Anorexia nervosa acquired its medical name in the 1870s, but descriptions commonly associated with the illness go back to Richard Morton’s work in 1689. William Withey Gull presented his account in 1873 and published it in 1874; the French physician Ernest-Charles Lasègue published a separate description in 1873. Those dates mark changes in medical recognition, not the first time a person became ill. Historical research on Morton, Gull and Lasègue helps separate those questions.
For people living with an eating disorder, another question matters: what happens when the person seeking care doesn’t match the picture in the clinician’s head? Changes to diagnostic language, and more recent research on body weight and autistic people’s treatment experiences, make this history relevant beyond the archive.
Content note: This article discusses eating disorders and historical deaths. It does not include calorie counts, target weights or instructions for eating-disorder behaviours.
When was anorexia nervosa first described?
Morton’s Phthisiologia, published in 1689, contains accounts that later historians have associated with anorexia nervosa. “The first recorded case” is therefore a claim about surviving medical descriptions and how we interpret them. It cannot establish who first experienced the illness.
The nineteenth-century milestones are clearer. Lasègue published his account of anorexie hystérique in April 1873. Gull presented his paper to the Clinical Society of London that October; his published account followed in 1874. A historical analysis of their competing claims to priority concludes that neither can simply be awarded sole priority for describing the condition. Gull is associated with the name anorexia nervosa; recognition developed through more than one physician’s work.
The name can also mislead. In general medical language, anorexia means loss of appetite. Anorexia nervosa is a specific eating disorder, not a synonym for every illness that makes eating difficult. Its features include restriction leading to significantly low body weight, fear of weight gain and disturbances in how a person experiences or evaluates body weight and shape. Atypical anorexia, discussed below, differs on the low-weight criterion. The US National Institute of Mental Health’s overview explains the distinction between eating disorders and ordinary changes in appetite.
That distinction becomes especially important when looking backwards.
Did anorexia exist before it had a name?
Records of extreme fasting long predate the diagnosis. Whether a particular historical person had anorexia nervosa is a different, harder question. Accounts written by religious supporters, doctors or observers rarely provide the kind of assessment a clinician would need today.
Catherine of Siena and religious fasting
Catherine of Siena’s severe fasting in fourteenth-century Italy is frequently discussed in this history. Historian Rudolph Bell’s 1985 book Holy Anorexia interpreted the fasting of medieval religious women through that lens. The interpretation should not be mistaken for a confirmed diagnosis.
Religious practices gave food restriction meanings that cannot automatically be reduced to a modern fear of weight gain. Nor does the presence of a religious explanation rule out illness. The responsible answer leaves room for uncertainty: these accounts document dangerous restriction, but they do not establish that every fasting saint had the same disorder, or that fasting itself is proof of anorexia nervosa.
A person from the past deserves more than a diagnosis assigned from a few dramatic details.
Sarah Jacob: a child placed under observation
Sarah Jacob, known as the Welsh Fasting Girl, died in December 1869 at the age of twelve. Claims that she could live without food attracted visitors and medical scrutiny. Nurses were brought in to watch her; she died during that observation. Her parents were subsequently convicted of manslaughter.
The National Library of Wales’s account, based on the defence brief and witness evidence, records the failure to provide nourishment and criticism of the doctors involved. Wellcome Collection also holds Robert Fowler’s 1871 account of the case, a contemporary document available under a Public Domain Mark.
These records establish a child’s death and serious failures in her care. They do not provide a sound basis for diagnosing her with anorexia nervosa now, much less for treating her death as evidence of pride or exceptional willpower.
The question that belongs at the centre of this episode is why observation continued while a child needed care. Curiosity about a patient cannot take precedence over the patient’s safety.
Why Karen Carpenter’s death is not the beginning
Karen Carpenter died from complications of anorexia nervosa on 4 February 1983, as the Carpenters’ official biography records. By then, Gull and Lasègue’s descriptions were more than a century old.
A history that opens with Carpenter can make the illness appear much newer than it is. It can also leave readers with an unnecessarily narrow frame: a famous young woman, a public image and a death.
Carpenter was a musician, a singer and a drummer. Her life deserves that context. There is no need to turn her body into a retrospective diagnostic exhibit, or to reproduce measurements to explain why the illness was serious.
Her death belongs in the history. It does not date the illness’s arrival.
In 2013, the diagnostic wording changed
The fifth edition of the American Psychiatric Association’s diagnostic manual, DSM-5, made two revealing changes to anorexia nervosa’s criteria.
