish’ in the obese, suggesting that the same meal leads to satiation in the former but not the latter group (Gautier et al. 2000). The majority of fMRI studies that assessed appetitive (i.e. ‘wanting’) or anticipatory responses to visual food cues have shown that obese subjects have a higher activation of the reward system (including the ventral and dorsal striatum, orbitofrontal cortex, insula, amygdala and hippocampus) in response to pictures of foods and/or high-calorie foods compared with normal-weight subjects (e.g. Rothemund et al. 2007). Others have shown higher activation in some and lower activation in other areas of the reward system (e.g. Martin et al. 2010). Consummatory responses can be studied through the delivery of gustatory or olfactory cues, i.e. the genuine consumption of the reward. Using these stimuli, some fMRI studies have shown that the reward system is more active upon receipt of the stimulus in obesity
Exp Physiol 99.9 (2014) pp 1121–1127
(e.g. Szalay et al. 2012). In others, the activation of somatosensory cortex, insula and frontal operculum in anticipation of a milkshake stimulus is higher in obese versus normal-weight adolescent girls, whereas activation of the caudate in response to genuine milkshake receipt is lower in the obese (Stice et al. 2008). Evidence from some of the PET studies indirectly supports the latter finding, with reduced dopamine receptor availability (using dopamine receptor D2 or D3 ligands) in the striatum of obese subjects and with dopamine receptor levels being negatively correlated with body mass index (e.g. Wang et al. 2001). In an attempt to explain these discrepant findings, a number of models have been developed. The one which is popular in the literature suggests that both appetitive/anticipatory and consummatory responses are higher in those at risk for obesity, and lead to the overconsumption of high-calorie foods (Carnell et al. 2012). Eventually, this repeated exposure to palatable foods leads to the downregulation of dopamine receptors and reduced consummatory responses, predominantly in the striatum. This state of high appetitive and low consummatory responses resembles the addiction model to recreational drugs. Consequently, obese subjects overconsume food/high-calorie food in an attempt to compensate for the reduced initial ‘kick’ they get from their consumption. At this stage, this dynamic model works as a vicious cycle, causing weight gain, which is incredibly difficult to break. The addiction model is further supported by a meta-analysis of 10 studies that compared whole brain responses to food cues in obese versus normal-weight individuals (Brooks et al. 2013). This showed that clusters within brain regions involved with the cognitive evaluation of food cues were activated more in the obese, whilst clusters in regions linked to cognitive control and interoception were activated less. It is noteworthy that, in contrast to the previously mentioned studies, this meta-analysis did not identify differences in activation between the groups in clusters within the striatum. The problem with the results of these numerous neuroimaging studies in obesity is that there are significant discrepancies between them, not only regarding which areas of the brain are activated and which not, but even regarding the direction of the activation. The most important reason behind the discrepancies probably lies in the nature of obesity itself, which is not a homogeneous condition. The same degree of obesity in two subjects does not necessarily imply similar aetiology, and therefore, imaging participants based purely on a crude measure like body mass index may lead to inconsistent results. Another reason is variability in the study design and neuroimaging protocols used (e.g. differences in the length of fasting, gender distribution, genetic predisposition, personality and psychological traits). C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
Exp Physiol 99.9 (2014) pp 1121–1127
Brain responses to food and weight loss
Brain responses to food during weight loss Dietary interventions. Very few neuroimaging studies have examined participants after prolonged caloric restriction or weight loss through dietary means. In the context of obesity, in a small fMRI study of obese subjects, 10% weight loss through a low-calorie diet for 6–8 weeks was associated with higher activation of brain reward (e.g. brainstem, parahippocampal gyrus) and decision-making systems (e.g. middle temporal gyrus, inferior frontal gyrus), but lower activation in the hypothalamus, emotional control (e.g. amygdala) and motor-planning regions (precentral gyrus) in response to food pictures, compared with baseline (Rosenbaum et al. 2008). In an extension of this study, the functional connectivity of the hypothalamus with visual processing (e.g. occipital fusiform and occipital cortex) and attention areas (e.g. anterior cingulate cortex) increased after weight loss, consistent with patients’ reports of higher ‘sensitivity’ and ‘perception’ of food cues (Hinkle et al. 2013). The differences in brain responses to food depending on the weight-loss treatment were highlighted in a study comparing fMRI activation in participants undergoing a lifestyle versus gastric banding intervention (Bruce et al. 2014). At similar weight loss, participants in the lifestyle arm exhibited higher activation in brain regions involved in food cue processing (medial prefrontal cortex and precuneus) than participants who had undergone gastric banding. Overall, these neuroimaging findings are consistent with the behavioural literature and suggest that weight loss through dieting activates compensatory brain responses that increase the reward, perception and preoccupation with food and translate to enhanced appetitive and consummatory behavioural responses to food. The latter resist further weight loss, promote weight regain and might explain, at least in part, why calorie-restriction-induced weight loss is short lasting in the vast majority of humans who attempt it. Pharmacotherapy. Drug development for obesity has
encountered plenty of failures and has left clinicians with the only option of orlistat, a drug that reduces dietary fat uptake from the gut. Centrally acting appetite suppressants, such as rimonabant, a cannabinoid CB1 antagonist, and sibutramine, a serotonin and noradrenaline reuptake inhibitor, were effective in causing weight loss but were withdrawn from the market due to increased risk of psychological and cardiovascular complications, respectively. These drugs not only decrease hunger but have also been shown in fMRI studies to reduce activation in key reward areas, including the OFC and amygdala, in response to palatable food cues (Fletcher et al. 2010; Horder et al. 2010).
