(p. 3) Do We Need Another Eating Disorder?
When people learn I study eating disorders, they often ask about anorexia because this is the first eating disorder that comes to mind. For many, it’s the only eating disorder that comes to mind. Out of curiosity, I searched Google for images of people with eating disorders. The first screen filled with 39 unique pictures (Figure 1.1). Of these, 95% of the people were young, 95% were white, 92% were female, 62% were underweight, and 3% were overweight (that is, one woman). The term “eating disorder” conjures up a powerful image of a young, emaciated white girl. This image has masked the different ways that eating can be disordered and who can be affected. If the Google search reflected reality, only 5% of the images would have been underweight, and they would have been far more diverse in terms of age, gender, race and ethnicity—but I’ll get into that in the next chapter.
While all eating disorders involve a disturbance in eating, there are different ways that eating can be disturbed. These differences translate into meaningful distinctions in the central features of the disorders, when and where they emerge, how they impact the lives of those afflicted, and how they respond to treatment. Distinguishing among eating disorders is critical to understanding their causes and ultimately finding their cures. But if only 5% of eating disorders look like anorexia, what do the other ones look like and how many are there?
The Official Eating Disorders
There are three official eating disorders in the most recent edition of the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM-5).1 These disorders include anorexia nervosa, bulimia nervosa, and binge-eating disorder (BED). For the established eating disorders, the central defining feature falls along the dimension of how much you eat—anorexia represents a problem with eating too little and both bulimia and BED represent problems with eating too much. Bulimia has the added feature of behaviors intended to compensate for eating too much that distinguishes it from BED. Purging disorder differs (p. 4) from each of these disorders. Individuals with purging disorder eat neither too little nor too much. If you were dining out with someone with purging disorder (or even able to observe them in the privacy of their homes), their eating wouldn’t look different from someone trying to watch their weight, which means they’d look like a lot of people. In purging disorder, the eating disturbance occurs in how that person responds to what they ate. As one woman described, “After I eat a regular meal I feel so disgustingly full that it’s just something I ‘need’ to do.” They feel compelled to get rid of food through self-induced vomiting, or through misuse of laxatives, diuretics, or other medications. They purge to avoid gaining weight or becoming fat. Although purging can occur in both anorexia and bulimia, it’s not the central feature of either disorder, and both disorders occur in individuals who have never purged.
You wouldn’t be able to tell that a person has purging disorder just by looking at them. They eat enough food to maintain a minimally healthy body weight. This distinguishes them from individuals with anorexia, in whom restricted food intake causes medically low body weight. On the outside, people with purging disorder look just like people without eating disorders. As one girl explained, “To all, I appeared to be a healthy, normal girl, but I was secretly destroying my body.” Their bodies can range from being somewhat thin to being significantly overweight or even obese, just like the bodies of people without eating disorders.
Individuals with purging disorder don’t consume more food than most people eat—distinguishing them from individuals with bulimia and BED, in whom (p. 5) bingeing involves a loss of control while eating an excessive quantity of food. For example, a person with bulimia or BED might eat an entire package of store-bought cookies and a box of sugary cereal with milk, consuming over 3,000 calories in a single sitting, because they couldn’t stop eating until all the food was gone. The average size of binges is approximately 3,600 calories in bulimia and BED, according to feeding lab studies. This represents an abnormally large amount of food that most people would experience as undesirable and unpleasant. Those with purging disorder eat no more food than most people eat, which means less than 1,000 calories in two hours, according to feeding lab studies. If you’re wondering how much food goes into 1,000 calories, this could accommodate a blueberry scone and a Matcha Green Tea Frappuccino at Starbucks. However, this is an upper limit. In purging disorder, the average consumption prior to purging is around 500 to 750 calories (meaning people purge after either the blueberry scone or the Frappuccino—neither of which is excessive).
Some individuals with purging disorder experience a loss of control over their eating and may even subjectively experience their food intake as large. For example, one woman described how she had eaten an “entire bag of potato chips.” With additional probing, she shared that she had gotten the bag from a vending machine. But for her, this single serving represented an enormous amount of food. Another woman’s largest out-of-control “binge” involved 40 calories. These perceptions represent distortions of reality and a feature of the illness. However, not all people with purging disorder experience their eating as out of control. Some just need to purge to feel control over the effects of food on their bodies: “We threw up because we ate normally and felt fat, or felt that it would make us fat.” So, if purging disorder is not captured by any of the three existing disorders, why aren’t there four eating disorders—anorexia, bulimia, BED, and purging disorder?
