Polar Opposites? The Social Construction of Bulimia and Anorexia Nervosa

Some might argue that bulimia nervosa is more “hidden” than anorexia nervosa — it is not always obvious that someone is suffering from bulimia (though, I would argue, it is not always obvious that someone is suffering from any eating disorder). Even when it is “discovered,” BN is often placed in opposition with AN — as if the two were polar opposites.

Indeed, attempts to define a phenotype (a set of observable traits or characteristics) for AN and BN tend to oppose the two and to suggest that the people who develop AN are inherently different from those who develop BN. While I believe there is some scientific evidence for personality differences between the two, the degree of diagnostic crossover and symptom variability in eating disorders makes me feel like this split is at the very least overly simplistic.

What is interesting is how BN has come to occupy a very different place in our collective social imagination than AN. We know that preconceived notions about what it means to be an individual with an eating disorder in general can have implications for things like treatment seeking (see here, here, and here) and feelings of legitimacy around eating disorders. We also know that misinformation abounds when it comes to the causes, presentation, and experience of eating disorders.

But how does society make sense of bulimia, and individuals with bulimia? How do people understand a disorder that expresses itself through behaviours that may or may not ever lead to visible changes in the body, when eating disorders are so frequently collapsed with their physical ramifications?

One of my favourite post-structuralist critical feminist writers (yes, I have favourite post-structuralist critical feminist writers; don’t judge), Maree Burns, has written about understandings of the bulimic bodies and how BN and AN can come to be hierarchically positioned in social understanding. I recently came across an article she wrote in 2004 (it is open access) that I’d like to explore here.

I do realize that the article is 10 years old; something that has struck me lately, however, is that we still seem to be grappling with similar issues in the eating disorder research field after years and years of debate. So, however unfortunate it may be, it would seem that conclusions drawn 10 years ago may still hold weight.

An important distinction

At the beginning of the post, I briefly touched on efforts to define a phenotype for AN and BN. To be clear, that is not the intent of the author of this article. Instead of looking at whether AN and BN are actually different, Burns looks at how AN and BN are understood to be different, and what impact this might have. She ties this into a broader discussion of the construct of femininity, particularly as it is articulated in Western societies.

This distinction is really important for interpreting Burns’ findings, and speaks to her theoretical leanings. As she notes in the article, using a feminist post-structuralist perspective allows us to locate meaning in discourse by looking at language used.

Rather than looking for some kind of truth about “what” or “who” individuals who binge and purge or individuals who restrict are, we are instead looking at how different ideas around these behaviours (and the individuals who practice them) come to impact their being-in-the-world.


For this study, Burns interviewed 15 women engage in bingeing and purging behaviours and 11 health professionals, including physicians, psychologists, psychiatrists, and a dietician, among others. She also looked for representations and coverage of BN in popular culture (e.g., TV, Internet, magazines), as well as in the psychological literature (e.g., scholarly articles).

In this body of data, Burns sought to tease out the overlap between a created dualism/hierarchy between BN and AN as represented in pop culture, “authoritative discourses” (e.g., experts’ opinions) and the reports of women with lived experience.


Control/Lack of Control

Burns discusses how BN has historically been constructed as a lack of control (e.g., Bruch, 1973) while AN has tended to be described as “over-control” (e.g., Jarman et al., 1997). From the outset, this seems to position individuals with BN as embodying a number of characteristics socially coded as negative. For example, seeing individuals with BN as lacking in control, they are positioned as “at the mercy of their compulsions” in contrast to a “disciplined” individual with AN.

Participants were well attuned to this dynamic, sometimes recounting AN as “the epitome of control.”

I think to get, to actually be an anorexic, you’d have to have like incredible, supreme control, and just to get to that place, something inside you must just push you there. Whereas I think with bulimia, it’s just (long pause) it’s a lot more common because it’s like a, a humans are fallible, that’s what they do. They indulge and they try to make up for it. (Becca)


Participants (both those with lived experience of BN and practitioners) also placed value on the ability to abstain from eating, sometimes describing individuals with anorexia as “overachievers” or “driven.” Burns argues that this reflects a broader value placed on control and success in individualistic culture; one practitioner even described BN as “failed anorexia” (which seems offensive, honestly, but there you go).


