Relationships Among the Brain, the Digestive System, and Eating Behavior: Workshop Summary (2015)

4

Future Directions: Is the Addiction Model for Drugs and Alcohol Appropriate for Food?

Moderator Joseph Levitt¹ opened the third session of the workshop by highlighting the popularity of the term “food addiction” not only in the media but also in the medical community. The notion that some foods may have addictive qualities now occupies a space next to low fat, low carb, and gluten-free in the American search for simple solutions to the obesity crisis, Levitt said. He observed that during the Korean War, malnutrition was the number one medical reason for rejection from military service; today, more than half a century later, the number one medical reason for rejection from military service is obesity (Christeson et al., 2010). “That is an enormous change,” Levitt said. But the question raised by Eric Decker² earlier during his introductory remarks needs to be addressed: Is addiction the proper model for examining overeating, overweight, and obesity?

Throughout the workshop, participants expressed varying opinions about how to interpret existing evidence for addictive-like eating behavior from neuroimaging studies based on criteria in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV). Some participants argued that existing evidence suggests that at least some eating behaviors, namely binge eating, can be characterized as addictive, while others opined that the data in this regard are inconclusive (see Chapter 3). The third session of the workshop, summarized here, comprised two counterpoint

______________

¹ Joseph Levitt, J.D., is with Hogan Lovells US LLP, Washington, DC.

² Eric Decker, Ph.D., University of Massachusetts Amherst, is a member of the Food Forum and was chair of the workshop planning committee.

presentations on the appropriateness of applying the addiction model for drugs and alcohol to eating behavior. Box 4-1 highlights key points made by speakers during this session.

THE ADDICTION MODEL IS APPROPRIATE FOR USE WITH FOOD³

Echoing what several other participants had previously said or implied, Nicole Avena identified obesity as “one of the main reasons why we are all here today.” Researchers have been trying for some time, she said, to understand how the concept of “food addiction” might play into the many factors contributing to the rising obesity epidemic. After providing an overview of factors contributing to obesity, she described her own research, along with other supporting studies, on the characterization of food addiction using a rat model. She noted that her work in the field extends back to when she was a graduate student with Bart Hoebel at Princeton University.

______________

³ This section summarizes the presentation of Nicole Avena, Ph.D., Mount Sinai School of Medicine, New York, New York.

Why Are So Many People Overweight or Obese?

A majority of the U.S. population is overweight or obese. Not only are overweight and obesity associated with multiple comorbid health concerns, such as heart disease and diabetes, but they also have significant psychological, social, and economic consequences (IOM, 2012; Tsai et al., 2011). But why, Avena asked, are so many people overweight and obese despite all the education and warnings they receive? Why do so many people have such a difficult time regulating their body weight? Avena emphasized that obesity is an endpoint, with no singular cause and many factors at play. For example, portion sizes have increased over the past several decades. Also, acquiring food today is much easier than it was for our hunter-gatherer ancestors. Other contributing factors include today’s sedentary lifestyles, genetic vulnerability, social norms regarding food, and stress and endocrine factors. Avena’s research program is focused on the taste of food and how food as reward may also be contributing to the obesity epidemic.

What Is a Food?

Avena raised the question: What is a food? She showed images of nutrition labels for two “foods”—one obviously baby carrots, with the only listed ingredient being “whole baby carrots, frozen, unprepared,” and the other unidentifiable by its long list of ingredients. The latter turned out to be Pop-Tarts. Avena suggested that before discussing how to determine whether certain foods can meet the criteria for being “addictive,” the term “food” may need to be defined or characterized differently from how it currently is.

Avena also suggested that it is necessary to consider what is meant by “eating.” There are multiple reasons why people eat—for hunger but also for hedonic purposes, that is, not because they are hungry but because they derive pleasure from eating or tasting a food. Avena clarified that when she describes research on “food addiction,” she is referring to hedonically driven eating. The types of foods often eaten for hedonic purposes tend to be hyperpalatable, ultra-processed foods that contain many added sugars, fats, and other ingredients (Monteiro et al., 2011).

