Neurobiologic Basis of Craving for Carbohydrates

Nutrition 30 (2014) 252–256

Contents lists available at ScienceDirect

Nutrition

journal homepage: www.nutritionjrnl.com

Review Neurobiologic basis of craving for carbohydrates

Tamara Ventura M.D., Jaime Santander M.D. *, Rafael Torres M.D., Ph.D., Ana Marıa Contreras M.D.

Department of Psychiatry, School of Medicine, Pontiﬁcia Universidad Catolica de Chile, Santiago, Chile article info abstract

Article history: Objectives: There is a relationship between emotional disorders, obesity, and craving for carbo- Received 23 January 2013 hydrates. This relationship complicates the success of treatments aimed at combatting obesity, Accepted 28 June 2013 which is considered to be the epidemic of the twenty-first century. We conducted a review of the neurobiologic basis for carbohydrate craving, with the hope that this understanding will enable the Keywords: design of more efficient therapeutic strategies. Dietary carbohydrates Method: We conducted a non-systematic literature search in PubMed using MeSH. Addictive behavior Results: Research on the basis of carbohydrate craving is varied, but may be grouped into five main Biological evolution Anxiety areas: the serotonergic system, palatability and hedonic response, the motivational system, stress – Eating behavior response systems, and gene environment interaction. Neurobiology Conclusions: The models that integrate motivational systems with palatability and hedonic response studies are the ones that we believe can best explain both craving for carbohydrates and related addictive phenomena. Research has contributed to a greater understanding of the neurobiologic basis of carbohydrate craving. The latter, in turn, contributes to an understanding of the implications, challenges, and possible therapies that might be put in place to cope with this phenomenon. Ó 2014 Elsevier Inc. All rights reserved.

Introduction The scientific literature supports the existence of carbohydrate- craving syndrome, in which carbohydrate intake medicates a The growing global obesity epidemic and its associated pa- dysphoric mood state [8].Thissyndromeisdefined as a food and thologies [1] have focused research on the variables associated mood disorder, characterized by an irresistible urge to consume with overeating and the barriers to achieving changes in nutri- sweets or starchy foods in response to negative moods [9].The tional habits [2]. Research has ranged from social and cultural criteria listed in Table 1 have been proposed to operationalize this variables, such as the availability of food or the effects of phenomenon [8]. Carbohydrate intake appears to temporarily advertising in childhood [2,3], to attempts at elucidating the improve mood in carbohydrate cravers, whereas individuals who mechanisms associated with the tendency toward excessive have no craving for carbohydrates usually complain of fatigue after consumption of fats and carbohydrates. Such research also has eating these foods [9], thus revealing a different relationship with suggested the possibility that there may be an addictive pattern this nutrient. to the consumption of these nutrients under certain circum- However, despite the important role of these relationships, stances [4–6]. the underlying neurobiologic mechanisms remain unclear and We know that emotional factors have become increasingly have not been directly related to emotional factors. Under- important based on associations found among obesity, anxiety, standing these phenomena potentially could be important for and mood disorders [5]. Carbohydrates are exceptionally in- several reasons. On the one hand, this would allow public health volved in these associations, which has led to further research specialists and clinicians to understand realistically the behav- into their patterns and roles. Sugar addiction, carbohydrate ioral changes of individual patients and the general population, craving, and emotional eating have been correlated with obesity considering how our biology works. On the other hand, this and being overweight. This has led to the establishment of a knowledge would allow specific pharmacologic strategies to be strong correlation between obesity, anxiety, and mood disorders developed that may lead to more successful outcomes when [4–10], all of which have important public health implications. dealing with obesity and its consequences. Given the importance of understanding the neurobiologic * Corresponding author: Tel.: þ56 2 27548872; fax: þ56 2 26651951. mechanisms underlying the phenomenon of carbohydrate E-mail address: [email protected] (J. Santander). craving, we aim to review the main neurobiologic hypotheses

