The plausibility of sugar addiction and its role in obesity and eating disorders

Summary

Background & aims

To consider the hypothesis that addiction to food, or more specifically sucrose, plays a role in obesity and eating disorders.

Methods

By considering the relevant literature a series of predictions were examined, derived from the hypothesis that addiction to sucrose consumption can develop. Fasting should increase food cravings, predominantly for sweet items; cravings should occur after an overnight fast; the obese should find sweetness particularly attractive; a high-sugar consumption should predispose to obesity. More specifically predictions based on the hypothesis that addiction to sugar is central to bingeing disorders were developed. Dieting should predate the development of bingeing; dietary style rather than psychological, social and economic factors should be predispose to eating disorders; sweet items should be preferentially consumed while bingeing; opioid antagonists should cause withdrawal symptoms; bingeing should develop at a younger age when there is a greater preference for sweetness.

Results

The above predications have in common that on no occasion was the behaviour predicted by an animal model of sucrose addiction supported by human studies.

Conclusion

There is no support from the human literature for the hypothesis that sucrose may be physically addictive or that addiction to sugar plays a role in eating disorders.

Introduction

Summary

1. Following sugar consumption a rat model has demonstrated physiological and behavioural changes consistent with addiction, although it was consumed under a highly prescribed and atypical feeding procedure.

2. By analogy with the rat data it has been suggested that human obesity and binge eating might reflect an ‘addiction’ to sucrose consumption, a suggestion that relies to a great extent on the suggestion that physical dependence occurs in a similar manner to that observed with drugs of abuse.

3. The day to day food preference of rats reflects palatability rather than sucrose content.

There is a widespread assumption that sugar consumption can lead to addiction: putting the words addiction and sugar together into the Google search engine produced 769,000 hits. A sample comment that illustrates the type of view often expressed in popular literature is that “sugar addiction can be just as strong as a drug or alcohol dependency”. Although at one time such opinions were largely dismissed by the scientific community, more recently animal studies of bingeing on sugar1, 2 or fat3, 4 have reported findings that have been used to bolster the view that food can be addictive, although the conclusions drawn by the scientists themselves and those relying on their results can differ. Those who have carried out the research have tended to concentrate their attention on eating disorders,1, 2 whereas more popular writing can portray the phenomenon as widespread if not virtually universal.

If addiction to food can be established in humans there are widespread implications. Dieting might not be the optimal response to obesity as it will lead to counter-regulatory mechanisms such as cravings and withdrawal symptoms. In fact Trotzky⁵ has treated eating disorders as addictive diseases using the twelve step programme that is familiar when considering addiction to other substances. If an animal model that is a homology of the human condition can be demonstrated, it would provide the means of establishing the underlying basic biology with consequent opportunities to establish novel treatments. There are also potentially widespread implications for food manufacturers and the fast-food industry. At an annual symposium of the “Confectioners Association and Chocolate Manufacturers Association” Susan Smith the Senior vice-president commented that: “They are looking at the tobacco model, turning their sights on sugar the same way they did on tobacco”. In 2003 Professor Banzhaf of George Washington University wrote to Burger King pointing out the possibility of future legal action: stating that “… foods of the type served at your fast food restaurants may produce addictive like effects … research strongly suggests that … at least some fast foods can act on the brain the same way as nicotine and heroin”.

It is apparent that we need to establish the veracity of such claims. Given the importance that may be placed on the research that considers the development of food addiction in animals, the relevance of these findings to the human condition is presently considered. Although animal studies can generate hypotheses, they need to be confirmed in humans who in addition are influenced by a cultural and social environment that adds a complexity not seen with rodents. The plausibility that sugar addiction plays a role in food intake, obesity and eating disorders is therefore considered. The aim was to derive predictions from animal studies and then to establish the extent to which they are supported by human research. It is not the purpose to consider the animal research in detail although it is outlined initially to allow relevant predictions to be made.

Definitions are arbitrary and the answer as to whether sugar is additive may depend on how it is defined. For some, addiction is a pharmacological term characterized by a compulsion to consume that is driven by cravings. Tolerance occurs so that over time to achieve the same response you need to increase the dose. There is dependence so that there are withdrawal effects if consumption does not occur, making it difficult to quit. However, the term addiction has evolved leading some to designate this type of pharmacological definition as physical dependence.

