We all know that substance abuse and dependence are serious, difficult-to-treat conditions. But ideas about what exactly constitutes substance abuse and addictions are changing.

Authors of the upcoming DSM-5 have proposed that substance use disorders (SUD) be categorized as “Substance Abuse and Addictive Disorders,” which would also include gambling addiction and would recommend other behavioral addictions (such as Internet addiction) for further study. Clinical experience, as well as a growing body of research, is asking whether food might also be considered a substance of abuse, and whether eating disorders are addictions.

The low cost and ready availability of highly processed foods have led some to speculate that these foods may contribute to eating disorders, just as newly available addictive substances like methamphetamine and MDMA/ecstasy have fueled a rise in addictions (Volkow N and Wise RA, Nat Neurosci 2005;8(5):555–560). As with addictions, remission rates in anorexia and bulimia are low. Binge eating and obesity are on the rise worldwide and are also difficult to treat. There is urgency, then, to develop other treatments for eating disorders, and viewing them through the lens of addiction may be fruitful.

Are eating disorders addictions? Our current eating disorder diagnoses are a mixed group: anorexia nervosa (AN), bulimia nervosa (BN), and eating disorder not otherwise specified (EDNOS), which currently includes the proposed binge-eating disorder (BED) and other atypical eating disorder presentations. Many patients with these diagnoses display symptoms that parallel the DSM-IV criteria for SUD: tolerance/needing more to achieve a desired effect (increased quantity); withdrawal symptoms when desired foods are restricted (often unpleasant emotional states); eating more than intended (binges); attempts to cut down (dieting); spending large amounts of time seeking the substance (excessive time and money spent on food and eating); functional impairment (social withdrawal, eating in secret); and continued use despite consequences (damage to teeth, esophagus, cardiac arrhythmias in BN; type II diabetes and heart disease in obesity) (Corsica J and Pelchat M, Curr Opin Gastroenterol 2010;26:165–169). Some subtypes of eating disorders seem to fit these criteria more closely (BED, BN, obesity) and others less so (AN); but all involve the obsessiveness, craving, preoccupation with a substance (food), and other cognitive symptoms that are characteristic of addictive thinking.

Similarities to Other Addictions
In the last decade, several observations of biological changes in eating disorders similar to those observed in addiction have become increasingly well established. First, neurohormonal systems are altered in the presence of food, particularly highly palatable foods such as simple sugars or high fat/high sugar diets. In fact, the mere taste, smell, and appearance of food have been shown to alter brain and body chemistry in patterns similar to the rewarding effects of abused substances. The most striking similarity to drug and alcohol abuse occurs when food is used in a restrictive/excessive pattern as seen in BED and BN, in which an episode of overeating is followed by either dieting or purging, which leads to further overeating. This pattern is associated with changes in the dopaminergic, opioid, and cannabinoid systems as well as in gut peptides (such as ghrelin, leptin, insulin) and, to a lesser extent, serotonin. These biochemical changes have been shown to both trigger and maintain the disordered eating patterns. In other words, in humans (and lab animals) who are predisposed to eating disorders, disordered eating leads to biological alterations that make the habit hard to break. In addition, as with other addictions, food may act differently in individuals with altered genetic systems. The evidence that food alters the brain in similar ways to other addictions provides a growing base for the food addiction concept.

A phenomenon similar to tolerance has also been described after prolonged consumption of foods, particularly highly palatable foods as noted previously. Studies have shown that dopaminergic reward systems are down-regulated with excessive food consumption, contributing to persistent overconsumption. This finding might explain why certain people find that the more they consume particular foods, such as highly concentrated simple sugars and carbohydrates, the more they ultimately need to satisfy cravings. Surprisingly, there’s also evidence of withdrawal from food. In studies of bulimic patients and of rats trained to develop binge-eating-like habits after intermittent feeding with highly palatable foods, the restriction of highly rewarding foods leads to elevated stress hormone levels, which are associated with anxiety, sleep disturbance, and increased cravings in humans, and increased fear sensitivity and anxiety in the rats. As explained by one of the pioneers in this field, Princeton psychologist Bart Hoebel, food addiction begins as hedonistic—for the pursuit of reward—but ultimately progresses to a type of negative reinforcement in which excessive food consumption is driven by the use of food to manage negative physiological states and affects (reviewed in Parylak S et al, Physiol Behav 2011;104(1):149–156). A similar process is described for other chemical and behavioral addictions and parallels the histories that most patients explain to us.

Measuring Food Addiction
Psychiatrists at Yale have recently developed a clinical instrument to measure food addiction. The Yale Food Addiction Scale (YFAS) (Gearhardt A et al, Appetite 2009;52(2):432–436) looks at the construct of food addiction as applied to people who meet the behavioral criteria for SUD with respect to food. This measure has yielded some interesting findings. The first, which I will call the milkshake study, found that women who met criteria for food addiction as defined by the YFAS showed higher levels of MRI activation in the caudate and medial orbitofrontal cortex (areas underlying craving and motivation) and lower activation in the lateral orbitofrontal cortex (self-control) upon seeing a milkshake, compared to women with lower YFAS scores. This finding echoes the biological findings above, that food targets the same reward systems as other drugs of abuse, and suggests further addiction-like features such as low levels of self-control.

A 2011 study of people with obesity found that only half who met criteria for BED had food addiction as measured by YFAS. Similarly, only 70% of people with food addiction are binge eaters. Two other recent studies found that between 8% and 11% of young adults with food addiction on the YFAS were actually normal or underweight (Meule A, Front Psych 2011;2(61):1–4). None of these studies found a correlation between YFAS score and body mass index, suggesting that not all food addiction results in obesity. In fact, just as drug or alcohol addicts may give the impression that their use is under control, even normal-weight bulimics and binge-eaters might have all the characteristics of true food addicts.

TCPR Verdict: 
Food can do things to the brain that addictive substances can do, and disordered eating patterns can yield biochemical changes like those seen in SUD. The concept of “food addiction” is not yet widely accepted, but an addiction-like profile is clearly over-represented in people with obesity and BED. Further research is needed to determine whether addictive behavior is increased in AN and BN as well. Exploration of these questions will advance our knowledge of not only eating disorders but also of the addictions as a whole.