1125
The most promising treatments for obesity are analogues of anorexigenic gut hormones. Gut hormones, including glucagon-like peptide 1 (GLP-1) and peptide tyrosine tyrosine (PYY), are physiologically produced by the gut in response to food intake and act as satiety signals to the hypothalamus and brainstem. They also act on the reward system to reduce the hedonic value of food, both in animals and humans. Glucagon-like peptide 1 analogues/receptor agonists are currently used clinically for the treatment of type 2 diabetes, because they cause reductions not only in glucose but also in weight. Indeed, in a recent fMRI study, intravenous infusion of GLP-1 and PYY in healthy, non-obese adults not only led to reduced ad libitum food intake but also had an additive effect to reduce activation across the brain reward system (amygdala, insula, caudate, nucleus accumbens, orbitofrontal cortex and putamen) in response to passive viewing of food pictures (De Silva et al. 2011). The hope is that the combination of gut hormone analogues will have additive or even synergistic effects in reducing caloric intake and improving food choices in obese patients through action on their specific brain receptors. This might make them a much safer and ‘cleaner’ class of obesity drugs compared with the already-mentioned centrally acting agents. Bariatric surgery. Bariatric surgery remains the most
effective treatment for obesity and its comorbidities. Whilst these procedures were designed to cause mechanical food restriction and caloric malabsorption, we now know that, for example, the Roux-en-Y gastric bypass works by increasing satiation and energy expenditure, but also by shifting food preferences from high-fat and high-sugar options to healthier choices (Miras & le Roux, 2013). The available fMRI studies in the field have consistently shown that Roux-en-Y gastric bypass decreases the activation of brain regions involved in hedonic control., including the lentiform nucleus, putamen and dorsolateral prefrontal cortex, in response to high-calorie but not to low-calorie food cues (Ochner et al. 2012). We have shown that not all bariatric procedures are the same. In our fMRI study, patients after Roux-en-Y gastric bypass had lower activation than patients after gastric banding surgery in brain reward systems, including the orbitofrontal cortex, amygdala, caudate nucleus, nucleus accumbens and hippocampus, in response to high-calorie food pictures (Scholtz et al. 2014). This was associated with lower palatability and appeal of high-calorie foods in Roux-en-Y gastric bypass compared with gastric banding patients. We are currently investigating whether gut hormones, such as GLP-1 and PYY, are the mediators
C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
1126
P. Behary and A. D. Miras
of this selective reduction in food reward that takes place after Roux-en-Y gastric bypass but not gastric banding surgery. Conclusion
In summary, our understanding of the human brain homeostatic and reward responses to food has increased dramatically since the introduction of functional neuroimaging in eating behaviour research. We now know much more about the ‘black box’ and how it responds to different types of food, the different internal states of hunger and satiation, the effects of personality and psychological states, how this system is dysregulated in obesity and how it reacts to different weight-loss treatments. The discrepancies and variability of functional neuroimaging findings have taught us that we should be very careful when designing, performing, analysing and interpreting eating behaviour studies. Like every new research technique, functional neuroimaging needs to evolve even more in terms of validity, reliability, resolution and versatility. Such improvements will enable it to provide more robust data on the highly complex human eating behaviour, accelerate the development of new obesity therapies and even be used not only as a research tool but also as a clinical stratification tool that can assist clinicians in choosing the right treatment, for the right patient, at the right time. Call for comments
Readers are invited to give their opinion on this article. To submit a comment, go to: http://ep.physoc.org/letters/ submit/expphysiol;99/9/1121 References Brooks SJ, Cedernaes J & Schioth HB (2013). Increased prefrontal and parahippocampal activation with reduced dorsolateral prefrontal and insular cortex activation to food images in obesity: a meta-analysis of FMRI studies. PLoS One 8, e60393. Bruce AS, Bruce JM, Ness AR, Lepping RJ, Malley S, Hancock L, Powell J, Patrician TM, Breslin FJ, Martin LE, Donnelly JE, Brooks WM & Savage CR (2014). A comparison of functional brain changes associated with surgical versus behavioral weight loss. Obesity (Silver Spring) 22, 337–343. Carnell S, Gibson C, Benson L, Ochner CN & Geliebter A (2012). Neuroimaging and obesity: current knowledge and future directions. Obes Rev 13, 43–56. De Silva A, Salem V, Long CJ, Makwana A, Newbould RD, Rabiner EA, Ghatei MA, Bloom SR, Matthews PM, Beaver JD & Dhillo WS (2011). The gut hormones PYY 3–36 and GLP-1 7–36 amide reduce food intake and modulate brain activity in appetite centers in humans. Cell Metab 14, 700–706.