Creating a New Diagnosis
Creating a new diagnosis means identifying a collection of symptoms or features that appear to go together and negatively impact people. The presence of distress or problems functioning in daily life distinguish a disorder from healthy behavior. For example, if symptoms cause dejection or problems at school, work, or in important relationships, this would be evidence that the condition is harmful and not healthy. Importantly, people do all sorts of harmful and unhealthy things that are not necessarily signs of mental illness. Have you ever crossed the street when the red flashing hand indicated that you should yield the right of way to vehicles that could crush your body? Yes? Repeatedly? Me too. So have most people walking in any semi-urban or urban setting. Even if something can or has been harmful (according to the Governors Highway Safety Association, over 1,500 people were killed crossing intersections in the United States in 20172), that’s not enough to make it a mental disorder. To be a mental disorder, the set of behaviors, thoughts, and feelings needs to be relatively rare and unusual, and cannot be considered normative either within the broader culture or within a subculture. Nor can (p. 6) something that a person chooses to do as an act of defiance against society or social norms be considered mental illness. The absence of choice in mental illness can be experienced as an inability to control what’s happening or as a compulsion to engage in a behavior. For example, vomiting in purging disorder, food restriction in anorexia, and even hand washing in obsessive-compulsive disorder are intentional. However, no one has chosen to feel compelled to engage in these acts. The key hallmarks that distinguish a mental illness from normality are (1) causing harm, (2) being (relatively) uncommon, and (3) being outside a person’s control.
Purging means inducing the forceful evacuation of matter from the body. It purposefully mimics the body’s response to a threat. For example, when a virus, bacteria, or poison enters the body, vomiting and diarrhea attempt to rid the body of the infection or toxin. Similarly, excessive urination occurs when the kidneys work to flush out toxins filtered from the bloodstream. As a biological function, vomiting, diarrhea, and excessive urination signify a problem, or threat to the body, and represent the body’s attempt to eliminate that problem and restore health. Reflecting this function, the word “purge” means to purify according to the Catechism of the Catholic Church.
In Catholicism, purgatory is where souls are cleansed of their sins before entering heaven. Throughout literature, we find examples of purging being used to eliminate moral contamination, such as the final chapter in Aldous Huxley’s Brave New World in which a Native American character seeks to rid himself of the ills of the modern dystopia by drinking mustard water to induce vomiting. This fictional account was inspired by the real-life Navajo sweet-emetic ritual—a purification rite employed to cure illness of bodily or spiritual origins. For this rite, the afflicted are brought into a special hut and given a basin containing an emetic that they drink to induce vomiting. The vomiting is viewed as central to eliminating pollution to restore bodily and spiritual health. While purging is a sign of malady across these diverse contexts, it is not a sign of mental illness because it has either a medical or cultural basis and is time limited. Purging is not maintained as an ongoing way of life.
Whether purging occurs due to a physical illness, as a cultural practice, or as an eating disorder, it signifies a deviation from what is “right”—a sign that something is wrong—and an attempt to re-establish goodness. In the context of purging disorder, food is viewed as a threat because it contributes to fat, which is experienced as vile and disgusting. People with purging disorder seek to rid their bodies of this threat. Because food is not a toxin or sin but is necessary to survival, the “threat” from food can never be completely eliminated, and purging becomes habitual. Habitual purging is a problem because purging is a violent act. As described in chapter 6, purging exacts an incredible toll on the body, contributing to a host of medical complications. Bodies are not equipped to withstand purging that recurs over weeks, much less months. One man who purged for over 40 years because of concerns about his weight described how over this time “stomach acids have made my teeth, especially the inside of my incisors, smooth.” In these ways, purging disorder differs dramatically from the wide range of behaviors that surround normal and healthy eating. However, establishing that deliberately vomiting (or inducing (p. 7) diarrhea or excessive urination) is not healthy or normal does not automatically translate into making purging disorder an official eating disorder. Purging occurs in both anorexia and bulimia, and some have argued that anorexia, bulimia, and purging disorder belong together as a single eating disorder.
Creating a new diagnosis means identifying a collection of symptoms and features that are different from established disorders. Early efforts at diagnostic classification for mental disorders drew from biology and medicine and sought to “carve nature at its joints.”3 More and more, we realize that distinctions we draw may be more functional than natural. A natural distinction is based in nature, like the distinction between dogs and cats. A functional distinction may not reflect any natural or biologically based difference. Instead, it may reflect demographic differences in who is affected by an illness, what treatments work for that illness, and illness course and outcomes. This creates a challenge because it means we employ culturally based values to choose what merits identification whenever we create a diagnosis.
Within the broader mental health field there is an ongoing debate between two camps: lumpers and splitters. Lumpers and splitters differ in their approach to identifying and distinguishing among mental disorders. Lumpers focus on commonalities, while splitters focus on differences. Lumpers would group under one diagnostic category what splitters would divide into separate diagnostic categories. For the lumpers, there is a strong case that mental disorders do not represent unique categories, but all reside along dimensions with one another. From this perspective, purging disorder is not different from other eating disorders in a way that requires its own diagnosis—differences are a matter of degree or severity, and purging disorder falls on a continuum somewhere between what we call anorexia and what we call bulimia (Figure 1.2). Some of my own work suggests that this may be true.
