Are Certain Foods Addictive?

In a recent article (1), Dr. Rippe highlights that lifestyle medicine practitioners need to ground their recommendations on sound scientific evidence and that this is complicated by the fact that scientific information is often distorted and conjecture is sometimes confused with proof. This includes, for example, that associations between variables found in crosssectional studies are presented as causal relationships or that associations between variables found in epidemiological studies are often confounded by important third variables. The author illustrates several examples of findings that are often misinterpreted and presented as true facts, although existing evidence needs to be critically evaluated. Those include the notions that (a) sugar causes obesity, (b) certain foods are addictive, (c) certain foods cause cancer, (d) exercise is not effective for weight loss, and that (e) there is a causal link between sugar consumption and diabetes. I think that the author makes an important point in arguing that scientific evidence often is distorted by researchers or the media and that researchers and practitioners in the field of health behaviors continuously need to take care of scrutinizing research findings. Although I agree with most of the statements made in that article, I also think that some of them on the current concept of food addiction warrant a more in-depth discussion. APPROPRIATENESS OF ANIMAL MODELS Firstly, it is stated that “much of the argument related to food and addiction is based on [. . .] animal data” and that those models “may be poorly mimicked in human beings when it comes to food consumption.” Presumably, the author refers to paradigms that show addiction-like consumption of sugar and neurobiological changes after several weeks of intermittent access to sugar (2). In these studies, rodents are, for example, food deprived for 12 h and then have access to lab chow or sugar for 12 h. These paradigms are often criticized for being artificial and, thus, to have low value for making inferences about a possible sugar addiction in humans. However, I would argue that these paradigms match quite well to the eating styles of some individuals. For example, individuals with bulimia nervosa (BN) engage in binge eating, but undereat on non-binge meals (3, 4). That is, food intake may be restricted throughout the day, followed by a binge episode (which usually includes high-calorie, e.g., high-sugar, foods) in the evening. The same eating topography may be seen in other weight-concerned individuals who try to restrict their food intake [“restrained eaters” (5)], although not exhibiting fullblown binge episodes. To summarize, animal models are indeed an important part of the food addiction hypothesis and human studies are lacking to support some of the results found in those studies. Nevertheless, the paradigm of intermittent access to food may parallel eating topography of some individuals with restrained or disordered eating behavior. EVIDENCE FOR FOOD ADDICTION BASED ON DSM-5 Secondly, it is argued that“there is very little evidence for food addiction” based on the DSM-5 criteria for substance use disorder (SUD). Most articles in which the concept of food addiction is discussed refer to the substance dependence criteria in DSM-IV. In 2013, the DSM-5 was published and diagnostic criteria for SUDs now include 4 additional symptoms [11 symptoms in total (6)]. To the best of my knowledge, only one study has examined the new DSM-5 criteria in relation to eating behavior yet. In that study (7), a semi-structured interview was conducted, responses of which were qualitatively analyzed. Results showed that obese participants with binge eating disorder (BED), and to a lesser extent also those without BED, met the full criteria for SUD. Although participants rarely met three of the four new criteria, most of them met the new criterion of craving, or a strong desire or urge to use the substance. Admittedly, findings of this study should not be overinterpreted as the validity of the semi-structured interview is questionable and sample size was small. Undoubtedly, future studies are urgently needed that examine if the new DSM-5 SUD criteria can be translated to eating behavior and if those criteria are met by individuals who engage in over eating or binge eating [for a more detailed discussion (see Meule and Gearhardt, submitted)]. However,dismissing the appropriateness of the new DSM-5 criteria with regard to food addiction in the first place seems unfair.