There are pathways in the brain that reinforce natural behaviors, such as sexual and feeding behaviors, causing individuals to engage in the behaviors repeatedly. Drugs of abuse activate those same pathways, according to Avena. Highly palatable foods, she said, activate brain reward systems beyond what is seen when healthy food is eaten (e.g., with rats, their rat chow), and instead act in ways that could potentially be putting these systems into “overdrive” (Avena and Gold, 2011). Given the overlaps in neural circuitry associated with eating and with drug use and abuse, the

empirical question Avena and her research team have been asking for many years is: Could some people be “addicted” to eating highly palatable foods rich in sugar and fat in ways that resemble drug addiction, and could such out-of-control eating result in increased body weight and obesity in some individuals?

Empirical Studies of Food Addiction Using Animal Models

Avena’s research involves the use of animal models to determine whether DSM-IV and DSM-5 criteria for substance use disorders (APA, 2000, 2013) apply when the substance in question is a highly palatable food instead of a drug. That is, instead of giving animals drugs, she and her research team give them delicious foods to eat. Avena emphasized that not all of the DSM-IV or DSM-5 criteria need to be met for diagnosis of a substance use disorder. Also of note, Avena was drawn to the use of animal models because they allow researchers to examine biological correlates of behavior often impossible to study in humans.

Avena started her inquiry into food addiction by looking at sugar (Avena et al., 2008b). She chose sugar for several reasons, including the fact that Americans consume on average 22 teaspoons of added sugar daily (NCI, 2010). Studies suggest that sugar appears to be one of the ingredients many people find particularly problematic, such that they have difficulty regulating the intake of foods rich in sugar when they try to cut back. There also have been several studies finding a correlation between sugar intake and obesity. Thus, for Avena, sugar appeared to be a good ingredient to examine first.

In one study conducted by Avena’s group (Rada et al., 2005), rats were given access to standard rat chow plus a sugar solution for 12 hours a day for 21 days and were observed to drink more sugar over time. By the end of the 3-week period, these rats were bingeing on the sugar solution and showed evidence of tolerance as they were consuming more and more each day, which suggests that an increased amount was needed to achieve the same effect (Rada et al., 2005). This bingeing behavior was particularly apparent during the first hour of access to the sugar solution following the 12-hour period of abstinence. Of interest, Avena noted, rats provided with chow and sugar ad libitum (one of the control groups) did not show this escalation in daily intake. Other control groups included rats that had access to sugar only on days 1, 2, and 21 of the experiment and rats that had 12-hour daily access to chow only (no sugar). It was only the rats that were bingeing on the sugar daily that showed increased intake over time.

Avena and her team were curious about whether the overconsuming rats in the test group were releasing dopamine in a way that was consistent with consuming a food or a drug. A hallmark of drugs of abuse is that they

can cause a release of dopamine in reward-related regions of the brain, such as the nucleus accumbens, every time they are administered. Food also can cause the release of dopamine, according to Avena, but the dopamine release normally wanes when the food is no longer novel and an individual becomes habituated to it. She and her team found that, indeed, rats that were overconsuming the sugar solution were releasing dopamine every time they had access to the sugar. This was as true on day 21 as it was on day 2. Rats in the ad libitum control group and the control group that drank the sugar solution only occasionally, on the other hand, did not show the same release of dopamine over time, nor did the rats that consumed chow and no sugar (Rada et al., 2005). According to Avena, these results suggest that there is something about sugar such that when rats overconsume it, they release dopamine in a drug-like way.

Other studies have shown that rats that overconsume sugar show physical signs of withdrawal, distress, and anxiety when the sugar is taken away or when they are administered an opioid antagonist, which blocks opioid receptors in the brain (Avena et al., 2008b; Colantuoni et al., 2001). Additionally, the reward-related regions of their brains show a decrease in dopamine levels, coupled with an increase in acetylcholine (Avena et al., 2008a). According to Avena, a similar dopamine-acetylcholine imbalance has been seen during withdrawal from many drugs of abuse, including cocaine, nicotine, and morphine.