Over time, a person may go from having anorexia to bulimia to purging disorder. Did they really suffer from three different eating disorders or did they have one eating disorder that changed with time? Further, the clear demarcations between anorexia, bulimia, and purging disorder get blurry when we consider the cases that exist between their boundaries. For example, by definition, a patient with anorexia will be underweight, and a patient with purging disorder will not. Between them, however, may be a patient who has lost a great deal of weight but not enough to be underweight and who purges but does not binge. According to the DSM-5, this person could be diagnosed with atypical anorexia nervosa, an otherwise specified feeding or eating disorder (OSFED). Atypical anorexia with purging represents a bridge between purging anorexia and purging disorder, with all three residing on a continuum of current weight and weight loss. Similarly, purging disorder falls on a continuum with purging bulimia according to the amount of food consumed before purging. We have interviewed countless women who purge after eating episodes that fall into a middle ground between diagnoses. They may be eating more than what most people would probably eat in a given situation, but it’s difficult to determine because people differ a lot in how much they comfortably eat. Sure, eating an entire large pizza in an hour is clearly excessive, (p. 8) but is eating all but one slice normal? What about eating half of the pizza in the first hour, waiting a couple of hours and then eating the second half? Ultimately, we don’t have a precise threshold for exactly how much pizza is too much pizza to consume in two hours. And that’s just pizza. Thresholds get even murkier when you move on to combinations of foods. Purging after eating an amount that is in this “gray zone” (not clearly excessive but also not clearly typical) falls on a continuum between bulimia and purging disorder.
Given these points, it’s fair to ask if we really need another eating disorder diagnosis. If there are no objective boundaries to be found in nature, can’t we be more parsimonious and make everything just one diagnosis— “eating disorder”? There are brilliant experts in my field who have advocated for exactly this approach. But there are at least two reasons to exercise caution before lumping everything into a single diagnostic category.
First, how do you decide which continuum matters most? Purging disorder could be placed on a continuum with anorexia, if the central feature is weight or weight loss. Or it could be placed on a continuum with bulimia if the central feature is amount of food consumed. But this view starts with the assumption that the central defining feature of the established eating disorder is the only one that matters and ignores the central feature of purging disorder—purging. As noted already, not everyone with anorexia or bulimia purges, and no one with BED purges. Thus, purging disorder resides on a continuum with other eating disorders if we ignore purging or treat purging as incidental. If we start with the central feature of purging disorder—the purging behavior—and ask if this resides on a continuum with anorexia, bulimia, and BED, the answer is no. Inducing purging is not a matter of degree. One does not have a “touch of purging”— a person either purges or they don’t. It’s binary. And many patients with eating (p. 9) disorders have never purged, creating a clear boundary between purging disorder and every eating disorder in which people don’t purge (Figure 1.3). And much of my research has supported this point of discontinuity.
Second, even if we could identify a single continuum (or set of continua) on which all eating disorders reside, there are places along that continuum at which the implications of illness change. For example, if we think about blood pressure, we can easily agree that hypertension and hypotension (blood pressure that is too high and blood pressure that is too low) reside on a continuum with one another. However, the causes of hypertension and hypotension and their treatments differ so dramatically that it is useful to identify them as separate conditions. To my knowledge, no one has proposed a broad category of “blood pressure disorder” that would merge these two together just because they fall on opposite ends of a spectrum. Even for aspects of purging disorder that do reside on a continuum with anorexia (body weight) or bulimia (amount of food consumed), at certain thresholds the factors contributing to and resulting from the severity of symptoms produce functional disparities. Using different names is an effect approach to capture meaningful boundaries between functionally distinct syndromes. These same reasons explain how we got the official eating disorders we have today despite the overlap among them.
Once upon a time, anorexia nervosa was a newly identified disorder, as was bulimia nervosa, and BED—how did each of these achieve that status? For each disorder, a repeating pattern emerged in which a critical mass of evidence pushed each condition from obscurity to widespread recognition, leading to an accepted name and definition. A balance between scientific evidence and sociohistorical context contributed to the recognition and creation of each of the current official eating disorders as “new” eating disorders.
(p. 10) When the Old Eating Disorders Were New
Anorexia is not only the first eating disorder that comes to mind for people; it’s also the first eating disorder that was formally identified and named in the medical literature. Sir William Gull, working in London, introduced the term “anorexia nervosa” to the medical field in 1874 to describe a series of patients, mostly young females, who had started starving themselves for no apparent reason. In a prior lecture given in 1868, Gull made a passing reference to hysteric apepsia to describe self-starvation in some patients, but it was really the 1874 case series that garnered attention from the newly forming field of psychiatry. Like the features we see today, patients were indifferent to their own state and quite physically active despite their emaciation. Charles Lasègue, working in Paris, named and described l’anorexie hysterique in several of his female patients in 1873. The condition closely mirrored what Gull described and was published a year before Gull’s paper, but the name Gull chose rose above the competing terms that were emerging from around the world. Indeed, during the final quarter of the 19th century, physicians from Italy, Russia, Australia, and the United States were describing mysterious cases of girls starving themselves with no medical cause. Moving from the late 19th century into the early 20th century, anorexia was recognized as a rare but serious psychiatric disorder because patients who did not recover were at considerable risk of death from starvation. Even though the diagnosis was included in the first edition of the DSM published in 1952, most psychiatrists never encountered a case outside of textbooks. At that time, anorexia nervosa was listed among a disparate set of disorders because there was no category for eating disorders.