The very “doing” of BN could be described as socially disruptive, Burns suggests. Her participants described BN as more of an active, physical endeavour:

[…] bulimia is is a more um, is a more physical thing, you actually have to physically go and find somewhere that you can throw up your food and you have to do it and then you have to sort of . . . come back in and carry on doing what you were doing. Sometimes you’d be in company and other times you wouldn’t, whereas anorexia is . . . I mean, there is no physical doing, you just, you just don’t eat. (Rosie)

Burns links this kind of categorization to a historical tendency toward the construction of female bodies as “sites of unwellness and infirmity.” The focus, in popular culture (both presently and looking back into history) is less on concern for the individual sufferer’s well-being and more on the shock value associated with the symptoms bulimia might entail, for example “gorging” on rich foods or “taking”/stealing food to binge.


Tied in to this fascination with women’s bodies and behaviours that circulates in popular culture and psychological discourses are links (subtle or explicit) to sexuality. Looking at the DSM, individuals with BN are described as being more likely to have impulse control problems and to be sexually active. Throughout pop culture and (perhaps more alarmingly) in clinical accounts of BN, there is a contrast between bulimic “bad girl sexuality” and anorexic “good girl sexuality.”

A psychologist Burns interviewed for this study articulated this linkage between appetites for food and for sensual pleasure:

She [friend with bulimia] just loved to eat. She was a very sensual person and she loved to eat and she liked to to sleep with strange guys (laughs) and you know I mean she she she / MB: yeah / she um she she really lived and and she couldn’t bear putting on weight so she used to vomit.

What I find most interesting about this particular area is that “really living” or having an appetite (for food, for sex, for life) could also be seen as positive; however, it becomes negative through clinical discourse when described as “promiscuity” or lacking in control. The degree to which this reflects our social understanding of femininity and sexuality is striking.

Dualism and Severity

Maybe there is nothing inherently wrong with the oppositional categorization of AN and BN. But what does this say about our understanding of the severity of the two? Further, what implications does this have for individuals suffering from either AN or BN (or for those who experience diagnostic crossover)?

Burns suggests that this dualistic construction informs understandings of the severity of AN vs. BN such that AN is understood as deviant by way of the appearance resulting from practices and the associated potential outcomes of these practices (e.g., ultimately, death). On the other hand, it is the practices themselves that are seen as deviant in BN.

While either AN and BN could result in death, AN may more obviously represent, to the gazer, “closeness to death.” The interpretation of the bodies of individuals with eating disorders explicitly informs others’ understandings of the severity of their practices.

So, “bulimic practices” are seen as deviant while “bulimic bodies” may fit the norm; on the other hand, “anorexic practices” seem to fit the norm (until a certain point) and may even be reinforced/praised/rewarded, while “anorexic bodies” hold a “shock value” and are read as deviant.


Burns also suggests that “the bulimic body” is more readily tied into dominant historical and cultural notions of femininity. To succinctly and perhaps inexpertly summarize Burns’ insights into this dynamic:

  • AN reflects a commitment to a “mind in control of body” that one would expect of accounts of the (rational, controlled) masculine body
  • BN is more explicitly linked to the physical, suggesting that the bodily appetite is in the driving seat
  • This “body ruling over the mind” is more in line with historical conceptualizations of female (irrational, uncontrolled, dangerous) bodies

There are nuances to this analysis that would make this post excessively lengthy if I were to really delve into them, but I also think it is important to note Burns’ acknowledgement of the paradox of understandings of anorexia: while AN in some ways reflects a “traditional” (passive) femininity, the mind-over-body control is more clearly linked to a historical understanding of masculinity.

With this, Burns also explores how the idea that AN and BN have culturally and historically occupied binarized positions, this does not mean there is one way of understanding AN and BN; AN in particular occupies an ambiguous place between positive and negative, normative and deviant. BN, on the other hand, has tended to more consistently occupied an abnormalized (but more invisible) place in the cultural imaginary.