Craving is a difficult behavior to assess with animal models, in Avena’s opinion. She considers it a highly subjective, psychological characteristic. Nonetheless, several researchers have assessed features of craving in rats in an effort to understand its biological basis (Avena et al., 2005; Grimm et al., 2005; Krasnova et al., 2014; Oswald et al., 2011). Oswald and colleagues (2011) provided M&Ms to rats that were either prone or resistant to binge eating, the catch being that the rats had to cross an electrified shock grid to get to the M&Ms. The researchers found that the rats prone to binge eating endured greater magnitudes of shock to obtain the treat relative to their binge-resistant counterparts. Other studies have shown that following an abstinence period, rats prone to bingeing on sugar increase their intake of sugar when it is made available (Avena et al., 2005) and work harder to gain access to sugar-associated cues (Grimm et al., 2005).

Avena and her team also studied cross-sensitization between overconsumption of sugar and drugs of abuse (Avena and Hoebel, 2003; Avena et al., 2004). They found that animals with a history of overeating sugar became hyperactive when administered a very low dose of amphetamine, a potent dopamine agonist, instead (Avena and Hoebel, 2003). Animals without a history of overeating sugar, on the other hand, did not show the same hyperactivity in response to the same dose of amphetamine. These results suggest to Avena that there is something about sugar consumption,

presumably the effect it is having on the dopamine system, that causes even a very low dose of amphetamine to have this effect. Greater evidence of cross-sensitization was observed when rats with a history of overconsuming sugar were provided with alcohol instead of sugar; they drank more alcohol than did control rats that were exposed to sugar, but did not overconsume it (Avena et al., 2004).

Avena mentioned that many other food addiction studies have focused on sugar-fat combinations and combinations of other foods. Many researchers have used as a test condition the cafeteria diet, whereby rats have been provided a wide variety of high-fat, high-sugar foods, as well as healthy foods. For example, Geiger and colleagues (2009) found that animals on a cafeteria diet became overweight or obese and when provided amphetamine in vivo or in vitro, released much more dopamine than chow-fed control rats. Moreover, when their cafeteria diet was replaced with regular lab chow, the rats did not show an increase in release of dopamine. Only when their cafeteria diet was reintroduced did these rats again show an increase in dopamine release. According to Avena, these results suggest that the cafeteria diet had changed the rats’ brains in a way that was similar to what is seen in rats that overeat sugar and that caused the animals to react to healthy food differently from the way the rats maintained on a healthy diet reacted.

Issues to Consider

Since this was a debate, Avena identified several issues raised by critics of this work to consider, or reconsider, in moving forward. First is the notion put forth be some experts that the construct of food addiction is “distracting” and diverts attention from the main causes of overeating and obesity (Rogers, 2013). According to those critics, obesity is better viewed as being due to a “toxic” environment (Rogers, 1999). Avena said, “I couldn’t agree more that a toxic environment is certainly a big part of this.” That said, she reiterated that overeating and obesity have multiple contributing factors and that research on “food addiction” does not preclude the legitimacy of any other factor; thus, there is no reason to consider it “distracting.”

Additionally, critics have noted that the construct of food addiction may be fitting for individuals with binge eating disorder but not for understanding obesity. That is a valid criticism, in Avena’s opinion, given that much of the laboratory work done to date has aligned food addiction with binge eating. To further understand how the food addiction construct may be helpful for understanding obesity, Avena reminded the audience that care is necessary in defining addiction. Usually when people think about addiction, they think about an extreme loss of control (Altman et al., 1996). But

quoting Rogers (2013), Avena stated that while an extreme loss of control may characterize binge eating, it does not describe well the repeated failure to resist energy-rich foods in large portions that gradually contribute to weight gain. Avena remarked that many people think of the typical addict as someone lying in a gutter with no job and no family. But in reality, the typical addict in American society is a mom driving her kids to soccer practice and smoking cigarettes—someone who is likely a fully functioning individual for whom withdrawal syndrome is not physically life-threatening. Avena suspects that addiction to palatable foods may be more like addiction to cigarettes than to drugs of abuse, and thereby produce the same type of milder and less pronounced loss of control that is associated with smoking.

Finally, some critics argue that stigma may be conferred when a person is diagnosed as a “food addict.” Again, this is a valid concern in Avena’s opinion, but scientific data suggest it is unfounded. She pointed out that recent studies suggest that the stigma conferred by being labeled a “food addict” is no greater and may even be less than that associated with being labeled “obese” (DePierre et al., 2013; Latner et al., 2014).

