How Do People Die From Anorexia Nervosa?

Anorexia nervosa kills through a combination of cardiovascular collapse, suicide, and progressive organ failure driven by prolonged starvation. People with anorexia face roughly three to four and a half times the risk of premature death compared to the general population, making it one of the deadliest psychiatric disorders. The mechanisms are varied and often overlap, which is part of what makes the illness so dangerous and so difficult to manage clinically.

How Elevated the Risk Actually Is

The mortality numbers around anorexia are striking even by the standards of serious mental illness. A study following patients for up to 40 years found that the risk of death from all causes was about four and a half times higher than expected, with deaths from natural causes alone nearly four times higher than in the general population.¹ A separate matched cohort study found that roughly three percent of patients with anorexia died before age 75, compared to less than one percent of matched controls.² That may sound like a small absolute number, but it reflects a young population: most people with anorexia are diagnosed in adolescence or early adulthood, so dying before 75 represents decades of life lost.

The risk is not evenly distributed. People first hospitalized for anorexia in their twenties had the highest risk of premature death, roughly eight times that of matched controls. Those hospitalized three or more times for the illness had nearly five times the risk.² And the longer the illness persists, the worse the outlook. One longitudinal study found that the mortality ratio more than doubled when illness duration exceeded 15 years, with alcohol abuse, low body mass index, and poor social adjustment independently predicting death.³

Cardiac Death and Why the Heart Fails

Cardiovascular complications are widely considered the leading cause of sudden death in anorexia nervosa. The heart is a muscle, and prolonged starvation shrinks it. Studies using echocardiography have shown that patients with anorexia have smaller heart chambers and thinner chamber walls than healthy controls, along with low blood pressure and abnormally slow heart rates.⁴ This cardiac wasting, where the left ventricle physically shrinks, has been confirmed through imaging showing significantly lower heart muscle mass in patients with anorexia compared to age-matched controls.⁵

Bradycardia, or an abnormally slow resting heart rate, is common in anorexia and is itself a marker of severe malnutrition. But the more immediately lethal cardiac problem involves the heart’s electrical system. Starvation and the electrolyte disturbances it produces, particularly low potassium, magnesium, and phosphorus, can alter the timing of the heart’s electrical signals. This shows up clinically as a prolonged QT interval, a measurement of how long the heart takes to reset between beats. When that interval stretches too far, the heart becomes vulnerable to chaotic, life-threatening rhythms.⁶

The most feared of these is a rhythm called torsades de pointes, a type of ventricular arrhythmia that can degenerate into ventricular fibrillation and cardiac arrest. A systematic review concluded that bradycardia and QT prolongation were the most common cardiac abnormalities in anorexia and that these could occasionally trigger exactly these fatal arrhythmias.⁷ That said, the picture is less clear-cut than textbooks sometimes suggest. A different systematic review found that despite the widely held belief that arrhythmias cause sudden death in anorexia, the direct evidence linking them is actually thin.⁸ The cardiac risk is real, but the specific chain of events in any given sudden death is often unclear, because many of these deaths happen outside clinical settings where cardiac monitoring would capture what went wrong.

Certain clinical markers sharpen the picture of who is most at risk. A review of sudden death in eating disorders identified long-term illness, chronic low potassium, chronically low albumin levels, and a QT interval exceeding 600 milliseconds as particularly concerning findings.⁹

Suicide

Anorexia nervosa carries one of the highest suicide rates of any psychiatric illness. In the 40-year follow-up study mentioned earlier, about 14 percent of all deaths among patients with anorexia were due to suicide, and the risk of dying by suicide was roughly 11 times higher than expected.¹ The matched cohort study similarly identified suicide as one of the leading causes of premature death associated with anorexia.²

The risk is not the same across all presentations of the illness. Suicide attempts are more common in patients who purge, whether through vomiting, laxative use, or other compensatory behaviors. Women with any eating disorder that includes purging have higher odds of attempting suicide compared to those with purely restrictive patterns.¹⁰ One study found that switching from the restrictive subtype of anorexia to the binge-purge subtype was one of the strongest predictors of suicide attempts, with a dramatically increased frequency when combined with major depression.¹¹ Other identified risk factors for death in severe anorexia include a history of previous suicide attempts and diuretic use.¹²

The relationship between starvation and suicidality is likely bidirectional. Malnutrition itself alters brain chemistry and worsens depression, hopelessness, and rigidity of thinking. Meanwhile, the despair of living with a chronic, treatment-resistant illness drives its own psychological toll. Clinicians treating anorexia increasingly recognize that suicide risk cannot be treated in isolation from nutritional status and vice versa.