Table 1 preference for carbohydrates in the diet [11]. Furthermore, in- * Carbohydrate -craving syndrome criteria [8] creases in brain serotonin improve mood levels in individuals Cravings during the afternoon or at night to eat sweets or starchy who are vulnerable to stress [11]. foods at least four times a week. The serotonergic hypothesis proposes that carbohydrate Eating carbohydrates in the afternoon or night at least four times intake associated with states of anxiety is an attempt at self- a week, supported by records. treatment, given the increase in brain serotonin that occurs af- Carbohydrate craving that occurs in conjunction with dysphoria, which is relieved by eating these foods. ter carbohydrate intake [11]. The behavioral expression of this hypothesis is the craving for carbohydrates that an individual * A carbohydrate is defined as a food consumed between meals that has a carbohydrate to protein ratio of 6:1. experiences within an episode of anxiety. After this macronutrient is consumed, the individual’s emotional state improves. It has been observed that during acute stress there is an in- that explain carbohydrate craving, and link carbohydrate crav- crease in serotonergic neuronal activity due to increased plasma ings with related phenomena, such as emotional eating and levels of cortisol and beta-endorphins [12]. An increase in sero- obesity, from a neurobiologic perspective. tonin regulates the hypothalamic–pituitary–adrenal (HPA) axis, Methods which participates in stress adaptation [13]. If this continues over time, the use of serotonin exceeds its synthesis capacity, which A focused review of the available literature was conducted by using the has been associated with the depletion of tryptophan (Trp), the database www.pubmed.org. The MeSH tool was used and the search matrix precursor of serotonin synthesis, and with depressive mood included the terms stress, carbohydrates, dietary carbohydrates, food, emotional eating, addictive behavior, and evolution. By combining these terms, 2187 articles levels [12]. Changes in brain serotonin are mainly due to the were found, of which some studies were selected according to the objectives of availability of Trp in the brain. Serotonin synthesis is known to this review. increase due to carbohydrates and to decrease or remain unal- tered from proteins in the diet [10]. In fact, it has been observed Results that Trp levels after a meal rich in carbohydrates and low in protein tend to increase along with serotonin, due to a 42% rise in We found dispersed information from different lines of brain Trp compared with diets rich in protein and low in car- research on carbohydrate craving. We then categorized these bohydrates [14]. findings into five groups that had similar research lines. The Based on the serotonergic hypothesis, it is inferred that other serotonergic hypothesis is presented first because it is the first means of increasing brain serotonin levels may constitute po- hypothesis to emerge and the one that has received the most tential treatments for obesity in individuals with anxiety. The use attention in the literature. The hypotheses that follow are pre- of serotonin, including fluoxetine, in rats with a history of food sented in terms of the degree to which they focus on physical restriction under stress, was shown to decrease compulsive consumption variables, thus beginning with palatability and eating (binging). However, the effect was not the same in rats ending with environmental aspects. A summary of these hy- with a history of food restriction that were not under stress [5]. potheses is given in Table 2. Controversy exists regarding the serotonergic hypothesis because some studies show conflicting results as to whether Serotonergic hypothesis increased brain serotonin induced by a diet rich in carbohydrates effectively translates into a change of affective state and appetite Brain serotonin is an important regulator of appetite, mood, [11,15,16]. Moreover, the positive effects on mood of increased and preference for certain macronutrients, among its other brain serotonin induced by a carbohydrate-rich diet occur in functions. Low levels of serotonin stimulate appetite and a patients suffering from anxiety due to carbohydrate craving, late

Table 2 Summary of the main hypotheses that explain carbohydrate craving

Neurobiologic hypothesis Highlights Serotonergic Increased brain serotonin improves mood [11]. Brain serotonin levels depend on the availability of its Trp precursor [14]. Dietary carbohydrates increase the passage of Trp through the blood–brain barrier, unlike proteins, which alter LNAA [14]. Faced with anxiety, an individual eats carbohydrates, which increase brain serotonin, thus improving mood [11]. Palatability and hedonic The pleasurable experience of eating food with high palatability immediately improves mood [17]. response This occurs in individuals with greater genetic sensitivity to sweet taste through the activation of the endogenous opioid system [19]. Faced with anxiety, an individual eats a food with high palatability, activating the hedonic mechanism, which improves mood [16]. Motivational system Carbohydrates act in the motivational system in the same manner as abused substances [6]. This increases dopamine and endogenous opioids, which are associated with a known pleasurable effect, [4] improving mood. If this behavior is repeated over time, structural changes in the brain are produced that generate dependence on highly palatable foods [4,6,23]. Stress response Faced with anxiety associated with stress, the HPA axis activates [30]. Highly palatable foods activate the motivational system and reduce the HPA axis, thus regulating the stress system [30]. Therefore, when faced with anxiety, highly palatable food produces a hedonic reward as well as reducing the state of anxiety [30]. Gene–environment Eating is a coping tool to relieve negative emotions. The behavior is learned through inadequate parenting and environment [9]. It also stems from an inability to distinguish hunger from other aversive internal states [9]. There is greater susceptibility in carriers of the A1 allele of the DRD2 dopamine receptor and carriers of the short allele of the serotonin transporter gene [33,34].