Exp Physiol 99.9 (2014) pp 1121–1127
Fletcher PC, Napolitano A, Skeggs A, Miller SR, Delafont B, Cambridge VC, de Wit S, Nathan PJ, Brooke A, O’Rahilly S, Farooqi IS & Bullmore ET (2010). Distinct modulatory effects of satiety and sibutramine on brain responses to food images in humans: a double dissociation across hypothalamus, amygdala, and ventral striatum. J Neurosci 30, 14346–14355. Gautier JF, Chen K, Salbe AD, Bandy D, Pratley RE, Heiman M, Ravussin E, Reiman EM & Tataranni PA (2000). Differential brain responses to satiation in obese and lean men. Diabetes 49, 838–846. Goldstone AP, Prechtl de Hernandez CG, Beaver JD, Muhammed K, Croese C, Bell G, Durighel G, Hughes E, Waldman AD, Frost G & Bell JD (2009). Fasting biases brain reward systems towards high-calorie foods. Eur J Neurosci 30, 1625–1635. Hinkle W, Cordell M, Leibel R, Rosenbaum M & Hirsch J (2013). Effects of reduced weight maintenance and leptin repletion on functional connectivity of the hypothalamus in obese humans. PLoS One 8, e59114. Horder J, Harmer CJ, Cowen PJ & McCabe C (2010). Reduced neural response to reward following 7 days treatment with the cannabinoid CB1 antagonist rimonabant in healthy volunteers. Int J Neuropsychopharmacol 13, 1103–1113. Martin LE, Holsen LM, Chambers RJ, Bruce AS, Brooks WM, Zarcone JR, Butler MG & Savage CR (2010). Neural mechanisms associated with food motivation in obese and healthy weight adults. Obesity (Silver Spring) 18, 254–260. Miras AD & le Roux CW (2013). Mechanisms underlying weight loss after bariatric surgery. Nat Rev Gastroenterol Hepatol 10, 575–584. Ochner CN, Stice E, Hutchins E, Afifi L, Geliebter A, Hirsch J & Teixeira J (2012). Relation between changes in neural responsivity and reductions in desire to eat high-calorie foods following gastric bypass surgery. Neuroscience 209, 128–135. Pelchat ML, Johnson A, Chan R, Valdez J & Ragland JD (2004). Images of desire: food-craving activation during fMRI. Neuroimage 23, 1486–1493. Rosenbaum M, Sy M, Pavlovich K, Leibel RL & Hirsch J (2008). Leptin reverses weight loss-induced changes in regional neural activity responses to visual food stimuli. J Clin Invest 118, 2583–2591. Rothemund Y, Preuschhof C, Bohner G, Bauknecht HC, Klingebiel R, Flor H & Klapp BF (2007). Differential activation of the dorsal striatum by high-calorie visual food stimuli in obese individuals. Neuroimage 37, 410–421. Scholtz S, Miras AD, Chhina N, Prechtl CG, Sleeth ML, Daud NM, Ismail NA, Durighel G, Ahmed AR, Olbers T, Vincent RP, Alaghband-Zadeh J, Ghatei MA, Waldman AD, Frost GS, Bell JD, le Roux CW & Goldstone AP (2014). Obese patients after gastric bypass surgery have lower brain-hedonic responses to food than after gastric banding. Gut 63, 891–902. Small DM, Jones-Gotman M & Dagher A (2003). Feeding-induced dopamine release in dorsal striatum correlates with meal pleasantness ratings in healthy human volunteers. Neuroimage 19, 1709–1715. C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
Exp Physiol 99.9 (2014) pp 1121–1127
Brain responses to food and weight loss
Stice E, Spoor S, Bohon C, Veldhuizen MG & Small DM (2008). Relation of reward from food intake and anticipated food intake to obesity: a functional magnetic resonance imaging study. J Abnorm Psychol 117, 924–935. Szalay C, Aradi M, Schwarcz A, Orsi G, Perlaki G, Nemeth L, Hanna S, Takacs G, Szabo I, Bajnok L, Vereczkei A, Doczi T, Janszky J, Komoly S, Ors Horvath P, Lenard L & Karadi Z (2012). Gustatory perception alterations in obesity: an fMRI study. Brain Res 1473, 131–140. van der Laan LN, de Ridder DT, Viergever MA & Smeets PA (2011). The first taste is always with the eyes: a meta-analysis on the neural correlates of processing visual food cues. Neuroimage 55, 296–303. Wang GJ, Volkow ND, Logan J, Pappas NR, Wong CT, Zhu W, Netusil N & Fowler JS (2001). Brain dopamine and obesity. Lancet 357, 354–357.
1127
Additional Information Competing interests None declared. Funding The Section is funded by grants from the Medical Research Council (MRC), Biotechnology and Biological Sciences Research Council, National Instititute for Health Research (NIHR), an Integrative Mammalian Biology (IMB) Capacity Building Award and an FP7-HEALTH-2009-241592 EuroCHIP grant and is supported by the NIHR Imperial Biomedical Research Centre Funding Scheme. A.D.M. has received funding from an MRC clinical research training fellowship and centenary award and from an NIHR lectureship.