Ultimately, Burns suggests that this dualistic construction of AN and BN can prevent women from being able to express different elements of their subjectivity. Perhaps especially for those who have experienced diagnostic crossover and have thus experienced both “anorexia-like” and “bulimia-like” symptoms, the hierarchization and dichotomization of BN and AN may lead to confusion about “what kind” of person one is, and who is the “real self.” This reflects the very real consequences of being labelled “anorexic” or “bulimic”; as one participant in this study reported:

I don’t know how the atmosphere had been created but I know that it felt right away that there was some kind of hierarchy between those labelled with anorexia and those labelled with bulimia and that anorexia represented this more kind of achievement of perfection and it was a cleaner disorder because you weren’t throwing up and there were just all of these things that um made – made that category. I mean that’s the ultimate achievement of anorexia is to kind of have it perfectly / MB: mmm / um . . . and I was labelled with bulimia and right away um the problem grew really really strong and difficult for me and decided that it needed to be called anorexia and um I went rapidly down hill in all ways and I I was in this supposedly very supportive environment but things took such a turn for the worse that I was um you know an inpatient in a matter of months in a couple of months / MB: right / and that was not being addressed at all in the program I felt like I couldn’t speak about it at that time but it was there was some kind of competition happening there that was really unhealthy.

Burns also suggests that a different understanding of AN and BN could facilitate a less pathologizing understanding of behaviours, bodies, and subjectivities. For example, a more fluid construction might place eating disordered behaviours on a continuum and sometimes engaged in simultaneously, rather than fixed at either end of a spectrum.

In the years since the article was written, I believe there has been a move toward a more “continuum-based” understanding of eating disorders. However, a polarized understanding of eating disorders does still circulate, particularly when you look at popular culture representations of eating disorders.

Something I see a lot is a lack of recognition of diagnostic crossover and the subtypes of AN; for example, it is far more common to see someone described as having anorexia and bulimia than to see them described as having “anorexia, binge-purge subtype,” in popular culture. I wonder about the impact of this kind of framing on general understanding of eating disorders.

Importantly, this article reveals the potential negative impacts that creating a hierarchy around eating disorder diagnoses can have. Further, I would suggest that it highlights the importance of not judging individuals’ eating/health status based on their appearance. I’m likely preaching to the converted, here, but it is an important point to reiterate.


Burns, M. (2004). Eating Like an Ox: Femininity and Dualistic Constructions of Bulimia and Anorexia Feminism & Psychology, 14 (2), 269-295 DOI: 10.1177/0959-353504042182

Liked it? Take a second to support Andrea on Patreon!


Andrea is a PhD candidate focusing on individual, familial, and health care definitions and experiences of eating disorder recovery. She has an MSc in Family Relations and Human Development and a BA in Sociology. In her Masters research, she used qualitative and arts-based approaches (digital storytelling) to explore the experiences of young women in recovery from eating disorders. Andrea has recovered from EDNOS. She can be reached at andrea[at]scienceofeds[dot]org.


  1. First: This blog name is a misnomer.

    “Indeed, attempts to define a phenotype (a set of observable traits or characteristics) for AN and BN tend to oppose the two and to suggest that the people who develop AN are inherently different from those who develop BN. While I believe there is some scientific evidence for personality differences between the two, the degree of diagnostic crossover and symptom variability in eating disorders makes me feel like this split is at the very least overly simplistic.”

    But that depends on what you consider “scientific evidence” and how you define anorexia nervosa and bulimia nervosa. Yes, you’ll find group differences if you take individuals with AN who’ve never crossed over and those with BN who’ve never had AN or crossed over and have been sick for >5 years (when chances of crossover are greatly reduced), but there’s a lot of middle ground where the vast majority of people sit. It is also unclear to what extent those differences are premorbid or the result of the illlness.