HPA, hypothalamic–pituitary–adrenal; LNAA, large neutral amino acids; Trp, tryptophan 254 T. Ventura et al. / Nutrition 30 (2014) 252–256 luteal phase syndrome, or seasonal affective disorders, and do suggested that there is a withdrawal syndrome for highly not occur in healthy individuals under normal circumstances. palatable foods, leading to the emergence of depressive states These effects of diet on affective states are probably due to a that could be relieved by eating these foods again [20]. relatively reduced amount of brain serotonin, induced by chronic This hypothesis implies that the use of drugs acting on opioid stress [13]. receptors, such as naltrexone, naloxone, and buprenorphine [5, Another critical aspect of the serotonergic hypothesis is that it 21], could be proposed as a possible treatment for carbohy- implicitly assumes that there is a delay between carbohydrate drate craving. intake and mood improvement, because this depends on the With respect to the critical aspects of this hypothesis, a entry of Trp through the blood–brain barrier and serotonin distinction has been made between the two components of the synthesis [17]. The latter is a process that takes at least 60 min eating reward: the “like” and “want” [16]. The “like” aspect cor- [18], whereas emotional eating immediately affects negative responds to palatability and depends on the opioid system, the mood [17]. This temporal dissociation has raised questions about hedonic component of eating, which means the pleasure derived the validity of this hypothesis. from the orosensory stimulation of food. The “want” aspect is under the control of dopamine and is a serious motivational Palatability and hedonic response hypothesis component in the desire to obtain food [16]. In our opinion, these systems are intrinsically connected, thus making this distinction Regulation of eating and energy homeostasis includes, in questionable [22]. addition to hunger and satiety signals, factors such as the sense of reward (or salience) of food, environmental keys, and cogni- Motivation system hypothesis tive factors [16]. This hypothesis proposes that carbohydrate craving during states of anxiety is due to the salience of this The motivation system hypothesis posits that an individual macronutrient, which calms the emotional state via hedonic displays craving for carbohydrates when faced with anxiety, and brain response induced through the endogenous opioid system after consuming carbohydrates, they generate neuronal and [17]. The expression of this hypothesis occurs because the indi- behavioral changes similar to those produced by addictive sub- vidual with anxiety eats highly palatable food that lowers his or stances. This leads to increased cravings, especially when faced her anxiety levels. with states of anxiety or anxiety from withdrawal from highly Eating chocolate immediately affects a negative mood, but palatable food. Thus, consumption is perpetuated and the not a positive or neutral mood. This effect has a short duration carbohydrate-craving syndrome comes into being [6,23]. and comes from the palatability of the food [17]. This hypothesis Substance dependence structurally and functionally alters the proposes that, because the effect on mood is immediate, it is brain’s motivational system, which is composed of the ventral unlikely to be due to the direct effects of the nutrients in neu- tegmental area (VTA), amygdala, prefrontal cortex, hippocampus, rotransmitters or to the psychopharmacologic effect of the active and nucleus accumbens, among other areas. Once changes have ingredients of chocolate, as these are produced 1 to 2 h after occurred in the brain, it is known that anxiety, among other consumption [17]. stimuli, increases the craving for consumption. During withdrawal There is evidence that patients with psychiatric disorders syndromes it is also common to find symptoms of anxiety that (mood, anxiety and eating disorders, and premenstrual syn- predispose the patient to repeated substance consumption [23]. drome) have greater hedonic response levels to concentrated A common etiology has been proposed for eating disorders sucrose solutions. This is a stable, innate, and inherited trait that and substance dependence [24]. In situations where there is is at least partly determined by genetic mechanisms in animals limited and intermittent access to sugar, consumption of sugar and humans, and is associated with higher hedonic response increases levels of dopamine and opioids in the same way as drug levels to sweet tastes, which increases the risk for developing use does [4,24]. This leads to changes in the expression or avail- obesity [19]. ability of dopamine receptors [3]. The role of glutamate, g-ami- There