What Is Anorexia Nervosa?

Anorexia nervosa is a mental illness with three parts in the American Psychiatric Association’s diagnostic manual, DSM-5-TR. The person eats too little for their body’s needs, so that their weight is significantly low for their age, sex, development and physical health. They have an intense fear of gaining weight, or keep acting in ways that interfere with weight gain.

The diagnosis also needs any one of three things: a distorted experience of one’s own weight or shape, weight or shape having too much influence on how the person judges themselves, or persistently not recognising how serious the low weight is.

So someone can know exactly how ill they are and still meet the diagnosis. Some people with anorexia do not believe the danger applies to them; others understand it well, and facts alone do not help someone who already knows them.

In one study, patients wrote two letters to their anorexia, one addressing it as a friend and one as an enemy. Among its benefits, they often named feeling looked after or protected, gaining a sense of control, and feeling special. The researchers suggest that these may help keep the illness going.

Someone can meet every other criterion for anorexia and still be at a normal weight or above, if they have lost a lot of weight to get there. The diagnostic manual calls this atypical anorexia nervosa.

One study followed 496 adolescent girls in a US city for eight years. By age 20, 2.8% had atypical anorexia and 0.8% the classic kind, about three and a half times as many.

That is one group of girls in one city, not a universal ratio. A 2021 review of 75 studies found atypical anorexia typically as common as the classic kind, or more common, in community studies. Among patients referred or admitted for specialist care, it was usually the rarer of the two.

Men get it. In Norway’s national patient register, about one man was diagnosed for every 13 or 14 women in 2010 to 2016, as a 2021 review reports. The reviewers say the figures for men are underestimates, because the illness is less often detected in men.

Children get it: anorexia is seen from about age 7. Eating disorders occur in midlife too, either as illnesses that began earlier or as new ones. A 2017 review found eating disorders in all races and ethnicities, though anorexia was somewhat less common among Black than White Americans. It found no consistent link between eating disorders and income or social class.

Two other illnesses get confused with it. Avoidant/restrictive food intake disorder (ARFID) is avoiding or restricting food so much that the body’s energy or nutritional needs go unmet, not because of worry about weight or shape. It can come from a lack of interest in food, dislike of how foods look, taste, smell or feel, or fear of what eating might bring. Bulimia is repeated binge eating followed by behaviour meant to prevent weight gain, usually at a normal weight or above.

Anorexia has one of the highest death rates of any mental illness.

It is sometimes called the deadliest, but some drug addictions have higher death rates. In a 2014 review of reviews, people with opioid use disorders died at 14.7 times the expected rate, against 5.9 times for anorexia. Apart from substance use, that review’s summary table had only one higher figure: women admitted to psychiatric care after giving birth, over the first year.

Someone with anorexia is roughly five to six times more likely to die than people of the same age and sex in the general population. Two reviews that pooled many studies put the figure at 5.86 in 2011 and 5.06 in 2025. The 2025 review covered 30 studies of 33,176 patients.

In the 2025 review, suicide caused 21% of deaths, the most of any single cause. Heart problems caused 19%, a figure drawn from fewer studies.

A Swedish study followed 6,009 women who had been treated in hospital for anorexia. Their risk of dying was still significantly raised 20 years or more after their first admission.

Starvation damages almost every major organ system, a 2015 review of anorexia’s medical complications found. The heart slows and can develop dangerous rhythms. Blood pressure drops. Bones thin, and break more easily. Periods often stop. Growth stalls in children. The gut slows down. Blood counts can fall. Salts in the blood can go wrong, and very low calcium can cause seizures, the Royal College of Psychiatrists notes. The brain physically shrinks, though that largely reverses when weight is regained.

Feeding someone back up after a long period of eating too little is itself risky if food is increased suddenly. That is why it is done with professional support, sometimes in hospital.

Recovery rates quoted for anorexia vary widely, and one reason is how long a study follows people.

Roughly half of the women who had not recovered by nine years had recovered at 22 years. In the Boston study, having major depression as well at the start strongly predicted having the restricting type of anorexia 22 years later.

A 2002 review of 119 study series also found better outcomes with longer follow-up. In that review, about 20% of surviving patients remained chronically ill on average, with large variations across studies.

Eating disorders are not choices but serious, biologically influenced illnesses, in the words of the Academy for Eating Disorders.

Anorexia runs in families. Twin studies estimate that genes account for 48% to 74% of the variation in risk of anorexia, a 2015 review found. That does not mean a single gene causes it: genes alone do not predict who will develop an eating disorder, according to the 2017 review.

A 2019 study of the genes of 16,992 people with anorexia and 55,525 without found eight places in the genome linked to the illness. Anorexia’s genetic profile overlapped with other psychiatric disorders, including obsessive-compulsive disorder, and with how the body handles sugar and fat. The authors say this supports seeing anorexia as a “metabo-psychiatric” disorder, and considering both metabolic and psychological drivers of it. Shared genetics does not by itself show that metabolism causes anorexia.