Psychiatrists use the term dependence rather than addiction. The American Psychiatric Association criteria for the clinical diagnosis of abuse and dependence is a maladaptive pattern of substance use leading to clinically significant impairment or distress, as manifested by one (or more) of the following, occurring within a 12-month period:

• Recurrent substance use resulting in a failure to fulfil major role obligations at work, school, home

• Recurrent substance use in situations in which it is physically hazardous

• Recurrent substance-related legal problems

• Continued substance use despite having persistent or recurrent social or interpersonal problems caused or exacerbated by the effects of the substance. (Diagnostic and Statistical Manual-IV (DSM-IV))

At the other extreme a lay definition of addiction has developed that may amount to little more than giving a particular activity a very high priority. Never missing an episode of a favourite soap-opera might be so described. An element of compulsion is a further step with behaviour being clinically considered if the compulsion is uncontrolled, albeit there is no harm being suffered by the patient or others. Medically there is a distinction between physical dependence with characteristic withdrawal symptoms and psychological dependence that is “uncontrolled, compulsive use”, even if nobody is harmed. Although an easy solution would be to reserve the term addiction for use with drugs of abuse, some feel that psychological dependency on such things as work, gambling, sex, computers, exercise, shopping, pornography or religion should be included. For others palatable food in general and sugar in particular should be part of this list. In practice these types of addiction are not always easy to distinguish as both physical and psychological mechanisms can co-exist.

Although there are many ways in which addiction has been defined a way of proceeding is required. Pelchat⁶ noted that the majority of evidence relating to food addiction relies on the similarities and differences between food and drug cravings. The renewed interest in the possibility of sugar addiction relies to a great extent on the work of Hoebel1, 2 whose argument relies on drawing biological parallels between the response of the body to sugar and drugs of abuse. Therefore the present review will draw on physical dependency rather than the psychological definition of addiction or the use of clinical diagnostic criteria. This approach seems reasonable as they commented that “…‘Food addiction’ seems plausible because brain pathways that evolved to respond to natural rewards are also activated by addictive drugs”.¹ In this spirit ‘sucrose addiction’ is predicted to be associated with craving, tolerance and withdrawal symptoms. The present objective is limited to considering the view that sucrose consumption might result from physical addiction in a manner homologous to drugs of abuse.

Avena et al.¹ reviewed the evidence that, given an appropriate feeding pattern, it is possible to show sugar addiction in rats. The starting point for this line of research was that: “Many people claim that they feel compelled to eat sweet foods, similar in some ways to how an alcoholic might feel compelled to drink. Therefore, we developed an animal model to investigate why some people have difficulty moderating their intake of palatable foods”.

Rats were deprived of food for twelve hours and then given twelve hours access to food, starting four hours into the dark phase, when they could consume laboratory chow and depending on the study either a ten percent solution of sucrose or a twenty-five percent solution of glucose. After being kept on this schedule for a month the animals showed signs of addiction. During the first hour of access there was a large intake of sugar, a phenomenon described as a ‘binge’; withdrawal symptoms were displayed; in a similar manner to drugs of abuse dopamine was released in the nucleus accumbens; opioid antagonists produced withdrawal symptoms.¹ Thus there was a body of research that supported the suggestion that rats kept under this behavioural regime manifest physiological and behavioural changes consistent with an addiction to sugar.

It was proposed that this animal model potentially offered insight into various human disorders. “We suggest that sugar, as common as it is, nonetheless meets many of the criteria for a substance of abuse … The rise in obesity, coupled with the emergence of scientific findings of parallels between drugs of abuse and palatable foods has given credibility to this idea.” The feeding regimen of the rats “shares some aspects of the behavioural pattern in people diagnosed with binge-eating disorder or bulimia. Bulimics often restrict intake early in the day and then binge later in the evening, usually on palatable foods”.¹

The concept of ‘sugar addiction’¹ relies on rats given the choice between a palatable sucrose solution and a much less palatable chow. Naturally in such circumstances they consume sucrose. The question is whether it is sucrose, sweetness or palatability to which they are responding? It needs to be demonstrated that similar behaviour could not be demonstrated with carbohydrate in general, artificial sweeteners or fat-rich palatable foods. Comparisons have been made between the reaction of rats to the provision of sucrose, a high-fat diet and a sweet-fat combination.⁷ The ability of the opioid antagonist naloxone to produce withdrawal symptoms was not observed with fat although it occurred when sucrose alone was provided, evidence that in this paradigm different types of palatable food produce different responses.