    So, I kind of disagree with that there’s scientific evidence for the personality differences between the two. There are a lot of methodological issues with personality studies, particularly in the ED subpopulation, and while I don’t disagree that there are differences when we are looking at the extremes, we have to remember that those make up the minority of the ED population.

    What attempts are you referring to, by the way? Whose attempts?

    I can just SEE Cathy rolling her eyes at the mention of Bruch, and I’m right there with her.

    How old where the people she interviewed? How experiences were the clinicians? (Young and inexperienced, hah?) That one professional with a friend with BN…. seriously. That person is a psychologist? Oh jeez.

    Also: “AN reflects both a “traditional” (passive) femininity, the mind-over-body control is more clearly linked to a historical understanding of masculinity.” This sentence doesn’t make sense to me. I feel it is missing a conjunction or something.

    • First: I’m not sure what you mean?

      Second: True, it does depend what you mean by scientific evidence. I should have been more specific and said that there have been some studies that have looked at differences and found a few correlations. Good point about the difficulty in determining whether the differences are premorbid or result from the illness. I think that there has been a recent turn toward understanding the great deal of diagnostic crossover and number of people who sit in the “spaces in between,” so it isn’t easy to polarize “anorexic” and “bulimic” personalities. It’s just not that simple. I think that the personality type/ED type divisions have been more so taken up in the popular literature than the scholarly literature, perhaps. The commentary around the polarization of AN and BN reflects an uptake in both the popular and scholarly literature of a relatively simplistic binarization of people with AN and people with BN; not saying that all/even most researchers/clinicians/people see the two as lying on completely opposite ends of a spectrum, but it happens. This could have improved over the past 10 years since the article was written. I have a lot of issues with the new DSM categories, for example, but the discussions around revising the categories therein seemed to take a more continuum-based understanding. By attempts I mean studies looking at, for example, personality traits & eating disorders, traits & eating disorders, etc.

      Three: Yes, Bruch is outdated. I don’t agree with the way she framed eating disorders, and I don’t think that Burns does either; understanding her framing of eating disorders, however, is important when looking at why some myths surrounding BN (and AN) continue to be perpetuated despite the common (in the research/clinical/skeptic fields) debunking of her ideas. Unfortunately, they’ve informed the socio-cultural imagination around eating disorders, especially for those who haven’t read critiques of her work.

      Four: She writes that most of the women with lived experience of BN were between the ages of 25-35. She doesn’t say how long the clinicians had been working with individuals with eating disorders, however, all of them were actively working in the field when they were interviewed. It is for sure somewhat alarming that these were the kinds of comments they had… but do you really think these perspectives are one offs, or that the participants were being honest about their experiences and perspectives? People stigmatize. They relate things to personal anecdotes. They project their own insecurities onto others.

      Re: the sentence, it’s a typo. It should read: “While AN in some ways reflects a “traditional” (passive) femininity, the mind-over-body control is more clearly linked to a historical understanding of masculinity.” Will edit.

      • My comment on the misnomer was a bit tongue-in-cheek. It is just that the content posted on the blog now-a-days (which I like, of course, otherwise I’d tell you to stop haha) is not at all what I envisioned when I first started SEDs. I don’t consider the above science (but I don’t consider a lot of what I’ve blogged about to be science either), but alas, the name is what it is. I’m not going to change it now. Something like “Eating Disorder Research” would be more applicable, I think.

        “I should have been more specific and said that there have been some studies that have looked at differences and found a few correlations.” Yes, this is definitely true.

        “I have a lot of issues with the new DSM categories, for example, but the discussions around revising the categories therein seemed to take a more continuum-based understanding. By attempts I mean studies looking at, for example, personality traits & eating disorders, traits & eating disorders, etc.” The issue is that the DSM categories are made by researchers who want to categorise things. Clinicians (good clinicians) know that the picture is much more complex than the DSM will allow but everyone wants to study a clean sample. It is difficult to look for genetic contributions, for example, in a heterogenous population. Then you don’t know what the hell you are looking for or will find. So I think it is important to understand the competing interests. Scientists and researchers want to categorize; it helps them do science — to an extent. (It also obvious hampers it when you only look for what you defined a priori and exclude other potentially relevant things, like non-fat phobic AN, for example.) But a continuum, while more realistic and relevant for clinicians and people’s experiences, makes research very hard because, well, you are introducing a lot of potential confounds and (causal) relationships become ever hard to decipher.