C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
Exp Physiol 99.9 (2014) pp 1121–1127
Themed Issue Issue (Obesity) (Obesity) Review Review Themed
Brain responses to food and weight loss Preeshila Behary and Alexander D. Miras
Experimental Physiology
Division of Diabetes, Endocrinology and Metabolism, Hammersmith Hospital, Imperial College London, Du Cane Road, London, UK
New Findings r What is the topic of this review? This report provides an overview of how functional neuroimaging technology has accelerated our understanding of human eating behaviour. r What advances does it highlight? The human brain responds not only to hunger and satiation, but also to the rewarding value of food. The latter is encoded in the brain reward system, which promotes the consumption of energy-dense food and is dysregulated in the context of obesity. In the majority of obese patients, dieting-induced weight loss is resisted by compensatory activation of the homeostatic and hedonic brain systems, whereas pharmacological and bariatric surgery interventions might be more effective in overcoming this brain response.
In this symposium report, we examine how functional neuroimaging has revolutionized the study of human eating behaviour. In the last 20 years, functional magnetic resonance and positron emission tomography techniques have enabled researchers to understand how the human brain regions that control homeostatic and hedonic eating respond to food in physiological and pathological states. Hypothalamic, brainstem, limbic and cortical brain areas form part of a well-co-ordinated brain system that responds to central and peripheral neuronal, hormonal and nutrient signals. Even in physiological conditions, it promotes the consumption of energy-dense food, because this is advantageous in evolutionary terms. Its function is dysregulated in the context of obesity so as to promote weight gain and resist weight loss. Pharmacological and bariatric surgical interventions might be more successful than lifestyle interventions in inducing weight loss and maintenance because, unlike dieting, they reduce not only hunger but also the reward value of food through their actions in homeostatic and hedonic brain regions. Functional neuroimaging is a research tool that cannot be used in isolation; its findings become meaningful and useful only when combined with data from direct measures of eating behaviour. The neuroimaging technology is continuously improving and is expected to contribute further to the in-depth understanding of the obesity phenotype and accelerate the development of more effective and safer treatments for the condition. (Received 1 April 2014; accepted after revision 9 June 2014) Corresponding author A. D. Miras: Division of Diabetes, Endocrinology and Metabolism, 6th Floor Commonwealth Building, Hammersmith Hospital, Imperial College London, Du Cane Road, London W12 0NN, UK. Email: [email protected]
Human eating behaviour and brain responses to food
The term ‘eating behaviour’ describes all the facets of the relationship of animals or humans with food. This relationship is characterized by the physiological, C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
psychological and social forces that are integrated in the brain and determine not only how much we eat but also what, when and how we eat. The study of eating behaviour is not only interesting from an academic point of view but is also a translational research priority due to the DOI: 10.1113/expphysiol.2014.078303
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
1122
P. Behary and A. D. Miras
exorbitant rise in obesity in most parts of the world. Eating behaviour is affected not only by physiological signals but also by psychological traits, cultural and social influences. Figure 1 illustrates some of the numerous factors that determine human eating behaviour. For the purposes of this report, we will focus on physiology and, in particular, on the responses of the brain to food cues. Until relatively recent times, the concept of eating behaviour was limited to the recognition of hunger and fullness as the major driving forces. Hunger is the predominant physiological signal to seek for or initiate an eating episode or meal. During the consumption of a meal, hunger subsides and satiation increases up to the point at which a meal is terminated. The homeostatic control of food intake takes place predominantly in the hypothalamus and brainstem, responds to peripheral hormonal and neural signals and is activated to counterbalance any drift in adiposity below a certain ‘set point’. The set point theory suggests that body weight is regulated around an ‘ideal’ level, which is determined at any time through the interaction of internal signals (i.e. the adiposity hormone leptin) and the environment. The evolutionary protective mechanism resisting fat loss
Exp Physiol 99.9 (2014) pp 1121–1127
has protected the human species from extinction through famine and war. In the last century, food and, in particular, high-calorie food has become readily available in many parts of the world. Now that our hunger and fullness can be readily satisfied, food is no longer merely fuel for survival but also a source of pleasure. Indeed, the reward value of food has emerged as a major force affecting eating behaviour and food preferences. Not only are our brains ‘hardwired’ to choosing energy-dense unhealthy foods, but their consumption in the context of fullness also suggests that the hedonic control of food intake can override the homeostatic control. In the modern environment, appealing energy-dense food is widely available at relatively low cost. Its characteristics, including appearance, texture, taste, smell and nutrient content, have been deliberately designed to maximize reward responses and promote overconsumption in the absence of hunger. The pleasantness of food alongside the emotional and cognitive aspects of eating behaviour are determined in the ‘reward’ system of the brain; this is comprised of a number of limbic and cortical areas that communicate with each other and with the hypothalamus, predominantly
Figure 1. Illustration of some of the factors affecting eating behaviour
C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
Exp Physiol 99.9 (2014) pp 1121–1127
Brain responses to food and weight loss
1123
Table 1. The main regions and functions of the brain system involved with the hedonic, cognitive and emotional evaluation of food Region of the brain