THE ADDICTION MODEL IS NOT APPROPRIATE FOR USE WITH FOOD⁴

Peter Rogers’s interest in the appropriateness of the addiction model for use with food stems from his interest in understanding human appetite and weight control and his work in caffeine psychopharmacology. During his talk, he questioned the definition of addiction, evidence for “food addiction” from animal and human studies, and the usefulness of “food addiction” in explaining and reducing overeating. But first, he noted that in his opinion, caffeine illustrates very well the distinction between dependence and addiction. Most people who consume caffeine—who he suggested represent the majority of people on the planet—are dependent on it. If they become tolerant to its psychostimulant effects and it is withdrawn, they become fatigued and tired. But few people who consume caffeine experience the extreme loss of control that is characteristic of addiction.

Also to set the stage for his talk, Rogers displayed a headline from a June 2013 edition of the Metro (UK): “Potatoes Give You ‘Drug Fix.’” The article read: “You might not have to shoot . . . it up to get a fix. But food is just as addictive as heroin and nicotine, research suggests. Substance abuse and high-glycemic foods—such as white bread and potatoes—trigger the same brain mechanism as that linked to addiction, according to Boston Children’s Hospital. They apparently cause excess hunger and stimulate

______________

⁴ This section summarizes the presentation of Peter Rogers, Ph.D., M.Sc., B.Sc., University of Bristol, Bristol, United Kingdom.

reward and craving in parts of the brain.” Rogers questioned whether those ideas and that sort of reporting help encourage healthier eating or weight loss.

What Is Addiction?

There are several ways to define addiction, noted Rogers. The Concise Oxford English Dictionary (Stevenson and Waite, 2011) defines it as a “a fact or condition of being physically dependent on a particular substance,” a definition that Rogers observed is aligned with how other workshop speakers had defined it. Other definitions are milder and imply something that someone simply likes to do frequently or is particularly interested in. For example, when people say they are “addicted” to soap operas, they mean that watching soap operas occupies a large amount of their time. And people who call themselves “chocoholics” usually are communicating that they probably eat more chocolate than they would like to rather than indicating a serious problem with that substance.

Rogers’s preferred definition of (drug) addiction comes from Altman and colleagues (1996): “Addiction is restricted to the extreme or psycho-pathological state where control over drug use is lost.” He emphasized the importance of “extreme state” and “loss of control.” Additionally, Altman and colleagues (1996) define (drug) dependence as “the state of needing a drug to function within normal limits; it is often associated with tolerance and withdrawal [symptoms], and with addiction as defined above.” They state, “Tolerance, sensitization, withdrawal and craving are phenomena that may accompany dependence.” Rogers emphasized that dependence and craving, while associated with addiction, are not necessary for addiction.

A Counterargument to the Case for Food Addiction

As a “scaffolding” for his talk, Rogers referred to Gearhardt and colleagues’ (2011a)⁵ synopsis of the case for food addiction (emphasis added by Rogers):

______________

⁵ Reprinted with permission. Gearhardt, A. N., C. M. Grilo, R. J. DiLeone, K. D. Brownell, and M. N. Potenza. Can food be addictive? Public health and policy implications. Addiction 106(7):1208-1212. Copyright © 2011. [1] = Volkow et al., 2008; [2] = Blumenthal et al., 2010; [3] = Avena et al., 2008b; [4] = Johnson and Kenny, 2010.

Rogers identified several key points of this summary worth reconsidering (i.e., the points in italics above), one of them being that foods share multiple features with addictive drugs. That argument, he explained, is based on an observed overlap between the brain mechanisms and behavioral processes involved in eating and those involved in psychoactive drug use or abuse. In Rogers’s opinion, that overlap by itself is not evidence of addiction; it merely shows that drugs of abuse have engaged some of the same mechanisms engaged by eating. It is often argued further that addictive drugs “hijack” those mechanisms, the implication being that drugs of abuse have particularly potent effects on those mechanisms and that foods have less potent effects. In Rogers’s opinion, that same point could be used to argue that foods therefore pose a relatively low risk of addiction.

