When completing my master’s degree in science and the public (through the Center for Inquiry and SUNY-Buffalo), I chose eating disorder misinformation as the subject of my thesis. This was important to me for several reasons, including that it involved several of my longstanding interests such as myths and misinformation (a typical skeptical subject); eating disorders (a subject I first became involved with when helping an ex-girlfriend struggle with bulimia); and the news media (the subject of my 2003 book Media Mythmakers: How Journalists, Activists, and Advertisers Mislead Us).
I wanted to understand and explain the processes by which valid scientific information about these important health disorders got translated—and often mistranslated—between clinical researchers and the public, mediated by eating disorder information clearinghouses (such as the National Eating Disorder Association), news journalists, and activist filmmakers.
Misinformation about eating disorders is not like misinformation about a car’s gas mileage or the weather. Eating disorders are mental illnesses with potentially lethal consequences. Sufferers and their loved ones deserve accurate, up-to-date information about the diseases, but upon closer inspection, trusted sources of information often turn out to be not so trustworthy. This important topic has received little or no attention in the mainstream media and, to the best of my knowledge, in academia. Part of this may be because the problem of eating disorder misinformation is multidisciplinary and includes journalism, public education, media literacy, science literacy, medicine, and psychology. Here I discuss one case study of flawed and misleading information about eating disorders presented by one of the largest and most prominent eating disorders information clearinghouses, the National Eating Disorders Association.
A brief overview of anorexia is helpful. Anorexia is diagnosed using criteria in the Diagnostic and Statistical Manual of Mental Disorders, issued by the American Psychiatric Association. They are: refusal to maintain body weight at or above a minimally normal weight for age and height; intense fear of gaining weight; disturbance in the way one’s body weight or shape is experienced (especially undue concern of body weight on self-evaluation); and amenorrhea (lack of normal menstrual cycles).
The most reliable estimates of anorexia place the average prevalence rate (the total number of cases in the population) for young females at 0.3 percent for anorexia nervosa (van Hoeken et al. 2003, 11), and “A mean incidence [the number of new cases in a population over a year] in the general population of 19/100,000 a year in females and 2/100,000 a year in males . . . estimated from twelve cumulative studies” (Treasure and Schmidt 2004). In other words, out of every 100,000 females in a given year, nineteen of them will develop new cases of anorexia, an incidence of 0.00019. By comparison, schizophrenia is about three times as common in the United States.
The National Eating Disorders Association
In a press release issued by the National Eating Disorders Association (NEDA) for National Eating Disorders Awareness Week (February 24–March 2, 2013), the NEDA included a section titled “U.S. Statistics on Eating Disorders.” It included statistics such as “There has been a rise in incidence of anorexia in young women 15-19 in each decade since 1930” and “The rate of development of new cases of eating disorders overall has been increasing since 1950.”
There was no citation or reference attached to the statistics, so I contacted a spokeswoman for NEDA inquiring where that data came from. I received the following reply from spokeswoman Kelly Williams of NEDA: “Here is the source for that statistic. Hoek, H. W., and van Hoeken, D. 2003. Review of the prevalence and incidence of eating disorders. International Journal of Eating Disorders, 383–396.” The journal article was not included as an attachment, but I soon located the paper. A careful review of the cited study found that the quoted information did not in fact appear; however, the following passage seemed to be closest to the statistic offered by the NEDA: “Lucas et al. (1999) reported an incidence of 73.9 per 100,000 person-years for 15–19-year old women over the period of 1935–1989, with a continual rise since the 1930s to a top rate of 135.7 for the period 1980–1989.” This seemed to be the most relevant statistic. It seems that the correct citation should not be Hoek, H. W., and van Hoeken, D. (2003), since the research does not appear in that study and is merely referenced in it, but should instead be Lucas, A.R., Crowson, C. S., O’Fallon, W. M., et al. (1999). The ups and downs of anorexia nervosa. International Journal of Eating Disorders, 26, 397–405.
However, there remained two other problems with the statistic cited by the NEDA. First, the date is incorrect; the Lucas (1999) research began in 1935, not 1930 as stated in the information provided by the NEDA (and which also appeared on their website). Second, that Lucas study is quoted within a more recent study (van Hoeken et al. 2003) which actually contradicts its conclusions. Van Hoeken et al. concluded that “The incidence of anorexia nervosa increased over the past century, until the 1970s” (p. 383, emphasis mine). They also state that:
There has been considerable debate regarding whether the incidence of eating disorders is, or has been, increasing. Various studies have reported diverging incidence rates, which may be due to methodologic problems. . . . The debate still continues about the extent to which there has been an increase in the true incidence (i.e., the incidence in the community) of anorexia nervosa in the 20th century. (388)
The Lucas study involved 174 women with “definite” or “probable” cases of anorexia nervosa (along with forty-three “possible” cases). Of this relatively small sample, 93 percent of the participants were white women with an average age at diagnosis of 21.5. It’s not clear why this study (of fewer than 200 white Minnesotan women) was selected by the NEDA out of all the corpus of research available to represent a statistic about the incidence of anorexia nervosa in the general population of the United States.
