The Enigmatic Persistence of Anorexia Nervosa

Anorexia nervosa was first described in the medical literature in 1689 by Richard Morton. It has been over 300 years since then and AN continues to be one of the deadliest psychiatric disorders. If not treated early, it runs the risk of becoming deeply entrenched and highly resistant to treatment.

Moreover, established treatments for related disorders like bulimia nervosa and depression, such as cognitive behavioural therapy and antidepressants, are rather ineffective in treating anorexia nervosa. Finally, even if significant physical and mental improvements are achieved in treatment, relapse rates for older individuals (even those in their 20s) remain high.

What makes anorexia nervosa so persistent and so hard to treat in individuals who develop it, particularly if it is not treated soon after onset? Why is recovery so hard?

In this paper, B. Timothy Walsh outlines a model based on cognitive neuroscience that attempts to answer these questions:

Its central element is that this behavior has become habitual and is grounded in the neural mechanisms that underlie habit formation and persistence. This formulation is based on a substantial body of animal and human research, much of which is focused on the neural mechanisms underlying substance abuse.


I’ve often said that restricting to me feels like the “default state”: without continual conscious effort to consume more food, I can easily slip into restriction almost by accident. In a sense, restricting is a habit — a learned or acquired pattern of behavior that occurs without conscious effort and can be provoked by various stimuli (or triggers).

It is crucial to emphasize that this is not innate:

In order to lead to significant weight loss, the dieting behavior must be engaged in repeatedly, over weeks to months, and the studies of eating behavior cited above demonstrate that it becomes markedly fixed.


How does this habit develop?

If you have ever taken an introductory psychology course, you’ve undoubtedly heard of operant (or instrumental) conditioning. (Remember Skinner’s box?) Very simply, it is a form of learning whereby behaviour is shaped by its consequences. If a rat realizes that pressing a lever leads to a reward, such as food, it will learn to keep pressing the lever. If, on the other hand, pressing the lever leads to a punishment, such as shock, the rat learns to avoid pressing the lever. When it comes to this type of learning, we are not too different from rats.

Importantly, after the behaviour is acquired because it continues to lead to reward, there comes a point when the behaviour becomes almost insensitive to the reward. That is, the behavior may continue even after it no longer leads to reward. This is the basis of habit formation.

How does this relate to the development and persistence of anorexia nervosa?

In the model presented here, which is based on these principles of learning, it is hypothesized that the dieting behaviors of individuals with anorexia nervosa begin as goal-directed actions that lead to weight loss, which is highly rewarding […] Over time, the dieting behaviors are engaged in persistently and repeatedly and thereby become overtrained and habitual […] Once they have become established as habitual, the behaviors are highly resistant to change and are a critical element in the persistence of the disorder.


When a rat presses a lever and receives food, the reward is clear. But what’s the reward in restricting food intake? What’s the reward in, essentially, starving? Walsh outlines two main reinforcers that contribute to forming the restriction habit and maintaining the anorexic state:

  1. Positive reinforcementCulturally maintained association of weight loss with self-control, achievement and attractiveness. 
  2. Negative reinforcementRestricting food intake as a way to regulate or cope with negative emotional states.

Overtime, the very act of restricting food intake becomes reinforcing (acting as a secondary or conditioned reinforcer) due to its association with weight loss and/or the removal  (or numbing) of a negative emotional state.


Many individuals begin to diet (or at least consciously modulate their food intake) during adolescence. A time of great physical and psychological changes. A time, also, when a lot of psychiatric disorders emerge.

Adolescence is a rough time for many (puberty [and everything that comes with it!], gaining independence, having to decide what to do with your life — at least in the immediate future). It is also a time when individuals are most sensitive to reward (more than children or adults!). Can this increased sensitivity to rewarding stimuli at least partly explain why adolescents are prone to developing substance abuse issues, and, perhaps, similarly explain the development of eating disorders such as anorexia and bulimia nervosa?