During the latter half of the 20th century, anorexia stopped being an extremely rare condition. The number of new cases seeking treatment was growing larger with each passing year, and programs specifically designed to treat this condition arose. Within these programs, clinicians started seeing patients whose problems resembled anorexia but did not fully match the diagnosis. More than 100 years after Gull introduced anorexia nervosa as a new disorder, Gerald Russell, also working in London, introduced the term “bulimia nervosa” in a 1979 article titled, “Bulimia nervosa: an ominous variant of anorexia nervosa.”4 His report described a series of 30 patients who were not underweight, who were bingeing and purging, and who were terrified of becoming fat. Publication of his article coincided with reports of binge–purge syndromes in normal-weight young women published in German and Spanish and concerns that bulimia or bulimarexia had become an epidemic on college campuses, affecting up to 15% of college women. Within a year, in 1980, bulimia became an official disorder included next to anorexia nervosa in the DSM-III, among disorders usually first evident in infancy, childhood, or adolescence.
For both anorexia and bulimia, there seemed to be a rapid shift in recognition in which experts from around the world began naming a condition that they were encountering too often to ignore and that was too different to be folded into existing diagnostic categories. Terms used to describe the conditions varied by author, with each choosing names they felt best captured the nature of the (p. 11) illness. In addition, some features varied across their descriptions. Like Gull’s and Lasègue’s descriptions of anorexia nervosa, several authors described a self-starvation disorder in young women. But neither Gull’s nor Lasègue’s patients expressed concerns about weight or shape—just a mortal fear of eating—while one of their contemporaries noted that his patient refused to eat to ensure that she could fit a ribbon around her waist because she didn’t want to become fat like her mother. Like Russell’s description of bulimia nervosa, bulimia in the DSM-III involved large out-of-control binges. But unlike Russell’s bulimia nervosa, DSM-III bulimia did not necessarily involve purging or compensatory behavior or body image concerns and was so broadly defined that it could accommodate cases of bulimia and BED. Indeed, some of the early literature on BED referred to it as “nonpurging bulimia” to distinguish it from most work conducted in those who binged and vomited. This changed after the 1987 publication of the DSM-III-R, in which bulimia was renamed “bulimia nervosa” and was defined by bingeing, compensatory behaviors (which included but extended beyond purging), and preoccupation with weight and shape.
BED had a less meteoric rise in recognition, potentially because it was absorbed in the original DSM-III definition of bulimia, potentially because it lacks the alarming features of emaciation or purging that differ so much from normal eating. Albert Stunkard, working in Philadelphia, named this condition in 1959, describing it in a subset of patients seeking weight loss for obesity. These patients had recurrent binge episodes, didn’t compensate, and differed demographically from the majority of those with anorexia—Stunkard’s patients were older, middle-aged individuals and far more of them were men. But very little additional work was completed on BED.5
The lack of work on BED changed when Robert Spitzer, who had served as the editor for the DSM-III and DSM-III-R, advocated for the inclusion of BED as a new eating disorder in the DSM-IV. Allen Frances, editor for the DSM-IV, resisted this proposal and expressed concern regarding over-pathologizing the common occurrence of overeating. Ultimately, a team of five eating disorder researchers, referred to as the DSM-IV Eating Disorders Workgroup,6 would examine the case for BED’s inclusion. In a published transcript of the DSM-IV workgroup’s deliberations, Dr. Frances commented, “It’s the proliferation problem. DSM-III-R has twice as many disorders as DSM-II. If this continues the system will become so cumbersome it’ll be unusable. In DSM-III and DSM-III-R, if a diagnosis was interesting and there was desire to stimulate research, it was put in. For DSM-IV, we’ve made a much higher threshold.”7 Additional concerns included the political ramifications of stigmatizing a segment of the obese population, that the disorder would be too common, that BED might not be distinct from bulimia, and that a premature definition of an understudied condition would result in it being accepted as true—the problem of reification. Reification of any diagnosis as “true” before enough evidence was collected could make it difficult to change, improve, or eliminate an invalid one. It seemed unlikely that enough evidence could be collected to support BED’s inclusion. In fact, at the time of the first discussion, it was unclear if there was enough time to even explore the option of its inclusion. (p. 12) Ultimately,8 the compromise in 1994 was to include BED in the DSM-IV as a provisional diagnosis in the broad, heterogeneous category of eating disorder not otherwise specified (EDNOS). Proposed diagnostic criteria were placed in the manual’s appendix. Another change to the DSM-IV was that the two official eating disorders, anorexia nervosa and bulimia nervosa, appeared in their own chapter, titled “Eating Disorders.”
The May 2013 publication of the DSM-5 introduced BED as an official eating disorder of equal standing with anorexia and bulimia. A search for “binge eating disorder” in the American Psychological Association’s PsychINFO database yielded 840 papers published between 1959 and May 2013, with 825 published since BED’s inclusion in the DSM-IV. The wealth of information that came from this work supported including BED as a new disorder in the DSM-5. This reveals an important relationship between scientific evidence and recognition of a new disorder. As pressures mounted to ensure that disorders met stringent criteria for inclusion in the DSM, the evidence required for admission increased exponentially across editions. Yet, the likelihood of obtaining that evidence is exponentially enhanced by including the disorder in the DSM (Figure 1.4). Returning to the DSM-IV workgroup’s early deliberations on BED, the following exchange9 highlights this dilemma clearly.