However, it appears that rats do not have a preference for sucrose consumption as there is a preference for sucrose in sham feeding studies, where after passing through the mouth it leaves the body, ensuring no post-ingestive effects occur. Dopamine is released from the nucleus accumbens with this procedure.⁸ The sweetness of fruit juices is rewarding as judged by “reward expectation-related neuronal activity” in the primate striatum, although it is produced by sugars other than sucrose.⁹ There is a preference for artificial sweeteners¹⁰ that in turn have been shown to influence the activity of the nucleus accumbens.¹¹ The intermittent access of rats to a saccharin solution rather than sucrose has also resulted in withdrawal symptoms when consumption stopped.² It appears that in part at least there is a response not to sucrose but rather to a sweet taste. In contrast Hoebel² suggested that in his rat model of addiction there was probably a greater response to sucrose as there are specific receptors on the tongue and in the gut. The taste receptor type 1 member 3, encoded by the TAS1R3 gene, responds to sweet tasting stimuli and is found in the gut as well as on the tongue.

More generally, is the response of a rat specifically to sweetness rather than palatability? Woolley¹² questioned whether the opioid regulation of food consumption reflects the macronutrient content rather than flavour. They studied the consumption of two types of food pellets that differed in flavour although they were nutritionally identical. A μ-opioid receptor agonist injected into the nucleus accumbens increased the consumption of both pellets in a similar manner if they were tested when only one of the two foods was present. However, when both flavours of pellets were presented simultaneously, the agonist increased and the antagonist naltrexone selectively decreased, the consumption of the preferred flavour. The authors suggested that based exclusively on flavour cues, opioid mechanisms in the nucleus accumbens increase the intake of palatable foods. Similarly the administration of naltrexone into the nucleus accumbens selectively decreased sucrose intake, although it had only a minimal influence on the consumption of less preferred chow.¹³ In addition a specific μ-agonist selectively increased the intake of sucrose, saccharin and a dilute saline solution.¹⁴ These findings demonstrated an important role for opioids in the nucleus accumbens in promoting the consumption of preferred palatable foods. When rats consumed a high-palatability sucrose solution the release of dopamine in the nucleus accumbens was dose dependent¹⁵ but palatable high-fat/sweet foods similarly induced dopamine release.¹⁶ The message is that it is palatability rather than sweetness or being sucrose that is critical in determining food preference.

This conclusion is supported by studying the impact of opioid drugs. As a generalization it has been known for many years that opioid agents enhance and opioid antagonists decrease feeding. In the rat the positive facial response to a sucrose solution was enhanced by morphine¹⁷ and decreased by opioid antagonists.¹⁸ The administration of morphine caused a short-term increase in food intake, and at least initially an increase in fat intake at the expense of carbohydrate.¹⁹ The opioid antagonist, naloxone, decreased fat rather than carbohydrate consumption in rats.20, 21 As it is known that for many rats fat is more attractive than carbohydrates these finding are consistent with the view that opioid mechanisms influence the intake of palatable foods. Such a suggestion is supported by the study of initial food preferences. As there is variability amongst rats in their preferences for carbohydrate and fat, Gosnell²² considered whether morphine was acting on food preferences. They distinguished fat-preferring from carbohydrate-preferring rats. Morphine increased carbohydrate intake in carbohydrate-preferring rats and increased fat intake in fat-preferring rats. Therefore morphine increased the intake of the preferred diet rather than a specific macronutrient. Similarly naloxone selectively decreased the intake of preferred foods and not sucrose content as would be predicted by the ‘sugar addiction’ hypothesis.²³