In addition to the positive and negative reinforcements, as well as the increased sensitivity to reward, Walsh outlines several other factors that may lead to restricting becoming ingrained:

  1. Persistent and repeated dieting –>overtrained and highly practiced –> entrenched behaviour.
  2. Intermittent rewards = stronger reinforcerI’m not sure I buy the argument that “daily weight loss is not inevitable, and the receipt of social accolades for becoming thinner occurs only occasionally.” While this is certainly the case for some, I would argue that for others, the positive effects of feeling like you’ve accomplished something by eating less (achieved a goal) and the effects of numbing emotions are much more reinforcing.
  3. Onset of anorexia nervosa typically coincides with stress and we learn better when we are stressed. So, behaviours that are acquired during a time of stress are more likely to become habits than when they are learned at other times.
  4. Weight loss itself leading to compulsive, obsessive, and rigid patterns of thought and behaviour. The most cited example of this is the Minnesota starvation study, where healthy subjects developed symptoms of obsessionality much like anorexia nervosa patients following a period of starvation.


The model presented here is quite similar to some previous models proposed by others, particularly the model proposed by Michael Strober, who hypothesized that:

Individuals who develop anorexia nervosa have an underlying propensity to anxiety and to fear-learning and a greater resistance to its extinction than unaffected individuals. Like the model presented here, Strober’s model is based on relatively recent advances in our understanding of how behavior is acquired and is shaped by internal and external stimuli. It suggests that negative emotion, specifically anxiety, about weight gain leads to enduring patterns of dieting.

Walsh emphasizes that the model presented in the paper does not challenge or contradict previous models — instead, it expands on them by proposing specific mechanisms by which dieting becomes rewarding and ultimately habitual.


There are a few implications of this model, but I’ll just mention two that I think are most important:

One, we need to treat eating disorders early — before the behaviour becomes entrenched. (Yeah, I’m looking at you insurance companies and stubborn doctors. Get with it.) Although we knew this even without this model, the model (if true) certainly further emphasizes the importance of early intervention. The data is clear: early onset and short illness duration are some of the strongest predictors of a complete recovery. (But, my fellow adult readers, please don’t let this decrease your motivation or leave you feeling hopeless. These are statistical associations — don’t let them dictate your efforts.)

Two, effective treatments probably need to really tackle the ingrained restricting behaviours in order for them to be successful in the long-term. That doesn’t tackling restriction is sufficient — it isn’t. But it is, obviously, necessary. I think when restricting is viewed as a strongly entrenched and powerful habit, it helps clinicians, patients, and families understand why  “eating a sandwich” is so freaking hard, and why “eating a sandwich” has to continue for a long time before it becomes a habit that overrides the dieting habits of anorexia nervosa.


As of now, this is just a model, and although separate elements are well-established, the model as it applies to anorexia nervosa isn’t. So, next is the fun part (well, for a research scientist): testing the model. One interesting question that can be studies is the ability of anorexia nervosa patients to develop and extinguish habits when (1) underweight, (2) partially recovered, and (3) fully recovered. That would be an interesting study — though, I wonder, whether the findings would be generalizable to dieting behaviour.

Walsh emphasizes that this model addresses only one aspect of anorexia nervosa. It does not directly address vulnerability, the onset of binge eating, excessive exercise, and so on.  But, in my opinion, it is not hard to integrate this model into other models that do address those things.

I’ll conclude with a great summary of the model provided by Walsh near the end of the paper:

The developmental and psychosocial context in which anorexia nervosa develops sets the stage for weight loss and, over time, the characteristic dieting behaviors themselves to become intensely rewarding. Several additional factors lead to dieting behavior becoming a well-established habit and very resistant to change. This formulation thereby provides a possible insight into the impressive persistence of anorexia nervosa once it has developed. However, other factors, both innate and acquired, undoubtedly also contribute to its chronicity, including biological, psychological, and environmental influences that reduce emotional and behavioral flexibility.

So, readers, what do you think?


Walsh, B.T. (2013). The enigmatic persistence of anorexia nervosa. The American Journal of Psychiatry, 170 (5), 477-84 PMID: 23429750

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


Tetyana is the creator and manager of the blog. She has an Honours BSc in Neuroscience and an MSc in Medical Science. She can be reached at tetyana[at]scienceofeds[dot]org.