This raises another question: If it’s true, as van Hoeken et al. (2003) state, that there is legitimate and “considerable debate regarding whether the incidence of eating disorders is, or has been increasing,” then why is that uncertainty not reflected in the National Eating Disorders Association statistic? The statistic is offered as established fact—not one of many contradictory studies. In science, it is of course common to have at least some published studies supporting contradictory hypotheses. For this reason, researchers addressing the overall weight of the research—the direction of evidence over dozens or hundreds of studies—are urged not to cherry-pick a few studies and present them as representative of the body of literature but instead to present examples of research both supporting and refuting a given hypothesis. To do otherwise is, at best, biased and sloppy research.
The van Hoeken study comes to a significantly different conclusion about the incidence of anorexia than the statistic that was pulled out of it: “the incidence of anorexia nervosa increased over the past century, until the 1970s” cannot be reconciled with “There has been a rise in incidence of anorexia in young women 15-19 in each decade since 1930,” and certainly not as it relates to the time frame from the 1970s to the present. A 2012 review of literature on the topic (Torres et al. 2012) notes that “the majority of research invalidating the increase of AN [anorexia nervosa] date from a more recent period. . . . In summation, after thorough review of the clinical research, we consider a position either for or against an increase in incidence of AN as speculation” (9). They conclude that “A substantial controversy exists as to whether or not AN is increasing. . . . The most recent studies suggest that the incidence of severe cases had increased over the past century, peaking in the 1970s, with a plateau from then to the present day” (16). If the incidence of anorexia nervosa has remained the same for at least the past thirty to forty years, then claims of an annual increase in the disease each year are clearly incorrect. In fact, Torres et al. (2012), in reviewing the research by van Hoeken et al., note that “an overall analysis of these and other studies led the authors to conclude that there was insufficient evidence of an increasing risk of AN during the 1980s compared to the 1970s” (9).
To help clarify the issue, I contacted a leading eating disorder researcher, Dr. Cynthia Bulik, director of the University of North Carolina Eating Disorders Program and coauthor of Decoding Anorexia: How Breakthroughs in Science Offer Hope for Eating Disorders. She told me, “We actually have fairly poor epidemiological data. Overall the prevalence of anorexia is fairly stable across time and Western populations” (Bulik 2012). Echoing Bulik’s assessment, recent data suggest that rates of eating disorders generally (and anorexia specifically) have remained the same or dropped in the past decade. A 2011 review found that patient hospitalizations with a principal diagnosis of an eating disorder dropped 23 percent in 2008–2009 (after peaking in 2007–2008), and that the percentage for all eating disorders “declined from 24 percent to 19 percent between 2007–2008 and 2008–2009” (Zhao and Encinosa 2011).
A closer look at the Lucas et al. (1999) study, which was apparently the source for the NEDA statistic that “There has been a rise in incidence of anorexia in young women 15-19 in each decade since 1930,” reveals instead that there is no data presented “since 1930,” and though the researchers note that “in both genders, the age-specific incidence rates were greatest for the age group 15–19 years” among the groups they studied, nowhere do they state or suggest that there has been a rise in incidence of anorexia among young women aged 15 to 19 that has been increasing each decade. The NEDA’s statistic—presented to the public as accurate, authoritative information about the incidence of anorexia—seems to be a Frankenstein-like amalgamation of phrases, categories, mis-read (or mis-typed) dates, and numbers taken out of context.
Elsewhere on its website on the “Anorexia: Overview and Statistics” page (https://tinyurl.com/og84zdb), the NEDA apparently again references the 1999 Lucas et al. study with this statistic: “An ongoing study in Minnesota has found incidence of anorexia increasing over the last 50 years only in females aged 15 to 24. Incidence remained stable in other age groups and in males.” Users of the website—laypeople and journalists searching for information on eating disorders in 2017—would reasonably assume that NEDA’s reference to anorexia increasing “over the last 50 years” refers to the past half-century, or about 1967 to the present (allowing, of course, for the fact that statistics on websites may be a few years out of date). However, as noted the Minnesota study covered fifty-five years (1935 to 1989) and offered no information about anorexia incidence over the past three decades. In fact, that study noted that “for females of all ages, there was a fall [in anorexia rates] in 1985–1989 from the highest rates observed in 1980–1984” (401).
