Understanding eating disorders: when eating becomes a fight

There is an evening that many people know in one form or another. At the table people talk and laugh, someone refills a glass, someone takes a second helping. And in one head at that same table a second soundtrack is running. It calculates. It compares. It is already planning what has to go differently tomorrow.

From the outside this looks like nothing. Sometimes it even looks like a role model. Like good posture, like consistency, like someone who has it together. It happens that this person gets a compliment for it.

From the inside it is a fight that never takes a break.

I want to place three things side by side here: what an eating disorder is, why it develops, and what happens in the body along the way. At the end this gets honest about the limits of this blog. Because health content can bring order. It can also make things narrower.

One note first. You will find no weight figures here, no calorie values, no description of behaviours that could be copied. That is a deliberate decision and follows the recommendations for safe reporting on eating disorders.

It is never only about the food

“Pull yourself together.” “Just eat something.” “You can see for yourself that this is not doing you any good.” These sentences are almost always well meant. They come from people who have to watch and cannot do anything. And they still miss, because they carry a false assumption: that someone here is making a decision.

In both large diagnostic systems, the ICD and the DSM, eating disorders sit among the mental illnesses. Not among the nutritional problems. That is not a formality. It describes an experience that largely escapes voluntary control, in the way a panic attack escapes it.

And it is common. A systematic review from 2019 gathered all prevalence studies of the years 2000 to 2018. The weighted lifetime prevalence was 8.4 percent in women and 2.2 percent in men. Across the study period the point prevalence rose from 3.5 to 7.8 percent. Roughly one man in fifty is affected in the course of his life, and it is exactly these men who get asked last.

On top of that comes a second, quiet burden. Two meta-analyses from 2023 found suicidal thoughts in about half of those affected, with very large heterogeneity between studies. The risk was also raised compared with other people living with mental illness. I mention this as an order of magnitude, not as a precise figure.

And now you know why the sentence “just eat something” does not simply pass without consequence. It can create shame, and shame can make talking about it less likely.

The forms: what hides behind the names

Almost everyone who looks into this topic first searches for themselves in a grid. Do I fit in there? Am I sick enough? That question almost always leads astray. “Sick enough” is not a medical category.

The forms are still useful, because they describe different cores and point to different treatments. A clinical review from 2020 stresses that the diagnostic map has widened. Anyone who knows only anorexia and bulimia overlooks two pictures with a considerable share.

And what about orthorexia?

Orthorexia describes a preoccupation with eating that is very strongly oriented towards purity and correctness. The term comes from Steven Bratman, from the year 1997. And to this day it is not a diagnosis in its own right in the ICD or the DSM. That is not a detail for specialists, it matters for you when you come across self tests online.

A narrative review from 2019, co-written by Bratman himself, came to a clear finding: no commonly shared definition, criteria still under debate, questionnaires with methodological weaknesses. Where criteria were proposed, all of them named three points: obsessive preoccupation with healthy eating, emotional consequences when a rule is broken, and a noticeable impairment in life.

And now you know why the question “which form do I have” is less useful than the question “how much room is this taking up right now”.

Where it comes from: genes, temperament, experience

One of the hardest questions parents ask in a consultation is: what did we do wrong. I have no comforting phrase for that. I have data, and the data tell a different story than the question of blame.

Let us start with predisposition. A Swedish twin study with more than 31,000 twins estimated the heritable share of susceptibility to a narrowly defined anorexia nervosa at around half. The confidence interval was very wide, so this is an estimate and not a measurement.

In 2019 a finding came along that shifted the view of the illness. The largest genome wide association study on anorexia nervosa so far included 16,992 affected people and more than 55,000 comparison persons. It found eight risk regions in the genome, and the genetic correlations reached not only into psychiatric traits but also into metabolism, including blood sugar and blood lipids, independent of the known gene variants for body weight.

That applies not only to anorexia. A methodological overview of family, twin and adoption studies describes consistent genetic contributions in bulimia nervosa and binge eating disorder too. How individually metabolism responds is covered at two places in this cluster in more depth: why a calorie inside your body is not a calorie and what role genes and ancestry play.