  1. Yeah, I’m definitely familiar with the concept of losing the reward but still sticking with the habit. It’s like…just the act of restriction itself becomes the reward, even if you don’t get anything out of it other than the comfort and satisfaction of engaging in the behavior. I suppose for me, even if there isn’t weight *loss*, preventing weight *gain* could be seen as the reward…

    And I agree with the importance of catching and heading EDs off early. When I had a previous bout of AN, it only lasted…hmm, less than a year, for sure. And it never got too severe, so when I decided I wanted to stop, to try to be “normal”, it was pretty easy. I was still in a mode where restricting was, if not difficult, then still work. I would have to specifically say “No, you are NOT going to eat that” and walk away from food. It was active rather than passive, I guess? So it was not hard to let it go and start eating normally again. Plus I had not lost much weight, so there wasn’t the physiological/brain effects to deal with.

    But this time…it went so much faster, I lost so much more weight in such a short period of time, there was no chance of someone else or myself stepping in “early” enough to stop it before it got so damn entrenched. Before my brain started crumbling and my body forgot how to function, and before the behaviors and the thinness became…requirements. I hate saying that, but it is how I feel.

    Um, anyway, I’m just babbling here, but I do think there’s a lot of truth to this model, and I just hope other folks do get that early intervention. Tetyana, you and I have talked about how not everyone “fully” recovers and how for some of us, management is a reasonable enough goal. But I do hope other folks can avoid getting to that place we’re at, and do have a chance at fuller recovery.

    • I agree on all points.

      Although, with regard to myself. I’m actually quite optimistic because I’m more on the bulimic spectrum than on the anorexic spectrum, and that’s (statistically) easier to treat and for me, mentally, much easier to deal with. I am totally cool with maintaining my current healthy weight. I’m not very rigid in my eating, and I don’t feel I’m in the mental prison of AN. I have a lot to work on, but I’m much more optimistic now than during my second relapse and significant weight loss.

      “But I do hope other folks can avoid getting to that place we’re at, and do have a chance at fuller recovery.”

      I hope so too.

      I also hope that countries like the US move to a more public health care system and that clinicians and health care workers start taking mental health, including eating disorders, seriously. I’m impatient… but I do know these battles with take a long, long, time.

  2. Tetyana

    Yes, yes, yes. Something I have been thinking and discussing in my back-handed non-scientific way for a couple of years now.

    Absolutely brilliant blog.

    xx Charlotte

  3. I really, really like this model – and this is a great blog post, Tetyana 🙂

    Attempting to recover from (restricting) AN is never easy, but in one’s 40s, it sometimes feels near impossible 🙁

    Despite having focused heavily on my eating behaviour (alongside some of the co-morbidities, thoughts and feelings that have accompanied my AN…) for the past 7 years, I still find eating enough really, really difficult. The urge to restrict is constantly present. I feel BETTER when I restrict: calmer and more ‘in control’ of my anxiety and my existence; more comfortable physically (I hate the feeling of fullness…) – and generally better.

    The problem is that restricting food causes me to lose weight – and I don’t want to lose weight. (For me, the objective of restricting and over-exercising has never been weight loss..). I have raised my weight by > 30 pounds over the past 7 years, but I am still underweight. My ‘habit’ of restricting was so difficult to break – and eating enough to just maintain my weight (let alone gain the extra 10 pounds I still need to gain) feels like torture at times.

    My AN did start during a period of intense and unimaginable stress (after being raped and repeatedly molested by someone outside of my family as a 10 year old) and both restricting food and over-exercising felt like ‘the solution’ to the intense anxiety. This behaviours pattern really did reduce my anxiety to manageable levels.

    Although I did have quite intense diet intervention as a child and teen, unlike some other kids with AN, who with better nutrition discover the joys of teen life, I could derive no satisfaction from such ‘joys’. I was terrified of boys and men in my teens; I imagined that they ALL wanted to rape me – and so I ducked out of teen life completely. After achieving a near healthy weight at age 17 – 18, I relapsed rapidly in my early 20s. *Sigh*

    For some kids, addressing eating behaviours (promptly) may be all that is needed. For others, with co-morbidities, including PTSD, additional therapies are required. I wish I had experienced such additional therapies in my teens. They might have made all the difference.

    • I agree with everything, particularly your last point. I think addressing co-morbidities is sometimes a prerequisite to making progress with eating disorder recovery (sometimes it is the reverse, of course).