Proposed function
Orbitofrontal cortex Amygdala Insula Dorsal striatum (caudate and putamen) Ventral striatum (nucleus accumbens) Hippocampus Anterior cingulate cortex
Dorsolateral prefrontal cortex
Receives inputs from the five basic senses (taste, olfaction, touch, hearing and vision), but also visceral sensory signals. Their integration leads to decision making, goal-directed behaviour and prediction of the anticipated reward value of specific actions Processes positive and negative emotions; plays a role in stimulus–reward association, learning and aversion Incorporates the primary taste cortex and is involved in the recognition of taste, memory and conditioned taste aversion, risk taking, anticipation, decision making and addiction Involved in learning and goal-directed behaviour through the association of a specific action with its expected reward value, including both food and non-food rewards Responds to both positive and negative stimuli from the environment and determines goal-directed behaviour. Novel stimuli activate the nucleus accumbens, which in turn initiates seeking behaviour and motivation to the stimulus. Responds directly to both the appetitive and consummatory reward of food and taste Involved in memory function; lesions of the hippocampus increase food intake but also appetitive behaviour Ventral subdivision is involved in assigning emotional valence and determining motivation in response to a stimulus, and also has autonomic and endocrine functions. Dorsal subdivision is mostly involved with cognitive control, including the processing of cognitively demanding information. Involved both with the interpretation and regulation of emotional responses but also with motivation for a specific cue and the assessment of reward and risk Inhibits actions which aim to obtain an immediate reward in response to its potential negative longer term effects
through dopamine, opioid and endocannabinoid neurotransmission. These areas include the orbitofrontal cortex (OFC), amygdala, insula, dorsal and ventral striatum, hippocampus, anterior cingulate cortex and dorsolateral prefrontal cortex, amongst others. The function of these brain regions has been determined mostly from animal studies using targeted stimulation and electrophysiological recordings. The allocation of specific roles to brain regions is an oversimplification considering the complexity of the human brain and behaviour, but is nevertheless useful for the basic understanding of the study findings in the next sections (Table 1 and Fig. 2). The study of human eating behaviour is exceptionally challenging due to the variability of its nature and of the tools used to assess it. Functional neuroimaging has been the most recent and exciting development in the study of eating behaviour in the last two decades because it allows the indirect measurement of brain activation in the living human. Functional magnetic resonance imaging (fMRI) and positron emission tomography (PET) are the mainstay investigative tools for appetite research. Functional magnetic resonance imaging provides both structural and functional information on brain activation by detecting changes in blood flow. It can also be useful for the anatomical identification of white matter tracks connecting regions of the brain and determining the functional connectivity between these regions. Positron emission tomography uses radioactive tracers that allow
the detection and quantification of surrogate markers of neuronal activation, including cerebral blood flow and glucose uptake, but also provides information on neurotransmission and neuroreceptor availability. Visual, gustatory, olfactory and tactile food cues have been used to study brain responses to food during neuroimaging protocols.
Figure 2. Illustration of the anatomical location of some of the regions of the brain system involved with the hedonic, cognitive and emotional evaluation of food Three-dimensional top-down view created using FMRIB Software Library (http://fsl.fmrib.ox.ac.uk/fsl/fslwiki).
C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
1124
P. Behary and A. D. Miras
Brain activation measurements are of limited value on their own; they need to be accompanied by data obtained from the more traditional eating behaviour methodologies (e.g. test meals, behavioural tasks, visual analog scale ratings), because this enables ‘blobs’ in the brain to become more useful and meaningful neural correlates of observed behaviour. The following are examples of studies demonstrating this principle; in healthy adults, the OFC, fusiform gyrus, inferior frontal gyrus and nucleus accumbens are activated preferentially to any visual food cues and also to high-calorie food cues (versus non-food or low-calorie food cues) in fMRI studies (van der Laan et al. 2011), and the increase in associated activation of the OFC when switching from the fed to the fasted state correlates with the increase in appeal bias towards high-calorie food (Goldstone et al. 2009). The cross-talk between the homeostatic and reward brain systems is illustrated in fMRI studies that have shown that the parahippocampal gyrus, amygdala, lateral OFC and inferior frontal gyrus are more active in response to food cues in the hungry than the fed state (van der Laan et al. 2011). In a seminal PET study, the release of dopamine in the dorsal striatum in healthy volunteers correlated with the subjective pleasantness of food (Small et al. 2003), whilst fMRI activation of the insula correlated with cravings for food when healthy volunteers were asked to imagine the sensory properties of their favourite food (Pelchat et al. 2004). Brain responses to food in obesity
In the last 10 years, there has been a marked rise in the number of functional neuroimaging studies in the context of obesity and eating behaviour in humans. In terms of differences in the activation of the homeostatic system, PET studies have showed that the activation of the hypothalamus in response to glucose or a meal is reduced in subjects of normal weight, whereas the pattern is much more ‘sluggish’ in the obese, suggesting that the same meal leads to satiation in the former but not the latter group (Gautier et al. 2000). The majority of fMRI studies that assessed appetitive (i.e. ‘wanting’) or anticipatory responses to visual food cues have shown that obese subjects have a higher activation of the reward system (including the ventral and dorsal striatum, orbitofrontal cortex, insula, amygdala and hippocampus) in response to pictures of foods and/or high-calorie foods compared with normal-weight subjects (e.g. Rothemund et al. 2007). Others have shown higher activation in some and lower activation in other areas of the reward system (e.g. Martin et al. 2010). Consummatory responses can be studied through the delivery of gustatory or olfactory cues, i.e. the genuine consumption of the reward. Using these stimuli, some fMRI studies have shown that the reward system is more active upon receipt of the stimulus in obesity