The second building block is temperament. A meta-analysis from 2023 with 95 studies and 32,840 participants found a consistent association between perfectionism and eating disorder symptoms: perfectionistic concerns correlated at r = 0.33, in the clinical subgroups at r = 0.40. Cross sectional data, so association and not cause.

The third building block is banal and the most uncomfortable one. An Australian school cohort followed 1,699 adolescents over three years in six survey waves. Strict dieting behaviour was the strongest single predictor of a newly occurring eating disorder, with an approximately 18-fold raised risk compared with non dieters. After adjustment, neither body weight nor amount of exercise nor sex was predictive.

And now you know why an eating disorder almost never has a single cause. It arises where a predisposition meets a temperament and a trigger comes along.

The Minnesota experiment and the reversed order

“Why do I think about food all day?” I hear this question often, almost always framed as a question of character. As though circling thoughts were proof of a lack of maturity. There is a very old study that shifts this reading.

Between November 1944 and October 1945, Ancel Keys and his colleagues ran an experiment at the University of Minnesota. Thirty six healthy young men, conscientious objectors, selected for physical and psychological stability, lived for months on markedly reduced food intake. The aim was practical: they wanted to know how starved people in Europe could be fed again after the war.

The finer observations are not in that historical account but in the original monograph by Keys and colleagues from the year 1950: rituals around meals, very slow eating, the collecting of recipes, social withdrawal and an almost complete mental occupation with the topic of food. I name the source because these details often circulate without attribution.

This reversal has a biological counterpart. A meta-analysis of MRI studies found grey matter reduced by 8.4 percent in acutely ill adolescents and by 3.1 percent in adults. After one and a half to eight years of remission, no meaningful difference was measurable any more in adults. For adolescents robust long term data are missing, so restraint remains appropriate here.

A third pointer, which I mention only as an observation. Three severely ill women received recombinant leptin over a few days in an uncontrolled case series, a hormone from fat tissue that drops sharply in underweight. In two of them the drive to move and the circling thoughts decreased, in all three mood improved rapidly. Three cases are not evidence, and the substance is not approved for this purpose. More on leptin and insulin and on the control loop of hunger and satiety is covered elsewhere.

And now you know why the question “why do I think about food constantly” is often not a psychological question but a physiological one.

What the body carries: hormones, bone, heart

There is a phrase that turns up often in medical letters and is frequently misunderstood: “physically unremarkable”. It means only as much as the examination behind it. In eating disorders the physical level is not a side stage but an ordered pattern of its own.

A large review by Madhusmita Misra and Anne Klibanski describes this pattern as a whole. Under sustained energy deficit the organism throttles reproduction and growth in order to secure basic functions. Menstruation can stop because the hypothalamus dials down the control. An acquired growth hormone resistance with low IGF-1 develops, cortisol sits relatively raised, leptin, insulin and amylin fall, ghrelin, peptide YY and adiponectin rise.

These are exactly the axes that stand at the centre of clinical psychoneuroimmunology: the HPA axis, the HPG axis, the thyroid. They all shift in the same direction, and not by chance but following a sensible logic: economy mode.

Bone writes the longest bill. Low bone density in anorexia nervosa goes along with altered microarchitecture, lower strength and clinical fractures. The uncomfortable message from the same review: recovery may improve many hormonal changes, but the catch up process in bone can stay incomplete. In the cited review, transdermal estrogen raised bone density in adolescents, and full catch up still did not follow.

In bulimia nervosa the medical danger sits elsewhere. A review by Philip Mehler and Melanie Rylander describes the complications as a direct consequence of the type and frequency of compensatory behaviour, and some of them can be very dangerous. I deliberately do not describe these behaviours here. The decisive point: it is about electrolytes and about heart rhythm. An unremarkable body weight is no all clear signal.

And now you know why an eating disorder as a rule needs two people treating it: someone for the psyche and someone who watches heart, electrolytes, hormones and bone.

Refeeding is not a side stage

“She is finally eating again. Now it will get better.” I hear this sentence from family members, and I understand the relief. Medically, though, a delicate stretch begins here.