Another argument put forth for the case of food addiction, one based on animal evidence, is that consumption of certain foods—those high in sugar and fat—causes reward dysfunction and sets in motion a vicious cycle of further overeating. That is, the changes that occur in the reward pathways of the brain lead to overconsumption, which leads to further brain changes, and so on. Again, the argument goes, there is a parallel with the effects of addictive drugs. The argument is based on evidence such as that reported by Johnson and Kenny (2010), who showed that rats exposed to a cafeteria diet of chocolate, pound cake, sugar frosting, and a variety of other energy-dense foods experienced increased body weight compared with rats fed a standard laboratory diet. Additionally, by implanting electrodes in the rats’ brains that delivered rewarding stimulation when the rats pressed a lever, the researchers found that the rats on the cafeteria diet had a higher current threshold; that is, they experienced less reward for the same amount of stimulation compared with rats fed standard lab chow. The researchers interpreted their results as evidence of brain dysfunction.

In Rogers’s opinion, there is an alternative to Johnson and Kenny’s (2010) “vicious cycle” conclusion. He suggested that the higher current threshold in the rats fed the cafeteria diet was a “positive adaptation” to limit further weight gain; that is, as the rats gained weight, they became less interested in eating. He traced his alternative explanation back to a 1983 study in which he offered rats either a variety of energy-dense foods or fat,

specifically lard, along with their standard diet (Rogers, 1985). Initially, rats in both groups showed an increase in body weight. Over time, however, the rats’ food intake decreased and their rate of weight gain plateaued. Rogers argued that the increased weight gain produced a negative feedback effect on appetite. When the rats were returned to their chow diet, they underate. Rogers interpreted their undereating to mean that the rats were no longer being stimulated by energy-dense palatable foods. While not discounting the significance of what he described as “our toxic obesogenic environment,” Rogers identified energy density as a major contributor to overconsumption, with energy-dense foods being less satiating on a calorie-for-calorie basis than energy-dilute foods and therefore more rewarding (Ledikwe et al., 2006; Stubbs et al., 1995).

Other studies, such as that of Epstein and Shaham (2010), have shown that rats exposed to addictive drugs experience a reduced brain stimulation reward similar to that experienced by the rats on a cafeteria diet in Johnson and Kenny’s (2010) study. Although the extended access to drugs caused a similar progressive disruption of the brain reward system, the reduction in brain stimulation reward declined rather rapidly when the drugs were withdrawn. That was not the case with the Johnson and Kenny (2010) rats fed a cafeteria diet, whose “reward dysfunction” persisted for a long time.

In summary, in Rogers’s opinion, Johnson and Kenny’s (2010) conclusions can be rewritten in a more positive way (see Figure 4-1). Their

FIGURE 4-1 Conclusions of Johnson and Kenny (2010), as reworked by Peter Rogers.
NOTE: Reprinted with permission from MacMillan Publishers Ltd.: Nature Neuroscience (13[5]:635-641), Copyright (2010).

observations, he said, do not point to dysfunction. Rather, they point to a functioning system whereby increased weight causes changes that oppose further weight gain.

Based on these animal studies and some parallel observations in humans, the case has been made that deficits in food reward mechanisms underlie human obesity and that individuals increase their consumption of palatable foods in an attempt to overcome that loss of food reward (Gearhardt et al., 2011b). According to that argument, one would expect to see increased food consumption in response to decreased brain dopamine. However, in a human study that involved reducing brain dopamine function by depleting the amino acid precursors to dopamine—tyrosine and phenylalanine—from the test subjects’ diets, Hardman and colleagues (2012) demonstrated the reverse: reduced dopamine function was associated with, if anything, a decrease in consumption of palatable food (p < 0.06). In Rogers’s opinion, these results further support the idea that the changes in brain reward being observed in animal studies reflect adaptive changes that counter overeating.