Walsh: The work group agrees these people [with BED] are out there.
Mitchell: But 18 months is quick. It’s taken us 10 years to nail down Bulimia.
Spitzer: Yes, it was my wisdom to include Bulimia in 1980 when it was new—like BED is now. That led to a decade of research which has helped us define it.
(p. 13) However, despite the incredible volume of research conducted prior to BED’s inclusion in the DSM-5, its official recognition was not free of controversy. In a Psychology Today blog, Dr. Frances ranked BED fifth in a list of the “ten worst changes” in the DSM-5 that should be ignored, writing, “5) Excessive eating 12 times in 3 months is no longer just a manifestation of gluttony and the easy availability of really great tasting food. DSM 5 has instead turned it into a psychiatric illness called Binge Eating Disorder.”10 In this blog and in interviews, Dr. Frances noted concerns about profit-related motivations for the publisher of the DSM and for pharmaceutical companies given an ever-expanding number of “patients” (a.k.a., customers). This response highlights the implications of creating a new disorder and the sociohistorical context within which purging disorder’s potential to become a new disorder is being weighed.
In some ways, purging disorder can be thought of as the inverse of BED. Individuals with BED binge but do not purge; individuals with purging disorder purge but do not binge. Between the two of them, purging is clearly more distinct from normality. Some of my favorite quotes from the ethnography on the DSM-IV workgroup’s deliberations would be profoundly disturbing if Dr. Frances’s comments referred to purging disorder instead of BED. These include Dr. Frances telling Dr. Spitzer, “I like your diagnosis. In fact, at times, I even have your diagnosis.”11 And later asking, “Are you testing for prevalence in your field trial? That’s a serious, serious problem. You’ve got to demonstrate that this diagnosis isn’t present in 50% of college girls.”12 Purging is such an alarming behavior that it’s unlikely Dr. Frances would casually joke about having purging disorder. Further, if 50% of college women were purging on a regular basis, we would see large-scale public health initiatives to address the problem rather than concluding that it wasn’t really that big of a deal. So, why wasn’t purging disorder included in the DSM-IV? Because no one had identified it; no one had even given it a name when decisions were being made for the DSM-IV.
Naming Purging Disorder in the Medical Literature
In 2005, I used the term “purging disorder” to describe women who were purging by self-induced vomiting, laxatives, or diuretics after consuming normal or small amounts of food to control their weight or shape. The title of the article, “Purging disorder: an ominous variant of bulimia nervosa?” was an homage to Russell’s paper on bulimia. I ended the title with a question mark because I understood how much more work would be needed before any conclusion could be drawn. I set out to name and describe a collection of symptoms and features that hung together and affected individuals in a particular way that seemed different from bulimia. Before this article, I wrote another published in 2001 but referred to the condition as “subjective bulimia nervosa.” And even before the 2001 article this clinical presentation had been described in a 1986 article written to inform the DSM-III-R workgroup’s examples of atypical eating disorders. But it had (p. 14) no name other than “atypical eating disorder group 1,” and the fourth example for an EDNOS in the DSM-IV (literally, EDNOS type 4). From 2001 to 2005, I stopped using “subjective bulimia nervosa” when I presented findings at conferences because I realized that name implied that purging disorder was a type of bulimia nervosa, which was looking less and less true according to my data. In addition, by doing this, I excluded from recognition those who purged without ever feeling a loss of control over their eating. Similar to the historical contexts in which anorexia and bulimia gained widespread recognition, reports of purging disorder started appearing from around the world, with cases in Italy, Australia, England, Germany, and Japan. Shortly after entering the 21st century, the field was encountering alarmingly high numbers of women who purged without bingeing and were not underweight—women with an eating disorder that did not fit into anorexia, bulimia, or BED diagnoses.
In the leadup to the DSM-5, the International Journal of Eating Disorders published special issues focused on the classification of eating disorder diagnoses in 2007 and 2009, and I was invited to contribute reviews of purging disorder to both. The 2007 article identified 10 empirical (data-based) papers, across which seven different names had been used in reference to the condition, including “subjective bulimia nervosa,” “EDNOS-P,” “purge eating disorder,” and “purging disorder.” Definitions differed across these papers, and even papers using the same name had somewhat different definitions, with “purging” sometimes being expanded to include fasting and excessive exercise. (For the record, fasting and excessive or compulsive exercise are compensatory behaviors that may be used in bulimia, but they are not actually purging behaviors because they do not involve forceful evacuation of matter from the body.) By the 2009 review, another 20 empirical papers had been published, reflecting a tripling of the scientific work on the topic. In contrast to the use of “purging disorder” in only 1 of the first 10 articles published on the condition (my 2005 article), “purging disorder” was used in 7 of the last 10 articles published (definitely not all mine). In addition to greater consistency in terminology, there was greater consistency in definitions.