What Starvation Does to the Organs

Beyond the heart, prolonged starvation systematically damages nearly every organ system. The body in severe anorexia is running on emergency reserves, and when those reserves run out, it starts consuming its own tissues.

The liver is one of the organs most visibly affected. Elevated liver enzymes are well documented in anorexia patients, reflecting ongoing liver cell damage. The suspected mechanisms include reduced blood flow to the liver and a process called autophagy, in which starving cells break down their own components for energy.¹³ In the most severe cases, the liver itself shrinks as the body depletes its stored energy sources. A study of end-stage malnutrition in anorexia found that liver volume reduction, combined with dangerous drops in blood sugar and triglycerides, represents what researchers described as a terminal stage where the body has consumed the liver as a “last resort” for energy.¹⁴

Bone marrow is another casualty. In healthy people, bone marrow is where blood cells are produced. Severe starvation can transform it into what pathologists call “gelatinous marrow,” a state where the normal fat and blood-producing cells are replaced by a gelatinous substance. This is associated with drops in all blood cell lines, a condition known as pancytopenia, which can be fatal.¹⁵ Gelatinous marrow transformation is not unique to anorexia and can appear in any condition of severe nutritional depletion, but it is an ominous sign of how far the body has deteriorated.¹⁶

Interestingly, despite often severely low white blood cell counts, patients with anorexia do not seem to get infections at higher rates than the general population. An older but frequently cited study found that even with markedly reduced white blood cells across all types, patients with anorexia had no more infections than controls.¹⁷ This remains one of the more puzzling immunological observations in the field.

Dying During Treatment

One of the cruelest aspects of anorexia is that refeeding itself can kill. When a severely starved person begins eating again, the sudden shift in metabolism can cause dangerous drops in phosphorus, a mineral critical for energy production at the cellular level. This phenomenon, called refeeding syndrome, is one of the most dangerous complications during treatment.¹⁸

When a starving body suddenly receives carbohydrates, insulin surges and drives phosphorus from the blood into cells, causing levels to plummet. In one study of hospitalized adolescents with anorexia, over a quarter required phosphorus supplementation during refeeding, and most reached their lowest phosphorus levels within the first week of hospitalization. The most malnourished patients were at highest risk. The patient with the most severe drop experienced dangerous heart rhythm disturbances.¹⁹ Severe hypophosphatemia can cause heart failure, respiratory failure, and seizures. The review of sudden death in eating disorders specifically flagged the cardiac effects of hypophosphatemia during refeeding as a recognized cause of death.⁹

This creates a clinical dilemma. Patients who are most malnourished and most urgently need nutrition are also the ones most vulnerable to dying from receiving it. Modern treatment protocols address this by starting nutrition slowly and monitoring electrolytes closely, but refeeding syndrome remains a genuine risk, particularly in settings without specialized eating disorder expertise.

Gastrointestinal Emergencies

Severe malnutrition can produce life-threatening gastrointestinal complications that are rare in the general population but surface with disturbing regularity in anorexia. One of these is superior mesenteric artery syndrome, where the loss of the fat pad between two major blood vessels causes compression of the upper small intestine. This obstruction can lead to massive stomach dilation, which at its worst can progress to stomach rupture or tissue death from restricted blood supply.²⁰ Case reports have documented situations where the stomach expanded so dramatically that imaging showed the stomach wall developing air within it, a condition called gastric pneumatosis that indicates tissue on the verge of rupturing.²¹

These events are individually rare, but they illustrate a broader principle: in severely malnourished bodies, systems that normally run on autopilot begin to malfunction in unpredictable ways. The loss of fat, muscle, and connective tissue changes the physical relationships between organs. Structures shift, compressions occur, and normal digestive functions can become medical emergencies.