Another argument for food addiction is the common occurrence of reported food cravings. Because of the association between craving and addiction, it has been suggested that perhaps an addictive process underlies those cravings (Gearhardt et al., 2011b). In Rogers’s opinion, however, craving is a normal part of the eating experience and reflects the result of an attempt to resist eating certain foods (Rogers and Smit, 2000). He explained that it is normal for people to develop an appetite for a food, such as chocolate, as a result of being reminded of chocolate, being in a place where they recently consumed chocolate, being in a mood similar to their mood the last time they consumed chocolate, or otherwise experiencing something associated with chocolate. But if people who develop an appetite for chocolate deny themselves the chocolate—for example, because they have gained a few pounds recently—the thought of eating chocolate does not necessarily go away. If anything, the thought becomes more elaborated to the point where it could be called a craving. Rather than a craving being related to the food being craved and a cause of eating, Rogers views it as a consequence of restraint.

Many of these counterarguments, Rogers said, can be linked to some of John Davies’s (1997) ideas in The Myth of Addiction. While some readers may regard Davies’s book as an extreme view of addiction—his essential argument being that addiction is a social construct—Rogers opined that the book makes for compelling reading. Among other arguments, Davies proposes that the idea of addiction may actually be unhelpful in relation to trying to change behavior.

Attribution of Food Addiction: Helpful or Unhelpful?

It has been argued that one way in which the attribution, or label, of food addiction may be helpful is by reducing the stigma of overweight and obesity and encouraging more support for obesity treatments. There is some evidence that this is indeed the case (Hoyt et al., 2014; Latner et al., 2014; Lund et al., 2011). For example, Hoyt and colleagues (2014) found that being exposed to the notion that obesity is a disease helped reduce body image dissatisfaction. However, the researchers also found that the same individuals expressed reduced concern about weight and higher-calorie food choices. Such results suggest to Rogers that while the attribution of food addiction may reduce stigma, it also diminishes personal responsibility and motivation to change (Ogden and Wardle, 1990).

Rogers and colleagues have themselves collected preliminary data on the effects of attribution of food addiction. They provided participants with a variety of passages either confirming or disconfirming the existence of food addiction, but told the participants to read the passages and comment on the font type, writing, and other features not related to the content. Then they provided the participants with high-fat cookies, crisps, breadsticks, and grapes. They found that participants who had been primed with the idea that food addiction is a valid construct on average ate a little more food but mainly showed more variable intake than participants who had not been primed with that idea.

In summary, in Rogers’s opinion, attributing overeating to food addiction may be counterproductive, at least for some people, with respect to successful eating control.

Food Addiction and Obesity

Rogers echoed the remarks of other workshop participants that most people who are obese do not display addictive-like eating behavior. In a study of the relationship between weight status and food addiction as defined by the Yale Food Addiction Scale (YFAS),⁶ a little over one-third of obese individuals met the YFAS criteria for food addiction (Meule, 2011). Another study found that only about 8 percent of obese individuals met the criteria (Pedram et al., 2013). Rogers suggested that perhaps the best case for food addiction is binge eating. But again, whether binge eating is a “food addiction” depends on the definition of addiction. In a study of 79 women with a diagnosis of binge eating disorder, Cassin and von Ranson (2007) found that most of the women were “food-addicted” based on one set of criteria for addiction, but fewer than half were “food-

______________

⁶ See Chapter 3 for a discussion of the YFAS.

addicted” based on a more stringent set of criteria. Binge eating may look like addictive behavior, and there is a known association between binge eating and obesity, such that people with binge eating disorder are more likely to be obese (Pike et al., 2001). Nonetheless, the prevalence of binge eating disorder is much lower than the prevalence of overweight and obesity (Striegel-Moore and Franko, 2003).

DISCUSSION WITH THE AUDIENCE

Following Rogers’s talk, both speakers participated in a panel discussion with the audience. Questions spanned a range of topics.

Intermittent Feeding in the Rat Model: Relevance to Human Eating Behavior

In her presentation, Avena had discussed results of her research with animal models aimed at determining whether DSM-IV and DSM-5 criteria for substance use disorders apply when the substance of desire is a highly palatable food instead of a drug. Edmund Rolls asked her why intermittent access to food was an important part of her rat model. What is happening when food is removed? Is the animal becoming stressed? How relevant is that type of forced removal of food to humans? Would anti-anxiety drugs abolish the observed phenomena?