Key conclusions of the 2009 review were organized around established criteria for considering a new diagnosis:
Criterion 1. Is There Ample Literature on Purging Disorder? With 48 total publications on purging disorder, including 30 empirical papers (6 from my lab and 24 from other researchers), there was ample literature on purging disorder. It wasn’t BED’s 840 publications as of May 2013, but it was considerably more than BED’s 15 publications as of 1994.
Criterion 2. Is There a Common Set of Diagnostic Criteria for Purging Disorder and Are There Assessment Tools Available for Measuring the Syndrome? Although this was clearly absent in the 2007 review, the definition of purging disorder had largely coalesced across more recent studies. Multiple groups were identifying purging disorder in those who purged without bingeing and who were not underweight. Further, the “gold standard” for eating disorders assessment, the Eating (p. 15) Disorders Examination, was ideally suited to identifying the disorder, distinguishing it from anorexia and bulimia, and capturing its severity and remission status.
Criterion 3. Is There Diagnostic Reliability? There was. People tend to be very clear on what the definition of purging is. There is high agreement among different interviewers about the presence or absence of purging, and high agreement over time in people’s own reports of whether or not they had purged.
Criterion 4. Can Purging Disorder be Differentiated from Other Eating Disorders (and Normality)? Here the literature had some important gaps. While every single study ever published supported that purging disorder was, indeed, not normal and was clearly a clinically significant disorder of eating, comparisons between purging disorder and other eating disorders were a bit sparse. All my research had compared purging disorder to bulimia—reflecting my early bias in thinking of it as a variant of bulimia. Unfortunately, a lot of the field followed suit, and very few studies compared purging disorder to anorexia or BED. That was true then, and it is less true now. But the biggest challenge in identifying this new disorder was determining whether it should be distinguished from the existing eating disorders or whether the definition of existing eating disorders should be expanded to include purging disorder. Indeed, the International Classification of Diseases published its 11th edition in June 2018, and it redefined a binge-eating episode as “a distinct period of time during which the individual experiences a subjective loss of control over eating, eating notably more or differently than usual, and feels unable to stop eating or limit the type or amount of food eaten.”13 Those who purge after feeling a loss of control over their eating could be diagnosed with bulimia using ICD-11 criteria even if they didn’t eat that much. That revision subsumes many individuals with purging disorder into a diagnosis of bulimia. Importantly, given that many patients purge without ever feeling a loss of control over their eating, it doesn’t account for all cases of purging disorder.
Criterion 5. Is There Syndrome Validity? For this article, we focused on a specific kind of validity—the extent to which a diagnosis helped clinicians know how to treat the illness and patients’ likelihood of recovery. Again, at the time of the 2009 review, considerable data supported that purging disorder looked different from bulimia on several factors, but there were limited data on course, treatment response, and outcome. These are all indicators of clinical utility, or the value of the diagnosis in a treatment setting—a key criterion for adding a diagnosis to the DSM-5.
We ended the review by discussing the pros and cons of four options for addressing purging disorder in the DSM-5. Option 1 was to do nothing—leave purging disorder as an unnamed example in a heterogenous group of eating (p. 16) disorders that just weren’t quite anorexia or bulimia. The pro was that this option certainly minimized proliferation of diagnoses and would not contribute to a diagnostic system that was unweildy or untrustworthy. However, it was already clear that purging disorder was a clinically significant problem that had been buried in a heterogenous group for which little to no useful information was being systematically obtained. For example, the largest and most representative epidemiological study of mental disorders conducted in the United States at the time couldn’t estimate how many people had purging disorder because participants were only asked about purging if they binged. For those who didn’t binge, all questions about purging were skipped. Fortunately, epidemiological data from other countries made clear the large number of people who would be ignored by doing nothing. Ultimately, we noted that doing nothing would limit collection of needed information to validate the disorder and to help the considerable number of people currently suffering it.
Option 2 was to redefine diagnostic criteria for official eating disorders to absorb purging disorder. The pros were that it minimized proliferation of diagnoses and it would support insurance coverage for treatment of those purging after eating normal amounts of food, even though they weren’t underweight and weren’t bingeing. The cons were that no revision to existing diagnostic criteria could accommodate all cases of purging disorder without damaging the clinical utility of anorexia or bulimia. The only way to include all cases of purging disorder in the definition of bulimia was to eliminate the binge-eating criterion. However, comparisons between purging disorder and bulimia showed significant differences on psychological and biological factors that might have major implications for treatment. The only way to include all cases of purging disorder in the definition of anorexia was to eliminate the low-weight criterion for anorexia. Although there weren’t many studies comparing purging disorder to anorexia, too much evidence supported the low-weight criterion for anorexia to change this aspect of its definition. Even if purging disorder could be absorbed into one of these diagnostic categories without impacting their clinical utility, there wasn’t enough evidence to determine whether it should go into a broader category for bulimia or a broader category for anorexia.