are at least two mechanisms described in the literature nobutyric acid, and opioids in the regulation of food intake also that can determine individual differences in terms of hedonic has been demonstrated. This suggests signaling in the nucleus response to sweet taste. The first mechanism is related to the accumbens on the hedonic component of eating [25,26]. Dopa- perception of sweet taste. Polymorphisms in the region of the mine also has a role in intake: dopaminergic projections from the TAS1R3 gene are associated with a preference for saccharin in 30 VTA to the nucleus accumbens determine how much of a certain strains of mice. Furthermore, it has been shown that variations of food will be ingested and by how much satiety or other metabolic the TAS2R38 gene, are associated with a preference for sweet or behavioral needs will be exceeded. This cognitive process is foods and beverages in children (5–10 years) and undergraduate strongly and constantly stimulated by the modern socioeconomic students (mean age: 22). The second possible mechanism is the environment, generally resulting in increased intake [27]. opioid system in the brain. This system regulates food intake by Bingeing, withdrawal syndrome, and craving for sugar also modulating the associated orosensory reward, instead of intake have been associated with neurochemical changes in the brain being modulated by energy needs. Individuals with a greater that also are produced with addictive drugs [4]. The withdrawal taste for sweets have a basally inhibited brain opioid system, and signs appear to be largely determined by opioid withdrawal in therefore are more sensitive to the effects associated with con- addition to a decrease of dopamine and release of acetylcholine sumption of these foods [19]. in the nucleus accumbens [4]. These effects are similar to those If the consumption of highly palatable foods is maintained produced by opiates and psychostimulants, although smaller in over time, it can lead to repeated release of endogenous opioids, magnitude. which induces hypersensitivity of the opioid receptor. This, in Recent evidence has shown that simple exposure to highly turn, may perpetuate binge eating of highly palatable foods. palatable food does not induce behavioral changes or neuronal Interestingly, when individuals addicted to exogenous opioids pathologic changes in the craving for carbohydrates and the are in withdrawal, they eat more sugary foods, probably to motivation system. Rather, repeated, intermittent, and excessive replace the action of opiates in the brain [5]. It has even been consumption of highly palatable food is required for these T. Ventura et al. / Nutrition 30 (2014) 252–256 255 changes to be established. It has even been seen that when low has been suggested that most children exhibit the natural palatability foods are eaten in an addictive manner, this had to biological response to distress (loss of appetite) because the have been primed with a highly palatable food [5]. normal distress response is associated with physiological re- This hypothesis implies that carbohydrate-craving treatment actions that are designed to prepare the individual for fight or should be similar to the treatment used in substance depen- flight, thereby suppressing the feelings of hunger and satiety. It dence, with the sole limitation being that in this case, it is has been suggested that emotional eating can be acquired as a impossible to completely withdraw from the dependent sub- result of inadequate parenting, which would have lasting effects stance [28]. on the neurobiologic response to stress. This may include a Critics of this hypothesis argue that it is not possible to draw hypoactivation of the HPA axis, with reversed neurovegetative a parallel between substance addiction and “supposed” food symptoms: hyperphagia and weight gain. These changes usually dependence. A disease model is different from a model of do not appear before puberty [18]. a complex physiological response that may eventually lead to A restrictive mother–child feeding style has been associated a somatic disease such as obesity [29]. with higher levels of consumption in the absence of hunger and with increased intake of restricted foods when free access to Stress response hypothesis them is permitted [32]. One study showed that carriers of the A1 allele of the dopa- The stress response hypothesis posits that glucocorticoids mine D2 receptor (DRD2) gene Taq1A polymorphism have and noradrenaline, which both participate in stress physiology, reduced availability of these receptors in the brain. Another increase their levels after the effect of a stressor, thus mediating study found that adolescents exposed to adverse experiences the