How the Restricting and Binge-Purge Subtypes Kill Differently

Anorexia nervosa has two recognized subtypes, and they tend to cause death through somewhat different pathways. The restricting subtype, where the person simply eats very little, produces more of the complications associated with pure starvation: bone marrow failure, liver damage, and dangerously low blood sugar. A study comparing the two subtypes during hospitalization found that patients with the restricting form had higher rates of these starvation-mediated complications.²²

The binge-purge subtype, where restricting is combined with episodes of binging and compensatory purging, carries its own distinctive dangers. Repeated vomiting and laxative or diuretic abuse create severe electrolyte disturbances, particularly low potassium, that directly threaten heart rhythm stability. These patients also face physical injuries from the purging itself, including tears of the esophagus that can lead to life-threatening infections in the chest cavity. The binge-purge subtype also carries higher rates of substance use disorders. A meta-analysis found that about 18 percent of people with the binge-purge subtype had a substance use disorder, compared to about 7 percent of those with the restricting subtype.²³ Substance abuse independently predicts death in anorexia, both through direct toxicity and by compounding the medical fragility that already exists.³

What Predicts Who Will Die

Research has identified several factors that, taken together, paint a picture of the patients at greatest risk. Older age at assessment, longer illness duration, and lower body weight are consistently associated with worse outcomes.¹² A history of suicide attempts is a strong predictor, as is the intensity of eating disorder symptoms and the use of diuretics, which signals both more severe compensatory behaviors and higher electrolyte risk. Poor social adjustment, an index of how well someone is functioning in daily life, independently predicts death over long follow-up periods.³

One pattern that emerges from the data is that anorexia’s lethality is strongly tied to chronicity. People who recover early tend to do well. Those whose illness becomes entrenched over years and decades face compounding medical and psychological damage that progressively narrows the window for recovery. The mortality ratio more than doubles when illness duration stretches beyond 15 years, and each hospitalization adds to the risk profile. This is one reason clinicians emphasize early treatment: not just because recovery is more likely early on, but because the cumulative physical toll of years of malnutrition creates problems that persist even if eating normalizes.

The Debate Over Terminal Anorexia

A controversial and still-evolving conversation in the field concerns what happens when anorexia becomes, by any honest clinical assessment, treatment-resistant. Some patients endure decades of repeated treatment, hospitalizations, and forced refeeding without sustained recovery. Their quality of life deteriorates under the weight of both the illness and the compulsory treatment meant to keep them alive.

In 2022, researchers proposed the concept of “terminal anorexia nervosa,” describing patients who meet certain criteria: a diagnosis of anorexia nervosa, age over 30, previous participation in high-quality treatment, and a clear and consistent determination by the patient that additional treatment would be futile, made with full understanding that their choices will lead to death. The authors argued that these patients deserve the same rights and palliative care options as patients with other terminal illnesses, including, in jurisdictions where it is legal, medical aid in dying.²⁴

This proposal has generated significant pushback. Critics argue that anorexia, as a psychiatric illness that distorts the patient’s own perception of their condition, makes it difficult to assess whether a person truly possesses the decision-making capacity required for end-of-life choices. Others worry that the label “terminal” could lead to premature abandonment of treatment. A separate model, described in the AMA Journal of Ethics, proposed a palliative care approach for severe and enduring anorexia that stops short of hospice or assisted dying. It instead aims to reduce suffering, improve day-to-day functioning, and honor the patient’s goals without framing the illness as inevitably fatal.²⁵

The tension here is genuine and unresolved. On one side is the recognition that some patients experience years or decades of compulsory treatment as a form of suffering in itself. On the other is the evidence that recovery is possible even after very long illness duration, and that the cognitive distortions characteristic of anorexia may undermine the informed consent process. How medicine navigates this question will shape the treatment of the illness’s most severely affected patients for years to come.