        Yeah, I agree. Freud is also outdated but we still seem to teach him to PSY100 students. Sigh.

        I wrote the comment and proceeded to check the ages :-). “It is for sure somewhat alarming that these were the kinds of comments they had… but do you really think these perspectives are one offs, or that the participants were being honest about their experiences and perspectives? People stigmatize. They relate things to personal anecdotes. They project their own insecurities onto others.” I have no idea. According to the research I’ve read and blogged about, no, it is unfortunately not a one-off. My experiences are obviously not typical, being in Toronto and all. (I believe Toronto General Hospital had the first ED inpatient treatment program or one of the first? And we have SickKids, too.) Like Michael Strober said at ICED2014: Never underestimate the incompetence of mental health professionals. I’ve met some pretty incompetent social workers and therapists.

        • Ah ok, I didn’t know if you meant the blog name in general or the name of the post! As I said on Twitter, I guess it depends on how you define science… maybe “Science (and Social Science) of Eating Disorders” 😛 Its kind of interesting how the blog has evolved from what you thought it might be, originally. Hopefully my feministing hasn’t detracted too much from its original aims! Its fun to have debates between various perspectives on both science and eating disorders- I never learn more than when I talk to people who take a different approach to research/writing/learning/life than I do.

          I totally agree re: the DSM being in existence largely due to a clinical need to classify things; I think it needs to be more explicit (somehow? somewhere?) that its categories need to be taken with a grain of salt and are not wholly indicative of the boundaries of possible symptoms/behaviours/affect/cognitions people could show, and is not the be all end all decider of how serious a disorder is. Unfortunately, despite critiques, it continues to exist as this giant monolith/the ultimate authority to some, I think. The newer version does take more clinical variation in diagnosing into account though, I think. I’ll never be the biggest fan though. This is (one of the reasons) why I really like conducting my research from a place where I can not base my sample on clinical diagnosis/clinical “recovery” (though that’s a whole other bag of worms), because it allows me to put the participant’s experience first… of course, it then limits the ability to generalize and/or be taken seriously, but hey, there are drawbacks to everything.

          Whoa-k, hope that rant/ramble made sense.

  2. Thank you. I think this is a really important subject, and one whose point is often lost in arguments over ‘evidence’ to support / deny the existence of phenotypes (as you said). Both clinicians and friends have assumed very different things about me depending on which diagnosis / symptoms I have disclosed (that anorexia and purging are mutually exclusive, or that b/ping is due to a compulsive personality etc). An understanding that a particular practice / behaviour (e.g restriction, purging) can have different meanings each time it is enacted would prevent this binary construction of AN and BN and of people with bulimia and anorexia as different types of people. ‘Restriction’ for me is often qualitatively different when part of AN-R / AN-BP / BN but I don’t feel that what it says about me or who I am changes because of that. I think a more bottom-up approach, as per the bi-polar classification system to categorising EDs would prevent some of the hierachy and dichotomy between diagnoses i.e a diagnosis of and ED + specifiers for restricting, binging, purging, low-weight as relevant. SEDs = accurate (=/= misnomer) as science is used to say things about x that affect how people experience x?!

    • I’m glad you enjoyed the post! I agree; I think sometimes the argument about whether there is or isn’t a common phenotype for AN and BN obscures some of the other factors/struggles going on here, like hierarchization/binarization based on observations about symptom types and body shapes/sizes. I totally agree, too, that binging , purging, restricting and other behaviours can hold vastly different meanings based on context and that this should be taken into consideration when looking at “eating disorder pathology.” Thanks for commenting!

