The Weight Requirement, and Other Ways We Diagnose Anorexia

This is a post that’s been a long time in coming. Unfortunately, every time I start it, I get upset and then have to leave it alone. So here we have it: what’s wrong with how we diagnose anorexia. Now with less ragequitting!

Anorexia has a weight requirement.

Refusal to maintain body weight at or above a minimally normal weight for age and height, for example, weight loss leading to maintenance of body weight less than 85% of that expected or failure to make expected weight gain during period of growth, leading to body weight less than 85% of that expected.


Please, someone define what a normal weight is. Because I know this totally isn’t controversial. Using BMI? It doesn’t account for muscle mass, so you’re leaving out  anyone who’s particularly athletic. Using clothing sizes? Ahahahahahahaha. 

Even supposing we had some accurate scale of normal height-to-weight ratios, should a diagnosis rest on a weight? If I’m restricting caloric intake, and I drop under 85%… Bingo, I have anorexia. Ding ding ding and all that. The next day, I enter treatment, or I eat a particularly large meal, or retain a little water, and suddenly I’m at 86%. Have I suddenly developed a disorder that is markedly distinguishable from what I had on Monday? I think not.

The emphasis on being excessively thin is also in the second criteria:

Intense fear of gaining weight or becoming fat, even though underweight.

Anorexia does not just happen one day, when you stop eating and magically, everyone can tell! This, like lots of the criteria, seem to be waiting to diagnose after the disease has progressed significantly.* That’s dangerous.

This is particularly true for people who start at an above-average weight. We fetishize the idea of heavy people losing weight through whatever means possible, up to and including unhealthy ones. (Biggest Loser, anyone?) It’s going to take an especially long time for family and friends to notice such severe weight loss–and of course, they’ll probably praise and reinforce it along the way–and all the while, the disorder will become more entrenched, wreaking havoc on the body it inhabits.


In postmenarcheal females, amenorrhea, i.e., the absence of at least 3 consecutive menstrual cycles. A woman having periods only while on hormone medication (e.g. estrogen) still qualifies as having amenorrhea.

Defined as non-menstruation for three cycles, this is competing with weight measurement for least helpful criteria. For one, it isn’t useful for anyone on hormonal birth control (withdrawal bleeding is not the same as getting your period). For two, it’s a fairly useless metric. Only cisgender women who can ovulate, are not pregnant, and who have reached puberty, but not menopause can use it. (Also, you qualify for this criteria automatically if you’re on birth control. Say what?) Further, there’s not a lot of research suggesting that amenhorrea occurs in a standardized way, or that it represents the severity of deprivation. Luckily, this is heading out with the publication of the DSM-5.

Waltzing between diagnoses.

In the span of six years, I met criteria for…

Anorexia nervosa (purging type)

Anorexia nervosa (non-purging type)

Binge Eating Disorder (BED)
Note: this is a proposed conditional diagnosis, given separate status from BN and AN. 

Bulimia Nervosa (non-purging type)

Eating Disorder Not Otherwise Specified (EDNOS)

…that would be all but two of the ways one can have an eating disorder. This suggests that we may just be quantifying eating disorders incorrectly. Those qualifying for anorexia diagnoses are automatically going to fall into EDNOS as they recover and gain weight (and consequently, no longer meet the first criteria for AN). Should we rename EDNOS as “Anorexia in Remission?” No, because then you leave out the others who were diagnosed as EDNOS for other reasons, such as not qualifying for a bulimia diagnosis. If EDNOS is made up of lots of people with very different manifestations of disordered eating, can we do any useful research about the diagnosis? Will we be able to draw any useful conclusions? Probably not.

And why does it all matter? After all, society recognizes that refusing to eat is bad, right? It matters because the research doesn’t look at “a population of women who have refused to eat at some point”. It looks at “300 female patients who had been diagnosed with anorexia in the last calendar year”. That means how we assess treatment, how we examine the genetic basis, how we study the disorder is a product of how we describe it.

And we’re doing it wrong.

*Also, the face of anorexia, besides being almost always a thin woman, is always white and upper class. Incorrect, and a scary myth to perpetuate.