Option 3 was to include purging disorder as an official eating disorder diagnosis—give it a name, diagnostic criteria, and a diagnostic code, putting it on equal footing with anorexia and bulimia. The pro was that this would support insurance coverage for treatment and promote research into the causes and consequences of the illness, which could ultimately lead to the prevention and treatment of purging disorder. The clear con was this meant creating a new eating disorder diagnosis before sufficient scientific evidence supported the decision. Further, once the diagnostic criteria were specified, all research on purging disorder would use those criteria—whether they accurately defined the disorder or not. This is the problem of reification raised during deliberations on the inclusion of BED in the DSM-IV.
Option 4 was to include purging disorder as a named condition with a provisional definition. This represented a compromise between Option 1 and Option 3 (p. 17) that maximized the benefits while minimizing the costs of each of those extremes. This option offered an opportunity for researchers to “study what we define” without setting in stone a definition that requires further investigation. Future editions of the DSM would need scientific study of purging disorder. And the first step in the scientific study of anything is to form the hypothesis that it exists. No field of science has gained much ground by accepting the null hypothesis that there is no difference between groups without ever testing it—this is the classic fallacy of equating an absence of evidence with evidence of an absence. A name and provisional definition would pave the way for tests of whether or not there were differences between purging disorder and anorexia, bulimia, and other eating disorders in terms of course, outcome, and treatment response. Such research would be enormously helpful to those suffering from purging disorder. In addition to these academic and practical benefits, this approach mitigated long-noted problems with prematurely introducing disorders in the absence of evidence supporting their clinical utility. Purging disorder would not be an official diagnosis and would not add to the overall number of diagnoses in the DSM-5. The final advantage of this option is that it allowed the needs of those currently suffering from purging disorder to be recognized.
One person wrote, “From my experience, I have always felt like a failed bulimic, but not quite an anorexic [ . . . ] like a grey area that doesn’t fit.” Another: “I have always thought that I fell out of the Bulimia and Anorexia ‘guidelines’ and that was a good excuse to tell myself I probably did not have a ‘disorder’.” “Every time I try to get help, the doctors will refer to my habits as binging and purging, and even when I explain otherwise I see them write binging on their papers [ . . . ] Over the past 4 months (after I started throwing up blood) I have been coming to terms with my disorder, and seeking treatment, but have had very little luck with finding someone willing to get past their preconceptions and stop assigning me symptoms that I don’t have.” Or “Many people feel that if you do not meet the DSM description, you do not have an eating disorder [ . . . ] I began to purge, but I did not binge. But because I was never underweight, not many noticed that I had a problem. I did at one point go see a counselor to have an intake, who at the end of the session basically told me I was fine.” Or even under the best circumstances, “When I was diagnosed in [ . . . ] I was told I had an ‘eating-disorder-otherwise-unspecified’ or something along those lines. It was oddly frustrating to hear I did not meet the requirements for bulimia. I had come to a point where I realized I needed help, had accepted I had a problem and found it hard to hear that my problem didn’t have a name.”
The 2009 review was commissioned by the DSM-5 workgroup, and we were asked not to express a preference for any of the options. After the review was completed, new studies might be published that would influence the final decision in ways that could not be predicted. Thus, any recommendations might be premature or viewed as biasing decisions that needed to be made within the structure of the DSM’s organization and process. This process was established to ensure consistency and rigor for all diagnoses under consideration—not just those in the eating disorders section. Of course, it was unlikely that information on (p. 18) long-term course and treatment response would emerge with so little time left before the fifth edition was scheduled for publication, and nothing earth-shattering appeared on purging disorder before the DSM-5 workgroup’s deliberations were completed. Thus, Option 4 continued to have the most benefits and the fewest costs, and Option 4 is largely what happened.
Purging disorder is a named condition in the DSM-5. It falls within the broad category OSFED, much as BED appeared as a named condition in the DSM-IV’s EDNOS. As a set, the disorders classified as OSFED do not have the same standing as anorexia, bulimia, or BED. Each of those eating disorders possesses its own diagnostic criteria, diagnostic code, and subsection of the chapter on feeding and eating disorders describing diagnostic features, associated features, prevalence, development and course, risk and prognostic factors, culture-related diagnostic issues (gender-related diagnostic issues), diagnostic markers, suicide risk, functional consequences, differential diagnosis, and comorbidity. The DSM-5 offers the following description for the newly named condition: “Purging disorder: Recurrent purging behavior to influence weight or shape (e.g., self-induced vomiting; misuse of laxatives, diuretics, or other medications) in the absence of binge eating.”14 No additional information is provided because purging disorder is not a separate, official eating disorder like anorexia, bulimia, or BED. Its diagnostic code is the same as that for atypical anorexia nervosa and night eating syndrome (NES)—although it shares no overlapping features with NES and does not even seem to reside on any identifiable continuum with this fellow OSFED.
