I used to think that eating disorders were an extreme diet (this was a very long time ago). I thought there were people who didn't diet, people who dieted, and people with eating disorders, all arrayed on a nice little continuum, one blending seamlessly into the next. I've since learned that scholars are still considering this question, albeit in a more academic, less hand-waving sort of way.
Eating disorders are currently classified into three major categories: anorexia nervosa, bulimia nervosa, and eating disorder- not otherwise specified. What we don't know is whether these categories are just humanity's attempts to corral people with eating disorders into three categories, or if these are, in fact, three different eating disorders. Nor do we know if eating disorders are dimensional ( ie, they exist along a continuum) or they are taxonic ( ie, they have their own category, sort of like humans are a separate species from chimps. We don't have a human-chimp continuum, the behavior of several of my relatives notwithstanding). Perhaps a better analogy would be the difference between a light on a dimmer switch (dimensional model) and a light with an on/off switch ( taxonic model). The methods for determining the difference between these use a lot of advanced mathematical models and some really fancy words--I'm not especially concerned in understanding precisely how the determination is made. Rather, I intend to look at whether we can make a distinction, and what that distinction is.
Like many new ideas in science, there is evidence in favor of both views. So let's start with a brief review of the evidence.
Eating disorders are dimensional
Thus far, one study has evidence that restricting anorexia exists along a continuum with "normal eating," although this study also found that bulimia and binge eating disorder do not exist along a continuum ( Gleavesetal, 2000a ). Perhaps one of the reason why restricting anorexia didn't appear quantitatively different from "normal eating" has to do with the prevalence of dieting behavior, which is more common than not in the college students surveyed in the study. In general, dieters have higher scores on the restraint/weight subsection of the EDI-2, but have similar psychopathology scores as non-dieters ( Lowe etal, 1996 ), which really helps muddy the waters.
Interestingly, the Gleaves study found that the binge/purge subtype of anorexia was much closer to bulimia nervosa than the restricting subtype of anorexia, indicating that binge eating and purging is, in fact, taxometric, whereas restrictive behaviors are more dimensional.
When looking at nonbehavioral eating disorder symptoms (such as fear of fatness, obsession with food/calories), researchers failed to find any sort of eating disordered category, which would indicate that yes, indeed, eating disorders exist along a continuum ( Tylka and Subich, 2003). A more recent study, building on this one, found that although eating disordered thoughts are common among people without clinical eating disorders, eating disordered behaviors are rather uncommon ( Miller, Vaillancourt, and Hanna, 2009 ), which has implications on what, precisely, is measured on future studies.
Eating disorders are taxometric
The evidence for both binge eating disorder and bulimia nervosa seem to indicate that these disorders exist as their own categories; that is, either you have them or you don't ( Lowe etal, 1996, Gleavesetal, 2000a, Gleavesetal, 2000b ). Of course, a diagnosis is rarely as simple as one of those notes you likely received in sixth grade that said "Do you like me? Check yes or no."* There are lots of issues still to be worked out, not the least of which is where do we draw the line?
Perhaps one of the most interesting studies found a middle of the road for this discussion: some eating disorder symptoms existed on a continuum, and some, such as binge eating, fear of fatness/compensatory behaviors, and drive for extreme thinness, did not ( Williamson etal, 2003 ).
At this point, the only consensus on the issue of eating disorder taxometry is the need for more research. Preliminary evidence suggests that bulimia and binge eating disorders are discrete symptoms; the evidence is less suggestive for restrictive anorexia. That being said, eating disordered thoughts are quite common, even if the behaviors are not, which could have significant implications on what is measured and studied in the future.
With the upcoming (and much-heralded) publication of the DSM -V in several years, these seemingly esoteric ideas could have a large impact on how we diagnose, treat, and prevent eating disorders.
Coming tomorrow: Classifying eating disorders, part two (What's temperament got to do with it?)
*Full disclosure: I neither sent nor received any of these notes.