In fact, contrary to the NEDA’s statistic, the Lucas et al. study states explicitly that “the rapid increase in anorexia nervosa seen from 1970 to 1984 has not been sustained” (403, emphasis added). Thus, even assuming that the findings of increased incidence of anorexia from 1970 to 1984 were valid, the authors state that the rate dropped or remained the same from at least 1984 onward. It’s not clear why the National Eating Disorder Association offers no data on the incidence of anorexia over the past thirty-three years (1984 to 2017); surely more recent studies have emerged showing that the disease has either increased, decreased, or stayed the same. As with other sources of misinformation about eating disorders, this NEDA statistic has been widely repeated and referenced in other eating disorder books, including in reference materials destined for middle schools (see, for example, Sonenklar 2011). An August 2017 Google search for that statistic yielded about 612,000 hits, mostly from reputable, academic sources.
As for NEDA itself, I asked NEDA specifically about the error and was sent more information, which also had the error. Kelly Williams emailed me: “Thanks for sending your findings. I forwarded to the person in charge of stats at NEDA.” That was in 2013. The wrong statistic still appears on the NEDA website today, years after I called attention to it.
In my research into eating disorder misinformation, I found many examples of flawed, misleading, and sometimes completely wrong information and data being copied and widely disseminated among eating disorder organizations and educators without anyone bothering to consult the original research to verify its accuracy. While this tendency is understandable and common, in many cases the information is made available through prestigious publishers. Peer-review and academic editors help minimize glaring errors by authors in scientific and medical journals, but mainstream book publishers are another matter. Editors for publishers such as Knopf and Random House are far more likely to employ fact-checkers for books on topics such as hard science than on social science. As with many subjects, eating disorder misinformation tends to creep into the public sphere where editorial vigilance is the lowest. Because expectations for factual accuracy and thorough scholarship may be relaxed for books written by nonspecialist writers about popular culture, opinion, and activism, misinformation is more likely to appear there than in eating disorder textbooks written by researchers and experts.
Many sources of eating disorder information feel that the responsibility to fact-check the information they provide falls in someone else’s purview: Scientific researchers, for example, rue the fact that journalists often misquote them but make little effort to correct the record. Writers and documentary filmmakers plagiarize statistics from other sources, assuming that someone somewhere must have verified the information. Because of this, the task of ensuring the accuracy and validity falls through the cracks.
The misinformation about eating disorders—spread by the National Eating Disorder Association—is unfortunately only one of many trusted sources of information that have spread myths instead of facts; others include the award-winning PBS television series Nova (see Radford 2015), bestselling books, and mainstream news media. For more on this problem, please see my master’s thesis, “Misinformation in Eating Disorder Communications: Implications for Science Communication Policy,” available for free at http://pqdtopen.proquest.com/pqdtopen/doc/1460763167.html?FMT=ABS.
- Bulik, Cynthia. 2012. E-mail correspondence with the author, May 7.
- Lucas, A.R., C.S. Crowson, W.M. O’Fallon, et al. 1999. The ups and downs of anorexia nervosa. International Journal of Eating Disorders 26: 397–405.
- Radford, Benjamin. 2015. Anorexia misinformation in the media: Case study of the PBS show Nova. Freethinking Blog (February 18). Available online at http://www.centerforinquiry.net/blogs/show/anorexia_misinformation_in_the_media_case_study_of_the_pbs_show_nova/.
- Sonenklar, Carol. 2011. Anorexia and Bulimia. USA Today Health Reports: Diseases and Disorders Series. Minneapolis, Minnesota: Twenty-First Century Books.
- Torres, Sandra, Marina Guerra, Filipa Vieira, et al. 2012. The epidemic of anorexia nervosa: Myth or reality?, Relevant Topics in Eating Disorders, Prof. Ignacio Jáuregui Lobera (Ed.) InTech, DOI: 10.5772/32134.
- Treasure, Janet, and Ulrike Schmidt. 2004. Clinical evidence-concise. British Medical Journal December: 256.
- Van Hoeken, Daphne, Jacob Seidell, and Hans Wijbrand Hoek. 2003. Epidemiology. Chapter 2 in Handbook of Eating Disorders, 2nd Ed, (Treasure, Schmidt, and van Furth, eds.)
- Zhao, Y., and W. Encinosa. 2011. An update on hospitalizations for eating disorders, 1999 to 2009. HCUP Statistical Brief 120 (September). Rockville, MD: Agency for Healthcare Research and Quality.