Culture plays a part, but not on its own. Television arrived in 1995 in Nadroga, a province of Fiji. Researchers surveyed one group of schoolgirls there within weeks of its arrival and another in 1998. The share scoring high on a test of disordered eating attitudes rose from 12.7% to 29.2%. Pressure to be thin is a recognised risk factor, but only a small share of those exposed to it develop an eating disorder, the 2017 review notes.

Researchers at the University of Oxford interviewed young women with anorexia at length. In one study, ten of them took a standard test of the capacity to make treatment decisions. As a group they showed excellent understanding and reasoning, and all but one clearly passed.

But the interviews found difficulties with thinking that the test missed, and changes in what they valued. Losing weight came to matter more than health, family, friendships or study, and some said they did not mind the risk of dying. For some, the illness felt like part of who they were.

The researchers argue that anorexia may severely compromise a person’s autonomy, particularly over eating and weight. So the illness can damage the ability to choose without removing it.

For children and young people, the National Institute for Health and Care Excellence (NICE), which writes the national guideline for England, says to consider family therapy focused on anorexia. Parents take a central role in helping their child manage eating, a role meant to be temporary, and then hand back independence. The guideline says the therapy should not blame the person or their family. If family therapy is unacceptable, unsuitable or ineffective, NICE says to consider individual therapy.

The difference was statistically significant only after treatment had finished, at 6 and 12 months. Even with family-based treatment, about half were not in full remission a year after it ended.

The 2019 Cochrane review of family therapy for anorexia covered 25 trials. It found only limited, low-quality evidence that family therapy may work better than usual treatment in the short term. It found too little evidence to say whether family therapy beats other psychological therapies.

In an Australian trial of 120 adults, three therapies brought similar, clinically significant improvement, but on average 28.3% of patients were in remission a year after treatment. In a German trial of 242 adults, two therapies and an improved form of usual care led to similar weight gain.

NICE lists three therapies for adults as options and says to explain them so that the person can choose which they prefer.

A 2019 trial’s authors wrote that efforts to develop effective drug treatments for anorexia have been disappointing. In that trial of 152 adults, the antipsychotic olanzapine led to a modest weight gain compared with a placebo, but brought no significant benefit for psychological symptoms. In a 2006 trial of 93 patients whose weight had been restored, the antidepressant fluoxetine did no better than a placebo at preventing relapse.

NICE says medication should not be offered as the only treatment for anorexia.

Relapse is common. A 2017 review of 27 studies found reported relapse rates from 9% to 52%, tending to rise with longer follow-up. It also found agreement that the risk is especially high in the first year after treatment.

Three things, if you only read this far.

Know the emergency signs. The Royal College of Psychiatrists’ guidance for relatives lists fainting when standing up, being unable to climb stairs, talking about ending their life, suddenly refusing food or water, and another illness such as diarrhoea or vomiting. If you see any of them, it says, encourage the person to go to the emergency department or contact their eating disorder team immediately. The College’s guidance for doctors warns that someone can be at high risk while appearing well with normal blood tests.

Do not comment on how they look, even kindly. Beat, the UK eating disorder charity, says comments about someone looking “healthier” or “better” are often taken to mean they have put on weight.

Get an assessment early, and if it is someone under 18, ask about family therapy. The guideline for England says people with eating disorders should be assessed and treated at the earliest opportunity.

In 2022 two clinicians and one of their patients proposed clinical characteristics for what they called terminal anorexia nervosa. They described adults with anorexia, older (for example, over 30), who had had high-quality care and who, with the capacity to decide, consistently judged that further treatment would be futile. They argued that such patients should have the care given to others with terminal illness, including medical aid in dying where it is legal.

The paper drew public dissent in journals and the media. Four clinicians wrote that the term is dangerous, that evidence suggests most people with anorexia eventually recover, sometimes after decades, and that the label risked unjustified deaths. Writers with lived experience of eating disorders argued that it did not adequately address how often good treatment is out of reach.

This one is unresolved. The Oxford patients described above scored well on a standard capacity test, and the researchers argue that such tests can miss the ways anorexia compromises the competence to refuse treatment.

Anorexia is a serious illness, not a choice, as the Academy for Eating Disorders says of all eating disorders. Anorexia has one of the highest death rates of any mental illness. People with it are roughly five to six times more likely to die than people of the same age and sex in the general population. Suicide is the most common single cause of those deaths.

You do not have to be underweight to have it. Atypical anorexia is typically as common as the classic kind in community studies, and as psychologically severe. Many people recover, though often only after many years. For under-18s, the guideline for England puts family therapy first; for adults no therapy has clearly beaten another, and no drug treats anorexia itself.

Download the conclusion

The verdict above as one image, with the two figures from the Boston study of 121 surviving women with anorexia. Of them, 31% had recovered by the 9-year follow-up and 63% at the 22-year follow-up. It is 1,200 by 630 pixels, and anyone may download it and post it.

Download the card (PNG, 1,200 × 630)