      We’ve talked many times about how I can relate to your situation, so I won’t go into it here, but thank you for writing about it. I think it is important for others to read about the reasons that AN patients restrict their intake — the initial causes and the causes that maintain it (which need not be the same, of course).

      Thanks for your comment Cathy 🙂

  4. Thank you thank you thank you for this:

    “Intermittent rewards = stronger reinforcer. I’m not sure I buy the argument that “daily weight loss is not inevitable, and the receipt of social accolades for becoming thinner occurs only occasionally.” While this is certainly the case for some, I would argue that for others, the positive effects of feeling like you’ve accomplished something by eating less (achieved a goal) and the effects of numbing emotions are much more reinforcing.”

    I once read a philosophy paper describing anorexia as not necessarily a weight-loss oriented, vanity behavior, but rather a behavior to achieve a sense of “transcendence”. In other words, the ability to suppress natural urges for food is a sort of triumph of willpower. Perhaps the media salience of weight loss = good provides an easy cognitive framework to attribute these behaviors to, but it doesn’t explain the manifestation of the disease over 100s of years. Also, I’ve found that most individuals that I’ve met with anorexia tend to NOT adhere to cultural trends, something I believe you’ve highlighted in the past.

    In regards to the development of anorexia, involvement of reward circuits, and excessive exercise, did the author mention any of the activity-based anorexia studies in animals? Apparently wheel-running behavior quickly develops with severe food restriction as a means of anxiety relief, appetite suppression, etc? My guess would be the “foraging” tendency gets put into overdrive in these animals but, without access to food, they must find an alternative way to reduce that drive to seek out food. Perhaps a similar underlying tendency develops in humans?

    • Yeah, that was mentioned in the paper, too.

      I don’t think this played any role in why I started restricting, but when I was already struggling with it, I do remember feel better than others for not eating. Very much so. I had this smug pride of being “above” eating. I could deny myself food (and sleep when I was angry) and that somehow made me better. (Of course, it didn’t, it just made me sick, but it really felt that way and I had to tell myself that those thoughts are idiotic). But, I don’t think that’s what initiated it for me, not at first, anyway, because it took a while for those thoughts to ‘kick in’.

      Yeah, he explicitly mentioned that he was NOT going to talk about excessive exercise for simplicity’s sake. I don’t know how generalizable animal model systems are with regard to this because they obviously lack a lot of the top-down processing I think that occurs in humans that may also contribute. I don’t know. Not too well-versed on ABA.

      • Yeah, it’s SO incredibly hard to figure out how these things start. I had no intentions of losing weight to begin with. I became vegetarian (for ethical reasons) and extremely budget-conscious, thus inadvertently restricting my caloric intake. Add in even more stress from school (I already had anxiety issues) and BAM! There I was: obsessed with planning miniscule meals. I can’t figure out how much of my descriptions are post-hoc explanations or true contributors to the disorder. I just know I have to congratulate myself anytime I do a really good job of paying attention to nutrition; and, dangit, I DO feel proud when I eat some local meat and have tons of energy to run around and be the active person I used to be. (I also have a tendency to get a bit perturbed by those that flaunt their vegetarianism….something I’ve got to work on. People have good reasons for doing so…)

        • “I can’t figure out how much of my descriptions are post-hoc explanations or true contributors to the disorder.”

          YUP, neither can I.

  5. Great post, Tetyana- I haven’t read Walsh’s article yet but now I will get on that… very interesting model for sure.

    You wrote: “I’ve often said that restricting to me feels like the “default state”: without continual conscious effort to consume more food, I can easily slip into restriction almost by accident.” – YES. Absolutely. This is one of the things that I find hardest to explain to others, actually. Despite the fact that I do consider myself recovered, when I’m going about my usual days I pretty much follow the exchanges that I was set in the step-down program from my program, which gives me enough energy, makes me feel good most of the time and allows me to maintain my weight. However, some people assume that the fact that I still largely follow a “meal plan” (though much more loosely) means that my recovery is somehow less solid. On the contrary, I find it keeps me MORE recovered, because I would likely slip into undereating if I were to move to completely intuitive eating. This is what tends to frustrate me about the idea of “intuitive eating” in general, actually. What if when your “intuitions” about food were developing you were not feeding yourself enough?