Why Cause of Death Can Be Hard to Determine

When someone with anorexia dies, determining the exact cause is often less straightforward than it might seem. Many of the lethal mechanisms overlap. A person with chronic low potassium, a shrunken heart, and a prolonged QT interval who dies in their sleep may have experienced a fatal arrhythmia, but without continuous heart monitoring at the time of death, this is often an inference rather than a confirmed diagnosis. The same ambiguity applies to deaths that may be intentional. Case reports describe deaths initially attributed to medical complications that on closer investigation showed signs of deliberate self-harm, such as a patient who fatally bled after a cardiac procedure, where the clinical team later suspected manipulation of the surgical wound with suicidal intent.²⁶

This diagnostic murkiness matters for the statistics too. Deaths coded as “cardiac” on a death certificate may mask underlying starvation; deaths coded as “natural” may involve undetected suicidal behavior. The reported proportion of deaths due to suicide, around 14 percent in the long-term data, is likely a floor rather than a ceiling. Some researchers suspect that the true contribution of suicide to anorexia mortality is underestimated, in part because of the blurred boundary between the slow self-destruction of the illness and acute suicidal acts. For families, clinicians, and epidemiologists alike, this ambiguity complicates both grief and prevention efforts.

1
PubMed. Overall and cause-specific mortality in anorexia nervosa; impact of psychiatric comorbidity and sex in a 40-year follow-up study
2
PubMed Central. Premature Mortality in Patients With Anorexia Nervosa: A Matched Cohort Study of Cause of Death
3
PubMed Central. A longitudinal investigation of mortality in anorexia nervosa and bulimia nervosa
4
PubMed. Effects of self-induced starvation on cardiac size and function in anorexia nervosa
5
PubMed. Reduced hemodynamic load and cardiac hypotrophy in patients with anorexia nervosa
6
PubMed. Cardiovascular complications of eating disorders
7
PubMed. Anorexia nervosa and heart disease: a systematic review
8
PubMed. Cardiovascular complications of anorexia nervosa: A systematic review
9
PubMed Central. Sudden death in eating disorders
10
PubMed Central. Suicide attempts in women with eating disorders
11
European Psychiatry. Switching to the bingeing/purging subtype of anorexia nervosa is frequently associated with suicidal attempts
12
PubMed. Factors predictive of ten-year mortality in severe anorexia nervosa patients
13
PubMed Central. Expanding the differential diagnosis for transaminitis in patients with anorexia nervosa
14
PubMed Central. Severe hypoglycemia with reduced liver volume as an indicator of end-stage malnutrition in patients with anorexia nervosa
15
PubMed Central. Decreased Volume of Bone Marrow Adipocytes With Sparse Gelatinous Marrow Transformation in a Patient With Pancytopenia With Anorexia Nervosa
16
PubMed Central. Starvation marrow – gelatinous transformation of bone marrow
17
PubMed. Leukopenia in anorexia nervosa. Lack of increased risk of infection
18
PubMed. Predictors of hypophosphatemia during refeeding of patients with severe anorexia nervosa
19
Journal of Adolescent Health. Hypophosphatemia during nutritional rehabilitation in anorexia nervosa: implications for refeeding and monitoring
20
PubMed. Severe gastric dilatation due to superior mesenteric artery syndrome in anorexia nervosa
21
PubMed Central. Gastric Pneumatosis and Its Gastrofibroscopic Findings in Life-Threatening Superior Mesenteric Artery Syndrome Complicated by Anorexia Nervosa in a Child
22
PubMed. A comparison of the metabolic complications and hospital course of severe anorexia nervosa by binge-purge and restricting subtypes
23
PubMed Central. The prevalence of substance use disorders and substance use in anorexia nervosa: a systematic review and meta-analysis
24
PubMed Central. Terminal anorexia nervosa: three cases and proposed clinical characteristics
25
PubMed. A Life-Affirming Palliative Care Model for Severe and Enduring Anorexia Nervosa
26
Forensic Science, Medicine and Pathology. Fatal bleeding after transfemoral coronary angiography in anorexia nervosa