The intermittent access, Avena explained, is a sort of “limited extended access.” The rats have their food and sugar available for the majority of the time they are awake and active, but are unable, for example, to get up in the middle of the night and eat. Although many people tend to think of humans as living in an ad libitum food environment, with constant access to food, humans engage in self-limiting eating patterns in many ways. People tend to eat in meals, and many who are trying to lose weight restrict their intake. In Avena’s opinion, the intermittent pattern used in her research caused the animals to eat in a way similar to how people who are having problems losing weight may eat.

With respect to whether anti-anxiety drugs would reverse the pattern, Avena replied that she and her team have not studied this question. They have studied some other types of pharmacological compounds, such as baclofen (a muscle relaxant also shown by some studies to assist in the treatment of alcoholism) and baclofen in combination with naltrexone (an opioid receptor antagonist), which have been shown to mitigate the effects of overeating.

Avena was asked whether overall consumption for rats with intermittent access to food was any different from overall consumption for rats allowed to eat ad libitum. She replied that, yes, the rats with intermittent access to

food not only binged when food was reintroduced into their environment but also consumed more food over the entire 24-hour period compared with the rats with ad libitum access.

Danielle Greenberg commented that even in the absence of thirst, rats drink vast quantities of water when food is provided only intermittently. It is a phenomenon known as schedule-induced polydipsia. Greenberg asked Avena whether her interpretation of the rats’ sugar-bingeing behavior when provided intermittent access to food is valid given that the same thing happens with water. Greenberg replied that an important control would be an intermittent feeding schedule whereby the rats would be provided with water only. Without that control, it is impossible to know whether the overconsumption of sugar water observed by Avena and her team is any different from schedule-induced polydipsia. Avena replied that she and her team typically include a control group that has intermittent access to chow only (no sugar), and these animals do not show the addiction-like brain changes and behaviors seen when rats are overeating sugar. Thus, it does not appear that intermittent access alone is responsible for the observed effects.

Sensory-Specific Satiety: Its Relationship to Overeating and Obesity

When asked whether sensory-specific satiety contributes to obesity, Avena replied that it is unclear whether people who are obese or are overeating are experiencing dampened sensory-specific satiety or becoming satiated by one particular food and then switching to another. Rogers remarked that food variety contributes to overconsumption, although the extent of its role is not clear. In his opinion, overconsumption is possible even with a very narrow range of food variety. In fact, even a single energy-dense food can promote overconsumption, at least in animals.

Sensory Components of Foods and Addictive-Like Eating

The speakers were asked whether any research is under way to determine whether other aspects of foods besides their ingredients, such as their sensory and visual appeal, may contribute to addictive-like eating. For example, if one were to dye a hamburger and its bun green, would that dissuade some people from overeating? Avena replied that much of what is going on with addictive-like eating in humans is related to conditioning to food cues. Food cues serve, in many ways, to reinforce or encourage some eating behaviors. In fact, Avena sees this even in rats. After just a few days of access to highly palatable foods, rats learn when to expect food. They become conditioned to the researchers walking into the room to give them Cheez Doodles or M&Ms. Avena remarked that it is important to keep the effect of these cues on eating behaviors in mind when thinking about how

foods are packaged and marketed. It is possible, for example, that simply seeing a specific food or beverage package can affect the brain reward system in a way that primes someone to want to eat the food.

Rogers disagreed to some extent. However effectively cabbage is packaged, for example, he does not believe it could ever become a binge food. He stressed the importance of the characteristics of the foods themselves. In his opinion, energy density is key. Taste characteristics are important as well. Many foods people like tend to have sweet and salty tastes. Rogers views packaging as something that merely reminds people of the value of the food in terms of its energy density and taste.