It removed the word “refusal” from the low-weight criterion because that wording implied intention. It also removed the requirement for absent menstrual periods, which excluded or failed to apply to several groups, including people who do not menstruate and people who continued menstruating while otherwise meeting the criteria. The APA’s explanation of its 2013 changes sets out those reasons.
Both changes matter to how we talk about patients. A word that sounds like a description can carry an accusation. A bodily sign included in a definition can make somebody else’s illness harder to recognise.
The same manual also recognised binge-eating disorder as a formal diagnosis and introduced avoidant/restrictive food intake disorder, or ARFID, beyond the earlier childhood feeding-disorder category. Anorexia nervosa is therefore one part of the wider history of eating disorders. Severe difficulty eating does not always belong under the same label.
Changing a definition does not create the people it now includes. It changes the categories available to describe their experience.
Research moved beyond a single explanation
The history does not support a tidy choice between an illness caused by culture and an illness caused by biology. Contemporary accounts of eating disorders recognise interacting biological, psychological and social factors. The NIMH guidance also explicitly rejects the idea that an eating disorder is a choice.
A major 2019 genetic study in Nature Genetics analysed 16,992 people with anorexia nervosa and 55,525 controls. It identified eight associated genetic regions and found genetic relationships with both psychiatric and metabolic traits.
Those are population-level associations. They do not identify a single cause for an individual’s illness, establish genetic inevitability or make social experience irrelevant. The study’s European-ancestry sample also limits how broadly its findings can be generalised.
For an individual perspective on the gap between assumptions and experience, Roseanne Murphy’s account of revealing her anorexia to friends and family describes how difficult that conversation could be. It is her experience, rather than a universal explanation of the condition.
Who still gets missed by the familiar picture?
The NIMH overview makes clear that eating disorders affect people across ages, racial and ethnic backgrounds, body weights and sexes. A person can be seriously ill without looking as an observer expects.
Atypical anorexia and the limits of appearance
Atypical anorexia nervosa describes a presentation meeting the criteria for anorexia nervosa except that the person is not at a low body weight, despite significant weight loss. The word “atypical” should not be read as “unimportant.”
An updated systematic review and meta-analysis published in August 2026 included 64 publications. It found substantial psychological distress among people with atypical anorexia, with higher eating-disorder-specific symptom severity than in the anorexia nervosa comparison groups. Physical complications also occurred, although the review found them less frequently than in low-weight anorexia nervosa.
That distinction matters. These findings do not mean identical medical risk in every group or every patient. They do mean the absence of low body weight cannot establish that someone is well.
In a 2019 first-person article for The BMJ, eating-disorder advocate Hope Virgo described seeking help during a relapse and being told she was not thin enough for support. Her account documents her own experience, not a prevalence estimate. It shows what a weight-based assumption can mean at the point of asking for care.
For England, NICE’s eating-disorder guideline says treatment decisions should not rest on BMI or illness duration alone. That is clinical guidance, not a universal statement about legal entitlement in every country.
Autistic people’s experience of treatment
Being counted in research is only part of being understood. A mixed-method systematic review first published in November 2024 examined eating-disorder treatment in autistic people and people with high autistic-trait scores.
It found poorer reported treatment experiences and indications of more intensive or longer inpatient care, alongside broadly similar improvements in eating-disorder symptoms and BMI. The authors identified a need for better autism-informed support. They also stressed the limits of evidence that relies heavily on trait scores rather than diagnosed autistic samples.
A screening score does not establish an autism diagnosis. Nor do these findings show that treatment cannot help autistic people. They ask us to look beyond an outcome measure and listen to how care is experienced. Communication and sensory needs belong in that conversation.
What this history asks of care now
A diagnosis should help a person access care. It should not become a test of whether they resemble the best-known patient, can explain their distress persuasively, or fit every assumption built into an older description.
For someone also living with chronic illness or another disability, the clinical picture needs room for both. NICE recommends collaboration between eating-disorder specialists and teams treating co-occurring physical or mental health conditions. Competing services should not leave the person doing all the coordination.
In URevolution, Lilian McCarthy’s essay on eating-disorder recovery and chronic migraine at college brings that overlap into everyday life, including the emotional and financial implications of buying clothes that fit a changing body. A history of diagnostic labels cannot contain all of that. Patient accounts need space alongside it.
The useful lesson is to keep testing our descriptions against the people seeking help, rather than requiring people to prove the descriptions right.
This article is a historical and editorial overview, not a diagnostic checklist or treatment plan. Treatment and recovery are possible. If you are concerned about eating, restriction or changes in your health, seek an assessment from a qualified healthcare professional. An eating-disorder service can help explain the support available locally; immediate danger requires urgent medical help.