Provisional diagnostic criteria were not provided for any of the OSFEDs in the DSM-5. This decision reflected the consequences of including provisional diagnostic criteria for BED in the DSM-IV. Revising BED’s criteria based on research was nearly impossible. Almost all studies of BED had used the provisional criteria in selecting participants. This fulfilled funding agencies’ expectations that methods could be reproduced across studies and contributed to the accumulation of a wealth of information on BED’s course, outcome, and treatment. But it also represented the problem of reification because we only had data on BED as it had been defined in the DSM-IV. In contrast, the descriptions for OSFEDs in the DSM-5 provide enough information to ensure some consistency across studies, do not preclude variations across studies that could be used to refine the definition, and largely follow what the field was already doing.
Although purging disorder is not an official eating disorder in the DSM-5, it is a specified eating disorder. Ultimately, Dr. Frances’s point about proliferation is right—there has to be a limit to the number of disorders that get named. If everything that is different from normal eating could be a disorder, then there would be an eating disorder for those who repeatedly chew and spit out their food without swallowing it, and one for those who follow strict rules about eating clean or correctly, referred to as orthorexia, and so on and so forth. To address this practical limit, the DSM-5 includes a broad “unspecified eating disorder” option that captures everything that merits diagnosis but has garnered insufficient evidence to be included as either an official eating disorder or as an OSFED.
(p. 19) Why We Need Purging Disorder
In the fairy tale “Rumpelstiltskin,” the Miller’s Daughter must name the man who has spun straw into gold in order to keep her son—reflecting a long-held superstition that the very act of naming an entity gives one power over that entity. The cultural belief that names have power is evident in the modern practice of diagnosis in which patients feel reassured when their collection of symptoms and signs has a name, even when the name provides little more than a shorthand reference for their described experiences. Naming creates the feeling that something is real, that one is not alone in having experienced the condition, and that there is hope for a good outcome. As one person wrote, “Finally a name for what is wrong with me. When you don’t fit into either category of eating disorders, no one can help you. I just wanted to say I’m glad you have found a name for the demon that haunts me.” The key challenge is that the process of naming provides so much reassurance that it can be difficult to relinquish names that do not work toward their ultimate goals—the goal of controlling mental illness through effective prevention and treatment.
The inclusion of purging disorder as a named and described condition in the DSM-5 has motivated considerable work. We now know much more about its development and course, risk and prognostic factors, consequences, and treatment. And what we’ve learned supports that purging disorder merits recognition as a new eating disorder. Based on what happened following the inclusion of bulimia in the DSM-III and BED in the DSM-IV, we can expect this body of research to continue to grow as we head toward the next edition of the DSM, the DSM 5.1. So, yes, we need another eating disorder. And, in the rest of this book, I want to share with you what we’ve been able to learn about this new eating disorder. (p. 20)
1. Some might argue that there are four official eating disorder diagnoses in the DSM-5, anorexia, bulimia, BED, and avoidant/restrictive food intake disorder (ARFID). ARFID was included in the DSM-5 as a revision to the DSM-IV’s feeding disorder (p. 146) of infancy and early childhood. The definition of ARFID, its diagnostic code, and its placement immediately after pica and rumination disorder, which are feeding disorders in the DSM-5, suggest that ARFID is a feeding disorder, not an eating disorder. However, the distinction between feeding and eating disorders is not entirely clear or universally agreed upon.
2. Governor’s Highway Safety Association (2019). Pedestrian Traffic Fatalities by State. https://www.ghsa.org/sites/default/files/2019-02/FINAL_Pedestrians19.pdf.
3. This phrase describes identifying biologically based differences to distinguish between two categories or classes. For example, in medicine, diabetes is divided into two types—type 1 and type 2. In type 1 diabetes, the pancreas no longer makes insulin and patients must take insulin through shots or a pump to survive. In type 2 diabetes, the pancreas continues to make insulin but the body is less responsive to that insulin, and the illness can often be managed through diet, exercise, and oral medication. Not producing any insulin and being resistant to the effects of insulin produce the same symptoms, but the differences in underlying biology and treatment justify separate diagnoses.
4. Russell, G. (1979). Bulimia nervosa: An ominous variant of anorexia nervosa. Psychological Medicine, 9, 429–448.
5. Within his own program of research, Dr. Stunkard was far more fascinated by night eating syndrome (NES). NES is another named and described condition included among the DSM-5 OSFEDs.
6. The DSM-IV Eating Disorders Workgroup included four psychiatrists and one clinical psychologist, of whom four were men and one (a psychiatrist) was a woman.
7. McCarthy, L. P., & Gerring, J. P. (1994). Revising psychiatry’s charter document: DSM-IV. Written Communication, 11, 147–192 (p. 172). doi:10.1177/0741088394011002001
8. I’m glossing over a lot of rich detail available from the ethnographic study of the DSM-IV Eating Disorder Workgroup’s process, included in McCarthy and Gerring (1994) referenced in the Source References section.
9. McCarthy & Gerring, p. 172.
10. Frances, A. (2012). DSM 5 is guide, not bible—Ignore its ten worst changes. Psychology Today. https://www.psychologytoday.com/blog/dsm5-in-distress/201212/dsm-5-is-guide-not-bible-ignore-its-ten-worst-changes.
11. McCarthy & Gerring, p. 172.
12. Ibid., p. 176.
14. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed., p. 353). American Psychiatric Association.