Exp Physiol 99.9 (2014) pp 1121–1127
(e.g. Szalay et al. 2012). In others, the activation of somatosensory cortex, insula and frontal operculum in anticipation of a milkshake stimulus is higher in obese versus normal-weight adolescent girls, whereas activation of the caudate in response to genuine milkshake receipt is lower in the obese (Stice et al. 2008). Evidence from some of the PET studies indirectly supports the latter finding, with reduced dopamine receptor availability (using dopamine receptor D2 or D3 ligands) in the striatum of obese subjects and with dopamine receptor levels being negatively correlated with body mass index (e.g. Wang et al. 2001). In an attempt to explain these discrepant findings, a number of models have been developed. The one which is popular in the literature suggests that both appetitive/anticipatory and consummatory responses are higher in those at risk for obesity, and lead to the overconsumption of high-calorie foods (Carnell et al. 2012). Eventually, this repeated exposure to palatable foods leads to the downregulation of dopamine receptors and reduced consummatory responses, predominantly in the striatum. This state of high appetitive and low consummatory responses resembles the addiction model to recreational drugs. Consequently, obese subjects overconsume food/high-calorie food in an attempt to compensate for the reduced initial ‘kick’ they get from their consumption. At this stage, this dynamic model works as a vicious cycle, causing weight gain, which is incredibly difficult to break. The addiction model is further supported by a meta-analysis of 10 studies that compared whole brain responses to food cues in obese versus normal-weight individuals (Brooks et al. 2013). This showed that clusters within brain regions involved with the cognitive evaluation of food cues were activated more in the obese, whilst clusters in regions linked to cognitive control and interoception were activated less. It is noteworthy that, in contrast to the previously mentioned studies, this meta-analysis did not identify differences in activation between the groups in clusters within the striatum. The problem with the results of these numerous neuroimaging studies in obesity is that there are significant discrepancies between them, not only regarding which areas of the brain are activated and which not, but even regarding the direction of the activation. The most important reason behind the discrepancies probably lies in the nature of obesity itself, which is not a homogeneous condition. The same degree of obesity in two subjects does not necessarily imply similar aetiology, and therefore, imaging participants based purely on a crude measure like body mass index may lead to inconsistent results. Another reason is variability in the study design and neuroimaging protocols used (e.g. differences in the length of fasting, gender distribution, genetic predisposition, personality and psychological traits). C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
Exp Physiol 99.9 (2014) pp 1121–1127
Brain responses to food and weight loss
Brain responses to food during weight loss Dietary interventions. Very few neuroimaging studies have examined participants after prolonged caloric restriction or weight loss through dietary means. In the context of obesity, in a small fMRI study of obese subjects, 10% weight loss through a low-calorie diet for 6–8 weeks was associated with higher activation of brain reward (e.g. brainstem, parahippocampal gyrus) and decision-making systems (e.g. middle temporal gyrus, inferior frontal gyrus), but lower activation in the hypothalamus, emotional control (e.g. amygdala) and motor-planning regions (precentral gyrus) in response to food pictures, compared with baseline (Rosenbaum et al. 2008). In an extension of this study, the functional connectivity of the hypothalamus with visual processing (e.g. occipital fusiform and occipital cortex) and attention areas (e.g. anterior cingulate cortex) increased after weight loss, consistent with patients’ reports of higher ‘sensitivity’ and ‘perception’ of food cues (Hinkle et al. 2013). The differences in brain responses to food depending on the weight-loss treatment were highlighted in a study comparing fMRI activation in participants undergoing a lifestyle versus gastric banding intervention (Bruce et al. 2014). At similar weight loss, participants in the lifestyle arm exhibited higher activation in brain regions involved in food cue processing (medial prefrontal cortex and precuneus) than participants who had undergone gastric banding. Overall, these neuroimaging findings are consistent with the behavioural literature and suggest that weight loss through dieting activates compensatory brain responses that increase the reward, perception and preoccupation with food and translate to enhanced appetitive and consummatory behavioural responses to food. The latter resist further weight loss, promote weight regain and might explain, at least in part, why calorie-restriction-induced weight loss is short lasting in the vast majority of humans who attempt it. Pharmacotherapy. Drug development for obesity has
encountered plenty of failures and has left clinicians with the only option of orlistat, a drug that reduces dietary fat uptake from the gut. Centrally acting appetite suppressants, such as rimonabant, a cannabinoid CB1 antagonist, and sibutramine, a serotonin and noradrenaline reuptake inhibitor, were effective in causing weight loss but were withdrawn from the market due to increased risk of psychological and cardiovascular complications, respectively. These drugs not only decrease hunger but have also been shown in fMRI studies to reduce activation in key reward areas, including the OFC and amygdala, in response to palatable food cues (Fletcher et al. 2010; Horder et al. 2010).
