Jump to Section 1. Introduction 1.1. General background 1.2. Addiction – it depends what you mean 1.3. A rat model 1.4. Sugar, sweetness or palatability? 2. Sugar and reward mechanisms 2.1. Introduction 2.2. Activation of reward circuits 2.3. Comparisons of dopamine release induced by food and drugs of abuse 2.3.1. Neurophysiology 2.3.2. Voltammetry 2.3.3. Microdialysis 3. Is sugar addictive in humans? 3.1. Food craving 3.1.1. The dimensions of food craving 3.1.2. Fasting 3.1.3. Factors influencing craving 3.2. Changes in sucrose preference over time 3.3. Sucrose, sweetness or palatability? 3.3.1. Palatability 4. Sugar consumption and obesity 4.1. Introduction 4.2. A sweet-tooth? 4.3. Dopamine and obesity 5. Bingeing 5.1. Introduction 5.2. Binge-eating disorder 5.3. Foods eaten while bingeing 5.4. Opioid antagonists 5.5. The impact of antagonists in the normal population 5.6. Opioid antagonists and eating disorders 5.6.1. If binge eaters are addicted to sugar then administration of an opioid antagonist would immediately induce withdrawal symptoms. Such responses would only be observed in those with a history of bingeing 5.6.2. The addiction hypothesis suggests that bingeing is an attempt to induce the release of endorphins to counter the adverse effects of withdrawal. It follows that as opioid antagonists further reduce endogenous opioid activity they should in the short-term further induce bingeing, in an attempt to stimulate endorphin release to reverse the effects of withdrawal 5.6.3. The response to opioid antagonists should occur with a dose capable of blocking opioid activity 5.6.4. The report that naloxone induces withdrawal symptoms in rats addicted to sugar1 but not fat7 leads to the prediction that if ‘sugar addiction’ is the mechanism that underlies bingeing, in those who binge opioid antagonists should selectively influence the choice of sugar rather than fat 6. Discussion Conflict of interest References

A series of predictions developed from the hypothesis that an addiction to sucrose consumption can develop have been discussed. If sugar addiction has played a major in the increase in obesity, fasting should increase food cravings predominantly for sweet items; cravings should occur after an overnight fast; withdrawal symptoms should prevent a decline in the preference for sucrose; the obese should prefer sucrose containing rather than other palatable foods or find sweetness particularly attractive; sucrose containing rather than other food items should predispose to obesity. These predictions have in common that on no occasion was the behaviour predicted by the addiction model supported by human studies. These findings give no support to much of the popular literature that proposes a widespread addiction to sugar.

Although both the animal literature147x147Marrazzi, M.A., Kinzie, J., and Luby, E.D. A detailed longitudinal analysis on the use of naltrexone in the treatment of bulimia. Int Clin Psychopharmacol. ; 10: 173–176

Crossref | PubMed | Google ScholarSee all References and the present consideration of human data agree that sucrose, when consumed as part of a normal diet, does not produce physical dependence, the animal literature has lead to a more specifically hypothesis. The suggestion is that it is the way that sucrose is consumed that may have a specific role in generating BED.1x1Avena, N.M., Rada, P., and Hoebel, B.G. Evidence for sugar addiction: behavioral and neurochemical effects of intermittent, excessive sugar intake. Neurosc Biobehav Rev. ; 32: 20–39

Crossref | PubMed | Scopus (324) | Google ScholarSee all References A summary of a symposium that considered food addiction concluded that “…even highly palatable food is not addictive in and of itself. Rather, it is the manner in which the food is presented (i.e. intermittently) and consumed (i.e. repeated, intermittent “gorging”) that appears to entrain the addiction-like process”.158x158Corwin, R.L. and Grigson, P.S. Symposium overview – food addiction: fact or fiction?. J Nutr. ; 139: 617–619

Crossref | PubMed | Scopus (44) | Google ScholarSee all References The present review therefore also examined various predictions derived from the hypothesis that an addiction to sugar is central to bingeing disorders. Dieting should predate the development of BED as it is only when dieting predated sucrose consumption did rats display ‘additive type’ behaviour; sweet items should be preferentially consumed while bingeing; dietary style rather than psychological, social and economic factors should predispose to binge-eating disorders; opioid antagonists should cause withdrawal symptoms; BED should develop at a younger age when there is a greater preference for sweetness. Again these various predictions have in common that on no occasion were they supported by human data.

The development of an animal model of eating disorders inevitably offers an implicit model of the cause of such phenomena: it is implied that eating disorders reflect the nature of the diet and eating style. There is, however, a second commonly discussed model: an eating disorder is a response to a pre-existing problem of a psychological or social nature, it is an attempt at a solution albeit not a successful approach. There are wide ranging theories about the aetiology of eating disorders emphasizing biological, psychological, social and family factors. Striegel-Moore159x159Striegel-Moore, R.H., Silberstein, L., and Rodin, J. Towards an understanding of risk factors for bulimia. Am Psychol. ; 41: 246–263

Crossref | PubMed | Scopus (523) | Google ScholarSee all References argued that this heterogeneity of variables suggests that unidimensional models of aetiology are unlikely to be valid. When Polivy and Herman160x160Polivy, J. and Herman, C.P. Causes of eating disorders. Ann Rev Psychol. ; 53: 187–213

Crossref | PubMed | Scopus (341) | Google ScholarSee all References reviewed the causes of eating disorders they considered socio-cultural factors such as media and peer influences; family factors such as enmeshment and criticism; negative affect; low self-esteem; body dissatisfaction. Although they also considered cognitive and biological factors it is clear that it is rarely suggested that the pre-existing diet is the causal factor. Even if nutritional mechanisms can in the future be shown to play a role in eating disorders, at the most they will only be a partial answer as psychological, social and economic factors are clearly necessary to explain much of the increase in obesity and eating disorders. Most experts will find implausible any suggestion that there is a single mechanism that plays a large role, and although we do not fully understand the details, one thing that is certain is the complexity. Any suggestion that problem eating to a large extent reflects a single causal factor is bound to be wrong. Although the scientists who work on animal models tend to be cautious when drawing conclusions, many in the general population have falsely attributed addictive properties to sucrose, even when consumed in a normal manner.

The purpose of this review was not to suggest that the widespread and cheap provision of palatable foods has not played a role in the increasing incidence of obesity. Rather it is important that the underlying mechanisms are understood so that this knowledge can generate appropriate responses. If addiction to sucrose plays an important role in the development of obesity then various conclusions will follow. If it is falsely believed that sucrose is addictive then inappropriate responses will be made and helpful behaviour will be avoided. Already some have been worried that dieting might not be the best solution if it generates cravings and withdrawal symptoms.5x5Trotzky, A.S. The treatment of eating disorders as addiction among adolescent females. Int J Adolesc Med Health. ; 14: 269–274

Crossref | PubMed | Google ScholarSee all References If sucrose is the villain then attempts to deal with obesity should concentrate on this ingredient. If sucrose is not central then inappropriately concentrating on this food item will ensure that more beneficial responses will be ignored. For example it has been frequently proposed that we should concentrate on fat as it plays a major role in determining palatability, has a more limited impact on satiation than other macronutrients, and plays a major role in the energy density of the diet.

A more general conclusion is that the use of rat behaviour as an experimental model of the human condition is fraught with difficulties. Before behavioural scientists extrapolate findings obtained with animals it is essential that they consider analogous human data. Only if there are marked parallels in the response of the two species will the use of animal models be informative.