relationship between anxiety and carbohydrate craving [30]. who develop emotional eating carry at least one DRD2 A1 allele. During acute stress, eating is suppressed due to the anorectic This suggests that the relationship between adverse experiences effect of the corticotropin-releasing hormone. However, this and the development of emotional eating can depend on a behavior increases during recovery due to the residual effects of genetic substrate [33]. The short (S) allele of the serotonin cortisol. Furthermore, cortisol, during acute stress, increases transporter gene (5-HTTLPR) also has been associated with this. available energy levels through lipolysis and gluconeogenesis. The S allele, compared with the long (L), can give a genotype that Additionally, as a mechanism to increase energy, glucocorticoids is more susceptible to stress and the consequent emotional produce hyperphagia, which ultimately leads to weight gain, as eating [34]. part of the pathophysiology of obesity [12]. The tendency to gain It has been observed that individuals who are both restricted weight over time has also been shown in chronic stress, which in their eating and vulnerable to uncontrolled eating are par- may be due to stress-related endocrine changes and behaviors ticularly susceptible to having negative emotions affect their associated with coping with the situation [12]. It also has been intake [35]. shown that individuals with high cortisol reactivity consume The gene–environment hypothesis places importance in the significantly more calories and sweet foods while under stress management of obesity to upbringing, emotional education, and than those with low reactivity of this hormone [10]. the teaching of coping mechanisms, especially in individuals A study in female rhesus monkeys continuously exposed to with genetic vulnerability. stressors verified that they consumed more carbohydrate calories compared with controls, which was associated with a flat- Discussion tening of the diurnal cortisol rhythm and a greater increase in serum cortisol when faced with acute separation. These data The results of this review, although only focused on carbo- support the hypothesis that daily exposure to psychosocial stress hydrate craving, allow us to begin to understand the neuro- increases the consumption of high-calorie foods [31]. biologic mechanisms that individuals and clinicians must Additionally, it has been found that acute stress reduces food overcome when attempting to achieve behavioral changes to intake unless there is access to highly palatable food during the address the obesity epidemic [1,2]. stressful period. This “non-homeostatic eating” has been hy- The first group of studies, which focus on the role of the pothesized to activate the motivational system and decrease the serotonergic pathways, link mood levels with a disposition to activity of the HPA axis, controlling the stress response, as evi- seek carbohydrates [11,12]. These studies also imply a lapse in denced by lower levels of cortisol. That is, the stress activates time between carbohydrate consumption and the effects on dopaminergic neurons and the stress response is controlled after mood that are inconsistent with reports that symptoms of anx- eating highly palatable food [30]. This is what enables con- iety subside [17,18]. In short, although the serotonergic hypoth- sumption of high palatability foods to continue over time, giving esis is attractive because it relates the consumption of rise to obesity. carbohydrates with an appetite for carbohydrates and with some A proper understanding of the mechanisms of the interaction psychiatric conditions, it does not allow for a proper under- between the stress system and the motivation system will enable standing of the carbohydrate-craving phenomenon. the eventual development of new therapies to combat obesity Alternatively, the association of mechanisms underlying pal- [30]. atability with an intense hedonic response that reinforces the motivational systems is in agreement with the rapid changes in Gene–environment hypothesis emotional symptoms observed after consumption of carbohydrates [16,17,19]. Such association also may be compatible with The gene–environment hypothesis suggests that inadequate the eventual reinforcement of sugar consumption, which would parenting in managing emotions generates carbohydrate craving lead to cravings and potentially addiction [5,6]. It must be added in adults as part of the coping response to states of anxiety. This that the normal response to stress further enhances the hedonic phenomenon is more common in genetically vulnerable adults. response following carbohydrate consumption, enhancing be- The distress that induces eating is highly prevalent in obese havioral reinforcement under conditions of stress [12,30,33].In adults but has a low prevalence in young children. That is why it other words, it is necessary to consider at least these lines of 256 T. Ventura et al. / Nutrition 30 (2014) 252–256 research: palatability and hedonic response, motivational sys- [6] Parylak SL, Koob GF, Zorrilla EP. The dark side of food addiction. Physiol – tems, and stress systems, in order to gain further understanding Behav 2011;104:149 56. [7] Christensen L, Pettijohn L. Mood and carbohydrate cravings. Appetite of the complexity of the system. 