1125
The most promising treatments for obesity are analogues of anorexigenic gut hormones. Gut hormones, including glucagon-like peptide 1 (GLP-1) and peptide tyrosine tyrosine (PYY), are physiologically produced by the gut in response to food intake and act as satiety signals to the hypothalamus and brainstem. They also act on the reward system to reduce the hedonic value of food, both in animals and humans. Glucagon-like peptide 1 analogues/receptor agonists are currently used clinically for the treatment of type 2 diabetes, because they cause reductions not only in glucose but also in weight. Indeed, in a recent fMRI study, intravenous infusion of GLP-1 and PYY in healthy, non-obese adults not only led to reduced ad libitum food intake but also had an additive effect to reduce activation across the brain reward system (amygdala, insula, caudate, nucleus accumbens, orbitofrontal cortex and putamen) in response to passive viewing of food pictures (De Silva et al. 2011). The hope is that the combination of gut hormone analogues will have additive or even synergistic effects in reducing caloric intake and improving food choices in obese patients through action on their specific brain receptors. This might make them a much safer and ‘cleaner’ class of obesity drugs compared with the already-mentioned centrally acting agents. Bariatric surgery. Bariatric surgery remains the most
effective treatment for obesity and its comorbidities. Whilst these procedures were designed to cause mechanical food restriction and caloric malabsorption, we now know that, for example, the Roux-en-Y gastric bypass works by increasing satiation and energy expenditure, but also by shifting food preferences from high-fat and high-sugar options to healthier choices (Miras & le Roux, 2013). The available fMRI studies in the field have consistently shown that Roux-en-Y gastric bypass decreases the activation of brain regions involved in hedonic control., including the lentiform nucleus, putamen and dorsolateral prefrontal cortex, in response to high-calorie but not to low-calorie food cues (Ochner et al. 2012). We have shown that not all bariatric procedures are the same. In our fMRI study, patients after Roux-en-Y gastric bypass had lower activation than patients after gastric banding surgery in brain reward systems, including the orbitofrontal cortex, amygdala, caudate nucleus, nucleus accumbens and hippocampus, in response to high-calorie food pictures (Scholtz et al. 2014). This was associated with lower palatability and appeal of high-calorie foods in Roux-en-Y gastric bypass compared with gastric banding patients. We are currently investigating whether gut hormones, such as GLP-1 and PYY, are the mediators
C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
1126
P. Behary and A. D. Miras
of this selective reduction in food reward that takes place after Roux-en-Y gastric bypass but not gastric banding surgery. Conclusion
In summary, our understanding of the human brain homeostatic and reward responses to food has increased dramatically since the introduction of functional neuroimaging in eating behaviour research. We now know much more about the ‘black box’ and how it responds to different types of food, the different internal states of hunger and satiation, the effects of personality and psychological states, how this system is dysregulated in obesity and how it reacts to different weight-loss treatments. The discrepancies and variability of functional neuroimaging findings have taught us that we should be very careful when designing, performing, analysing and interpreting eating behaviour studies. Like every new research technique, functional neuroimaging needs to evolve even more in terms of validity, reliability, resolution and versatility. Such improvements will enable it to provide more robust data on the highly complex human eating behaviour, accelerate the development of new obesity therapies and even be used not only as a research tool but also as a clinical stratification tool that can assist clinicians in choosing the right treatment, for the right patient, at the right time. Call for comments
Readers are invited to give their opinion on this article. To submit a comment, go to: http://ep.physoc.org/letters/ submit/expphysiol;99/9/1121 References Brooks SJ, Cedernaes J & Schioth HB (2013). Increased prefrontal and parahippocampal activation with reduced dorsolateral prefrontal and insular cortex activation to food images in obesity: a meta-analysis of FMRI studies. PLoS One 8, e60393. Bruce AS, Bruce JM, Ness AR, Lepping RJ, Malley S, Hancock L, Powell J, Patrician TM, Breslin FJ, Martin LE, Donnelly JE, Brooks WM & Savage CR (2014). A comparison of functional brain changes associated with surgical versus behavioral weight loss. Obesity (Silver Spring) 22, 337–343. Carnell S, Gibson C, Benson L, Ochner CN & Geliebter A (2012). Neuroimaging and obesity: current knowledge and future directions. Obes Rev 13, 43–56. De Silva A, Salem V, Long CJ, Makwana A, Newbould RD, Rabiner EA, Ghatei MA, Bloom SR, Matthews PM, Beaver JD & Dhillo WS (2011). The gut hormones PYY 3–36 and GLP-1 7–36 amide reduce food intake and modulate brain activity in appetite centers in humans. Cell Metab 14, 700–706.