The mechanism can be explained in a picture. After a long deficit the metabolism switches to economy mode. When energy arrives again, insulin rises, and insulin shifts phosphate, potassium and magnesium out of the blood into the cells. Imagine a house in which all the lights were off for months. If someone suddenly flips every switch at once, the old wiring does not take that easily. That is why blood levels can drop, and that is why this concerns heart rhythm, muscles and the nervous system.

Three additional lenses go with this, each with an honest classification. First zinc. In a small controlled trial from 1994 with 35 women treated as inpatients, weight development under zinc ran faster than under placebo (p = 0.03). A single small paper, high dose, and not repeated in comparable quality since. I mention it as an observation, not as a recommendation.

Second the gut. A study with 16 patients found lower diversity of gut flora at admission and at discharge than in healthy comparison persons. So the gut stayed altered even after weight gain. Sixteen people are very few, and an association is not a cause.

And now you know why the return to regular eating can be the medically most demanding moment and not the easiest one.

What carries treatment, and what is not enough on its own

The most important question is: what changes something. The evidence here is clearer than you might expect, and more modest than you would wish. Both belong together.

The German S3 guideline on diagnosis and treatment of eating disorders, AWMF register number 051-026, is unambiguous: the best evidence exists for psychotherapy, as first choice treatment in all three main forms. In adolescents with anorexia nervosa the family should be actively involved. The current version can be viewed via awmf.org.

In bulimia nervosa and binge eating disorder, cognitive behavioural therapy in its enhanced form has the broadest support. A randomised trial with 154 outpatients tested a focused variant against an enhanced one that additionally works on mood intolerance, perfectionism, self esteem and interpersonal difficulties. After 60 weeks, 51.3 percent of the sample had eating disorder features close to the population average, and the result did not depend on the diagnosis. People with pronounced underweight were excluded.

A meta-review from 2022 summarised all meta-analyses and network meta-analyses. Family based approaches surpassed the active control condition in anorexia nervosa, in adults with bulimia nervosa individual cognitive behavioural therapy had the broadest effectiveness, and in binge eating disorder psychotherapy and lisdexamfetamine surpassed the control. One finding by the authors belongs here just as much: treatment effectiveness is moderate overall, and the guidelines diverge from one another.

And on the question of whether this is worth it. A cohort at Massachusetts General Hospital followed 228 women over a mean of 22.1 years. After 22 years, 62.8 percent of the anorexia group and 68.2 percent of the bulimia group counted as recovered, while after 9 years it was only 31.4 percent in the anorexia group. About half of those not yet recovered at that time became so later.

If you are looking for a medical counterpart for the physical side, an appointment can be booked below this article. The more important step, though, remains connecting with psychotherapy, and as a rule that should come first.

And now you know why the order has to be right: psychotherapy first, medical support in parallel, nutrition professionals in the team and not on their own.

When health content itself becomes the problem

This section is the reason I wrote the article. It is about the limits of this blog and of my own work.

Remember the finding from the Australian school cohort: strict dieting behaviour was the strongest single predictor. Not weight. Not sport. The restriction behaviour itself. Then any content that suggests restriction is a potential factor. Including a well meant one.

The same goes for self measurement. This cluster holds a whole article about how you can look at your own eating with data using a glucose sensor over 14 days. For many people that is a useful tool. For a person with a tendency towards control, the same sensor can become an authority that grades every meal. The same applies to tracking apps, scales and in part to exercises on noticing satiety. What creates freedom for some can pull the loop tighter for others.

And now you know why this article deliberately holds no number you could measure yourself against. Not out of caution about criticism, but because a number in the wrong head turns into a target.

For family and friends: what carries and what adds weight

Whoever stands alongside almost always feels helpless, and helplessness can produce frantic action. The following points are not an instruction and not a recipe. They summarise what has shown itself to hold, from treatment research and from conversations in my practice.

And now you know why the best question to ask someone affected is rarely “are you eating enough”, but rather “how are you doing with this right now”.

Common questions about eating disorders

Eating disorders in a wider context

Hunger, satiety, mood and metabolism belong together. From here, paths lead into other areas of this blog that look at the same axes from a different direction.