2001;36:137–45. On the one hand, it also must be taken into account that [8] Corsica JA, Spring BJ. Carbohydrate craving: a double-blind, placebo – human have evolved in an environment with a relative scarcity controlled test of the self-medication hypothesis. Eat Behav 2008;9:447 54. [9] Konttinen H, Mannisto S, Sarlio-Lahteenkorva S, Silventoinen K, of resources [2,36], where obtaining vital resources is a key Haukkala A. Emotional eating, depressive symptoms and self-reported food objective. In this regard, identifying the most efficient nutrients consumption. A population-based study. Appetite 2010;54:473–9. is a fundamental adaptive capacity [2,36]. Their consumption [10] Lieberman H, Wurtman J, Chew B. Changes in mood after carbohydrate consumption among obese individuals. Am J Clin Nutr 1986;44:772–8. triggers a pleasurable response, which, by acting on motivational [11] Beulens JWJ, Bindels JG, De Graaf C, Alles MS, Wouters-Wesseling W. Alpha- systems, will seek repetition, as with everything else that is lactalbumin combined with a regular diet increases plasma Trp-LNAA ratio. biologically desirable. Thus, carbohydrate craving appears to be Physiol Behav 2004;81:585–93. [12] Takeda E, Terao J, Nakaya Y, Miyamoto K, Baba Y, Chuman H, et al. Stress part of a biologically conducive system. If we add stress response, control and human nutrition. J Med Invest 2004;51:139–45. which also is evolutionarily selected through phylogeny, the [13] Markus CR. Dietary amino acids and brain serotonin function; Implications power of this aforementioned circuit increases, probably because for stress-related affective changes. Neuromol Med 2008;10:247–58. of the possibility of quickly storing available energy resources in [14] Orosco M, Rouch C, Beslot F, Feurte S, Regnault A, Daude V. Alpha-lactalbumin-enriched diets enhance serotonin release and induce anxiolytic and order to face aversive stimuli [2,36]. Consequently, we can posit rewarding effects in the rat. Behav Brain Res 2004;148:1–10. that the carbohydrate-craving mechanism for maintaining [15] Markus CR, Olivier B, Panhuysen GE, Van Der Gugten J, Alles MS, Tuiten A, emotional homeostasis could have been evolutionarily selected et al. The bovine protein alpha-lactalbumin increases the plasma ratio of tryptophan to the other large neutral amino acids, and in vulnerable sub- due to its environmentally adaptive functions. jects raises brain serotonin activity, reduces cortisol concentration, and Of course, there are certain genes that might increase improves mood under stress. Am J Clin Nutr 2000;71:1536–44. vulnerability to this phenomenon. For example, it could be [16] Lemmens SG, Martens EA, Born JM, Martes MJ, Westerterp-Plantenga MS. Lack of effect of high protein vs. high-carbohydrate meal intake on stress- suggested that individuals with higher reactivity to cortisol or related mood and eating behavior. Nutr J 2011;10:136. the carriers of the A1 allele of the dopamine DRD2 receptor [33] [17] Macht M, Mueller J. Immediate effects of chocolate on experimentally or the S allele of the serotonin transporter gene [34] could be induced mood states. Appetite 2007;49:667–74. [18] Diksic M, Nagahiro S, Sourkes TL, Yamamoto YL. A New method to measure more vulnerable to presenting a craving for carbohydrates, this brain serotonin synthesis in vivo. I. Theory and basic data for a biological effect being amplified in the presence of psychiatric disorders. model. J Cereb Blood Flow Metab 1990;10:1–12. As a result of this review, it is clear that carbohydrate craving [19] Kampov-Polevov AB, Alterman A, Khalitov E, Garbutt JC. Sweet preference predicts mood altering effect of and impaired control over eating sweet involves systems that were neurobiologically selected thr- foods. Eat Behav 2006;7:181–7. oughout our evolutionary history to sustain life. It is for this [20] Iemolo A, Valenza M, Tozier L, Knapp CM, Kornetsky C, Steardo L, et al. reason that confronting this phenomenon will require much Withdrawal from chronic, intermittent access to a highly palatable food more from clinicians and public health specialists than mere induces depressive-like behavior in compulsive eating rats. Behav Phar- macol 2012;23:593–602. indications or restrictions that will be difficult to sustain in an [21] Rao RE, Wojnicki FH, Coupland J, Ghosh S, Corwin RL. Baclofen, raclopride environment of a plentiful food supply. Long-term education and naltrexone differentially reduce solid fat emulsion intake under limited – campaigns could be required to encourage people to change access conditions. Pharmacol Biochem Behav 2008;89:581 90. [22] Havermans RC. “You Say it’s Liking, I Say it’s Wanting .”