Exp Physiol 99.9 (2014) pp 1121–1127
Fletcher PC, Napolitano A, Skeggs A, Miller SR, Delafont B, Cambridge VC, de Wit S, Nathan PJ, Brooke A, O’Rahilly S, Farooqi IS & Bullmore ET (2010). Distinct modulatory effects of satiety and sibutramine on brain responses to food images in humans: a double dissociation across hypothalamus, amygdala, and ventral striatum. J Neurosci 30, 14346–14355. Gautier JF, Chen K, Salbe AD, Bandy D, Pratley RE, Heiman M, Ravussin E, Reiman EM & Tataranni PA (2000). Differential brain responses to satiation in obese and lean men. Diabetes 49, 838–846. Goldstone AP, Prechtl de Hernandez CG, Beaver JD, Muhammed K, Croese C, Bell G, Durighel G, Hughes E, Waldman AD, Frost G & Bell JD (2009). Fasting biases brain reward systems towards high-calorie foods. Eur J Neurosci 30, 1625–1635. Hinkle W, Cordell M, Leibel R, Rosenbaum M & Hirsch J (2013). Effects of reduced weight maintenance and leptin repletion on functional connectivity of the hypothalamus in obese humans. PLoS One 8, e59114. Horder J, Harmer CJ, Cowen PJ & McCabe C (2010). Reduced neural response to reward following 7 days treatment with the cannabinoid CB1 antagonist rimonabant in healthy volunteers. Int J Neuropsychopharmacol 13, 1103–1113. Martin LE, Holsen LM, Chambers RJ, Bruce AS, Brooks WM, Zarcone JR, Butler MG & Savage CR (2010). Neural mechanisms associated with food motivation in obese and healthy weight adults. Obesity (Silver Spring) 18, 254–260. Miras AD & le Roux CW (2013). Mechanisms underlying weight loss after bariatric surgery. Nat Rev Gastroenterol Hepatol 10, 575–584. Ochner CN, Stice E, Hutchins E, Afifi L, Geliebter A, Hirsch J & Teixeira J (2012). Relation between changes in neural responsivity and reductions in desire to eat high-calorie foods following gastric bypass surgery. Neuroscience 209, 128–135. Pelchat ML, Johnson A, Chan R, Valdez J & Ragland JD (2004). Images of desire: food-craving activation during fMRI. Neuroimage 23, 1486–1493. Rosenbaum M, Sy M, Pavlovich K, Leibel RL & Hirsch J (2008). Leptin reverses weight loss-induced changes in regional neural activity responses to visual food stimuli. J Clin Invest 118, 2583–2591. Rothemund Y, Preuschhof C, Bohner G, Bauknecht HC, Klingebiel R, Flor H & Klapp BF (2007). Differential activation of the dorsal striatum by high-calorie visual food stimuli in obese individuals. Neuroimage 37, 410–421. Scholtz S, Miras AD, Chhina N, Prechtl CG, Sleeth ML, Daud NM, Ismail NA, Durighel G, Ahmed AR, Olbers T, Vincent RP, Alaghband-Zadeh J, Ghatei MA, Waldman AD, Frost GS, Bell JD, le Roux CW & Goldstone AP (2014). Obese patients after gastric bypass surgery have lower brain-hedonic responses to food than after gastric banding. Gut 63, 891–902. Small DM, Jones-Gotman M & Dagher A (2003). Feeding-induced dopamine release in dorsal striatum correlates with meal pleasantness ratings in healthy human volunteers. Neuroimage 19, 1709–1715. C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
Exp Physiol 99.9 (2014) pp 1121–1127
Brain responses to food and weight loss
Stice E, Spoor S, Bohon C, Veldhuizen MG & Small DM (2008). Relation of reward from food intake and anticipated food intake to obesity: a functional magnetic resonance imaging study. J Abnorm Psychol 117, 924–935. Szalay C, Aradi M, Schwarcz A, Orsi G, Perlaki G, Nemeth L, Hanna S, Takacs G, Szabo I, Bajnok L, Vereczkei A, Doczi T, Janszky J, Komoly S, Ors Horvath P, Lenard L & Karadi Z (2012). Gustatory perception alterations in obesity: an fMRI study. Brain Res 1473, 131–140. van der Laan LN, de Ridder DT, Viergever MA & Smeets PA (2011). The first taste is always with the eyes: a meta-analysis on the neural correlates of processing visual food cues. Neuroimage 55, 296–303. Wang GJ, Volkow ND, Logan J, Pappas NR, Wong CT, Zhu W, Netusil N & Fowler JS (2001). Brain dopamine and obesity. Lancet 357, 354–357.
1127
Additional Information Competing interests None declared. Funding The Section is funded by grants from the Medical Research Council (MRC), Biotechnology and Biological Sciences Research Council, National Instititute for Health Research (NIHR), an Integrative Mammalian Biology (IMB) Capacity Building Award and an FP7-HEALTH-2009-241592 EuroCHIP grant and is supported by the NIHR Imperial Biomedical Research Centre Funding Scheme. A.D.M. has received funding from an MRC clinical research training fellowship and centenary award and from an NIHR lectureship.
C 2014 The Authors. Experimental Physiology C 2014 The Physiological Society
Downloaded from Exp Physiol (ep.physoc.org) at California Digital Library on September 13, 2014