    I also really liked the emphasis on the reward/habit forming elements; this makes intuitive sense to me and helps to move eating disorders out of the whole “she just wants to be skinny” over-simplification. In general, I think the model really helps to clarify the key difficulties in treating eating disorders and why this can be so incredibly hard, particularly in the case of longer-course disorders. I hope someone does integrate this model with others (or creates something with more factors) that include binge eating, vulnerability and excessive exercise, among other things.

    • “This is what tends to frustrate me about the idea of “intuitive eating” in general, actually. What if when your “intuitions” about food were developing you were not feeding yourself enough?”

      Seriously. I count calories for the same reasons you do the exchanges thing. I need to, because even after a long time of eating ‘enough’, I still fall into restricting if I don’t keep things in check. It is too easy.

      I agree with your other points :-).

  6. Hmm, this is very interesting. Thanks for the write-up, Tetyana.

    For me, it wasn’t losing weight that was rewarding, but simply restricting intake. Remarks about my thinness only ever made me uncomfortable, and since I was quite thin even before AN-onset there were no social accolades as I got thinner. Hell, I didn’t even know I *was* getting thinner. I was so out of it when I first got sick and started seriously restricting, at age 15, that it didn’t really occur to me that what I was doing would lead to weight loss. I ate less and less and moved more and more simply because eating and being still had become terrifying. I guess the compulsion was driven by Walsh’s negative reinforcement (“restricting food intake as a way to regulate or cope with negative emotional states”), but I didn’t experience it as such at the time.

    Anyway, I definitely relate to the idea of replacing restricting habits with eating habits, and replacing moving habits with sitting-on-my-butt habits (haha), as a big part of the work of recovery.

    • Thanks for your comment D,

      “I guess the compulsion was driven by Walsh’s negative reinforcement (“restricting food intake as a way to regulate or cope with negative emotional states”), but I didn’t experience it as such at the time.”

      I think part of why (and others) probably don’t is because it is a complicated concept (negative reinforcement), and at a young age, it is difficult to understand *why* restricting would make someone less anxious for example. So, the behaviour gets blamed on, explained in, or rationalized in different ways, by the patient and those around him/her.

      But, that’s just my hypothesis.

  7. Yes Tatiana, I think this is something as a patent and caregiver to a young adult, and so many patents I have shared this experience over time is that while probably the toughest part of treatment, is long term maintenance of a meal plan with full support of all professions involved. I’m not speaking to exactly how that’s accomplished but that it must be the gold standard 1st line goal.
    How that is accomplished may encompass ion different behavioral approaches but if everyone involved are working together the prognosis is most positive.
    If we can realize that the act of restricting itself contributes to the inability of the sufferer to be responsible fir maintains the plan for a longer period of time that supportive care should continue.

    • Thanks for your comment Lisa,

      I agree.

      If only good treatment options were available to everyone who needed them…

  8. Great write up. Thank you Tetyana. Something about the habit model makes great ‘intuitive’ sense. Also makes sense as to why a behavioral model like FBT can have the success it does in many instances. Persistent and repeated new behaviour (eating adequately and consistently) for an extended period of time allows for the ‘new’ behaviour to become ‘ingrained’ and habit forming. External support in the formation of the new habit was instrumental in making the new habit ‘stick’ in our experience. It was so obvious doing meal support that my girl couldn’t do it on her own. ‘Eating the sandwich’ was absolutely, positively ‘so freaking hard’. It is one thing when your mom can show up for lunch every day for a year (and make all other meals/ and sit/distract/watch/support)…it’s a whole other thing when you are a YA trying hard to work to recovery on your own. It takes such a looong time, as you mention, for the counter habit to become ingrained. We found it took repeated, consistent exposure (like ‘6x a day/ 3 meals 2 snacks, day in day out for an extended period of time’ kinda consistent exposure) for the new habit to really take hold. All well and good when you are 11 and living at home with built in meal support…not so simple when living independent. Truly great ‘follow up’ care and ‘transition from treatment’ support would recognize just how long it takes to establish the new habit and work to provide better mechanisms or processes to help with the ‘new habit’ formation.

Leave a Reply

Your email address will not be published.

19 − eleven =