. On the difficulty their habits, creating a culture that is able to modulate or modify of disentangling food reward in man. Appetite 2011;57:286–94. people’s spontaneous behavior [1,2,28]. Moreover, research [23] Kalivas PW, Volkow ND. The neural basis of addiction: a pathology of – opens the door to the use of pharmacologic strategies both to motivation and choice. Am J Psychiatry 2005;162:1403 13. [24] McGeea HM, Amarea B, Bennettb AL, Duncan-Vaidyaa EA. Behavioral ef- trick palatability systems, with carbohydrate substitutes, for fects of withdrawal from sweetened vegetable shortening in rats. Brain example, and to decrease hedonic response, acting on opioid or Research 2010;1350:103–11. other pathways [5,21]. [25] Olszewski P, Levine A. Central opioids and consumption of sweet tastants: When reward outweighs homeostasis. Physiol Behav 2007;91:506–12. We understand that this review has a number of limitations; [26] Arce M, Michopoulos V, Shepard KN, Ha QC, Wilson ME. Diet choice, for example, it does not cover other foods that have been shown to cortisol reactivity, and emotional feeding in socially housed rhesus mon- be the subject of potential abuse and dependence, like fat and salt keys. Physiol Behav 2010;101:446–55. [27] Trinko R, Sears R, Guarnieri D, DiLeone R. Neural mechanisms underlying [6]. Although research in this area is under development, it has obesity and drug addiction. Physiol Behav 2007;91:499–505. been possible to illustrate the complexity of the carbohydrate- [28] Davis C, Carter JC. Compulsive overeating as an addiction disorder. A re- craving phenomenon and the potential neurobiologic barriers to view of theory and evidence. Appetite 2009;53:1–8. overcome in order to achieve substantial changes in people’s [29] DiLeone RJ, Taylor JR, Picciotto MR. The drive to eat: comparisons and distinctions between mechanisms of food reward and drug addiction. nutritional behavior. We hope that future advances in the study of Nature Neuroscience 2012;15:1330–5. neuroscience associated with carbohydrate craving may help [30] Maniam J, Morris MJ. The link between stress and feeding behavior. – clarify the hypotheses described here because they are comple- Neuropharmacology 2012;63:97 110. fi [31] Yen JY, Chang SJ, Ko CH, Yen CF, Chen CS, Yeh YC, et al. The high-sweet-fat mentary. Therefore, new ndings should offer greater coherence food craving among women with premenstrual dysphoric disorder: to the models currently under development. Emotional response, implicit attitude and rewards sensitivity. Psycho- neuroendocrinology 2010;35:1203–12. References [32] Brown A, Lee M. Maternal child-feeding style during the weaning period: Association with infant weight and maternal eating style. Eat Behav 2011;12:108–11. [1] World Health Organization. Global health risks. Geneva, Switzerland: [33] Van Strien T, Snoek HM, van der Zwaluw CS, Engels RCME. Parental control World Health Organization; 2009. and the dopamine D2 receptor gene (DRD2) interaction on emotional [2] Lieberman LS. Evolutionary and anthropological perspectives on optimal eating in adolescence. Appetite 2010;54:255–61. foraging in obesogenic environments. Appetite 2006;47:3–9. [34] Caspi A, Hariri A, Holmes A, Uher R, Moffitt T. Genetic sensitivity to the [3] Story M, French SA. Food advertising and marketing directed at children enviroment: the case of the serotonin transporter gene and its implications and adolescents in the US. Int J Behav Nutr Phys Act 2004;1:1–21. for studying complex diseases and traits. Am J Psychiatry 2010;167:509–27. [4] Avena NM, Rada P, Hoebel BG. Evidence for sugar addiction: behavioral and [35] Fay SH, Finlayson G. Negative affect-induced food intake in non-dieting neurochemical effects of intermittent, excessive sugar intake. Neurosci women is reward driven and associated with restrained–disinhibited Biobehav Rev 2008;32:20–39. eating subtype. Appetite 2011;56:682–8. [5] Corwin RL, Avena NM, Boggiano MM. Feeding and reward: perspectives [36] Berthoud H. Interactions between the “cognitive” and “metabolic” brain in from three rat models in binge eating. Physiol Behav 2011;104:87–97. the control of food intake. Psychoneuroendocrinology 2007;91:486–98.