    • Just to respond to what I meant by the misnomer: I was half-joking. It is just that when I started the blog two years ago, I didn’t envision writing articles like this (well, and I don’t, thankfully Andrea does better job than I ever will) and while I’m super happy to have them here and have her contribute, it is not science, hence, I feel the name is a bit of a misnomer. But it is okay. It is not a big deal. I’m not going to change the domain name anytime soon.

  3. I would like to comment on the hierarchal differences between anorexia and bulimia as far as I’ve seen it in my clinical experience.

    I presented as EDNOs – purging type anorexic I believe was the diagnosis that they gave me at the time. I had a close friend in treatment who they diagnosed as purely bulimic. Those who more heavily restricted were placed in inpatient or an all day outpatient program while those who presented with predominantly bulimic symptoms and who usually were at a higher weight were placed in a 2 hour outpatient program.

    I found this very frustrating as my close friend was doing very terribly, much more so than me at the time, but because she was a higher weight and did not restrict to a high degree could not receive more intensive treatment. The clinicians did not appear to take her issues very seriously and many of the bulimics were put through treatment as quick as possible because there wasn’t really a weight gain goal and that appeared to be all they could really do rather than resolving behaviors. The frustrating thing was that this did present as a hierarchal concern because all of the attention and resources were placed on those who restricted rather than those who purged. There was a lack of attention on behavior resolution and the entirety of focus was on weight. I personally saw many bulimics who cycled through the program feeling shitty about themselves and their disorder and who later heavily restricted, losing a lot of weight and THEN could get attention and treatment. There was a lot of internal shame and guilt at not having a “real” disorder by the clinicians until it got significantly worse.

    I don’t really know what others experiences of this are, but I’m assuming it’s fairly common in terms of how the disorders are stigmatized and by how insurance companies bill ED treatment and its ties to weight as a major indicator of a “real” issue.

    • Thanks for commenting, Lauren, and sharing your experiences. It does sound really frustrating that your fiend was not given the level of care that would have been appropriate for her given the severity of her symptoms. Its upsetting that clinical decisions would be made more around body weight that symptoms/distress. I can’t see this doing any favours to individuals seeking treatment who might already feel “not sick enough.” I remember feeling this way when I was in treatment for EDNOS, and the literature speaks to this dynamic, as well. I hope that both you and your friend are doing well despite these upsetting circumstances!

  4. This is fantastic. I mean, both the paper and your post. I relate so much to this that it’d probably take me weeks to write down all my thoughts on this. So I’ll make this very short (hopefully). The paper was written in 2004, that’s right, but these social constructions of BN and AN are still the norm where I live, and I lost count of how many clinicians I met who assumed this kind of framework. When I was struggling with fears of diagnostic crossover, I was once told by a therapist that it was just “understandable” (!) that I had developed AN and that it was unlikely that I’d develop BN, as I’m (supposedly) a very “rational”, “disciplined” person who’s interested in “abstract” issues, etc. Needless to say, I felt so uncomfortable that I couldn’t remain still. There it was right in front of me the grotesquely sexist and oversimplified construction of mental illness: on the one hand, me, the guy with AN – disciplined, controlled, abstract, rational, all the characteristics that our cultures came to identify with masculinity; on the other hand, the bulimics – embodied, sensual, concrete, incontrollable, the aspects that we came to identify with femininity. Add to that the issue of weight stigma surrounding these constructions, and you have a pretty explosive mixture of dangerous polarizing notions.

    What really infuriates me sometimes is that this kind of construction and the obvious hierarchical structure that it implies does absolutely zero to help us either understand or treat EDs. It’s a falsification both of the clinical reality and of the scientific evidence (admittedly preliminary, insufficient, ambiguous, etc) about EDs, in my view. Not to mention that these constructions can have a stigmatizing effect that is way too harmful. I’ve seen this myself so many times – both in my treatment and in the lives of friends with EDs – that I really don’t believe it’s simply a matter of coincidence. Rather, it is something inherent in the way EDs are often presented, understood and dealt with in the culture where I live.

    Damn, I ranted. Sorry. : )

Leave a Reply

Your email address will not be published.

7 + seven =