Sugars Versus Artificial Sweeteners: Dopamine Release and Overeating

Avena was asked whether there is any evidence that her sugar model in rats is applicable to artificial sweeteners. She replied that artificial sweeteners are difficult to study in rats because rats do not taste them the same way that humans do. For example, rats cannot detect the taste of aspartame. However, Avena and her team have done relevant work with sugar sham feeding, whereby they provide the animals with a sugar-containing food that they can taste and ingest but not necessarily digest (because it exits through the stomach via a gastric cannula). This work has revealed that sweet taste alone is sufficient to elicit the dopamine release associated with the addiction-like behaviors. Of interest, Avena noted that other researchers have observed the same release of dopamine in reward regions of the brain when sugar is infused into the gut. She remarked that consuming caloric sugar produces a “double whammy” effect because dopamine is released in response to both the taste and postingestive effects of surgar, whereas artificial sweeteners stimulate only some dopamine release in response to the sweet taste.

When asked whether it would be possible to test the effects of artificial sweeteners in primates, Rogers commented that in some ways, such research has already been conducted. Many human studies have compared the appetite and long-term body weight effects of beverages (or foods) sweetened with artificial sweeteners and the effects of sugar-sweetened beverages (or foods) or water. In Rogers’s opinion, all of the evidence to date converges on the notion that “intense sweeteners” help people eat less and lose body weight. (Rogers noted that he prefers the term “intense sweetener” over “artificial sweetener,” because some so-called artificial sweeteners arguably are not artificial.)

Overeating, Obesity, and Socioeconomics

An audience member observed that people in communities of lower socioeconomic status, in which food is purchased at corner stores and not

in supermarkets where fresh foods are abundant, appear to develop an attachment to highly palatable foods and an apparent aversion to healthy foods. She asked whether there is a solution to this problem. Avena observed that she has seen much of this behavior in Harlem and noted that several organizations in New York City are making efforts to educate children about the importance of seeking out fruits, vegetables, and other healthy foods.

Rogers agreed that education is important and added that convenience and access are also parts of the problem. Many so-called healthy foods require more skill and equipment to prepare relative to many energy-dense foods. Rogers pointed to the replacement of french fries with apple slices in children’s meals by McDonald’s as a successful example of nudging eating behavior in a healthier direction. Fergus Clydesdale, University of Massachusetts Amherst, noted that some healthy foods are in fact available in frozen or canned form.

Another audience member observed that the higher rate of obesity in resource-poor communities is a much larger problem than a lack of education. Most of these communities are so stressed that it does them a disservice to “simply say that you can educate them out of this.” Avena reiterated the importance of education but also agreed that to think that education will be enough is “overly ambitious and overly hopeful.” She said, “We need to think beyond that and think outside the box in terms of how we might diversify some of our approaches.”

Rogers remarked that obesity shows the same socioeconomic patterning as other unhealthy behaviors, including drug use and alcoholism, and agreed that stress in certain communities may underlie those unhealthy behaviors. But there are other elements as well, in his opinion. He pointed to attitude as a potential problem, with studies in the United Kingdom showing that in some communities obesity is considered the norm and diabetes is considered an inevitable part of getting older. Thinking about those unhealthy conditions as normal or inevitable, in his view, probably undermines the idea that one can or should try to change one’s behavior. He suggested developing approaches to tackle those attitudes.

Pleasure Versus Addiction: Neuroimaging Evidence

A member of the webcast audience observed that listening to jazz music has been shown to elicit the same dopamine response in the striatal system as that activated by consumption of hedonically pleasing foods. He asked whether, given that observation, Avena was willing to make the argument that listening to jazz music is addictive. Or are current biological methods unable to distinguish between addiction and pleasure?

Avena replied that she did not think this observation warrants calling

listening to jazz an addiction and reiterated that no single feature defines a substance as addictive. While palatable foods are known to activate brain reward systems, so are many other pleasurable activities. But whether those pleasurable activities are addictive depends on how the reward systems are engaged; the extent to which they are engaged; and whether other concomitant features of addiction manifest, as has been shown with palatable foods. She said, “When we talk about addiction, we are talking about a multifaceted, multifeatured issue that we need to study from multiple angles.”

For Rogers, it is useful to think of the addiction risk posed by a substance. He cited a case report of “carrot addiction” in the Australian and New Zealand Journal of Psychiatry (Kaplan, 1996). In his opinion, carrots pose a low risk of addiction. Yet Rogers noted that, as this report demonstrates, it is possible to show addictive behavior in relation to eating carrots.

This page intentionally left blank.