is anorexia the most deadly psychiatric disorder?
“it's the most deadly of the psychiatric disorders.”
a host, a neuroscientist said on Huberman Lab, 2026-09-14, at 55:13. hear it in context →
first, the part that holds
anorexia nervosa does carry one of the highest mortality risks in psychiatry. people with it die at roughly six times the expected rate, and outside the substance use disorders, no condition in the main mortality meta-review ranks higher.
what we found when we went and looked
we read the 2014 meta-review of mortality across 20 mental disorders, which pooled 20 systematic reviews covering over 1.7 million patients and over a quarter of a million deaths. it gives anorexia nervosa an all-cause standardized mortality ratio of 5.9 (95% ci 4.2-8.3). opioid use disorder comes in at 14.7 (12.8-16.5), more than double that, and amphetamine use at 6.2. the authors' own summary is that "those with the highest all-cause mortality ratios were substance use disorders and anorexia nervosa", which puts anorexia near the top but not at it. so "the most deadly of the psychiatric disorders" holds only if addiction is left out of the comparison, and the clinical manual used in the us lists substance use disorders as psychiatric disorders.
assertion by assertion
| assertion | verdict | what the record says |
|---|---|---|
| anorexia nervosa has the highest mortality of any psychiatric disorder | overstated | in the 2014 meta-review, anorexia nervosa's all-cause smr is 5.9 (4.2-8.3). opioid use is higher at 14.7 (12.8-16.5), with confidence intervals that do not overlap. amphetamine use is 6.2 (4.6-8.3), which overlaps with anorexia. the authors name substance use disorders and anorexia nervosa together as the highest-risk group. the comparison uses ratios against the general population, not absolute death counts. [1] |
| people with anorexia nervosa die at a much higher rate than the general population | confirmed | a 2011 meta-analysis of 36 studies (17,272 patients, 755 deaths) found a weighted crude mortality of 5.1 per 1,000 person-years and an smr of 5.86 for anorexia nervosa. the figures were 1.93 for bulimia nervosa and 1.92 for eating disorder not otherwise specified. a 2020 review cites newer cohorts with smrs or hazard ratios of 5.2 to 6.5 in specialist-care samples, but only 2.3 when primary care is included. [2] |
| leaving out substance use disorders, anorexia nervosa has the highest all-cause mortality ratio | confirmed | among the non-substance disorders in the 2014 meta-review, anorexia nervosa (5.9) is above personality disorders (4.2), schizophrenia (2.5), bipolar disorder (2.0-2.3) and depression (1.6). this is the version of the claim the record supports. it rests on 20 reviews, only one of which was rated high quality. [1] |
| a large share of anorexia deaths are suicides | confirmed | in the 2011 meta-analysis, one in five people with anorexia nervosa who died had died by suicide (1.39 per 1,000 person-years). the 2014 meta-review gives a suicide smr of 31.0 (21.0-44.0) in a specialist-care sample. that is second to borderline personality disorder at 45.1. [1] |
both directions
for the speaker: every source we read puts anorexia nervosa among the deadliest psychiatric conditions. the 2011 meta-analysis found an smr of 5.86, about three times the ratio for bulimia nervosa or other eating disorders. the 2014 meta-review names it alongside substance use disorders as the highest-risk group. its smr ranks above schizophrenia, bipolar disorder, depression and personality disorders. its suicide risk (smr 31) is second only to borderline personality disorder. a 2020 review places it well above the pooled relative risk of 2.2 for mental disorders as a whole. if the speaker meant "among non-addiction psychiatric disorders", the record backs them.
against: the claim says "the most deadly", with no exceptions. on the measure usually used to make this claim, the standardized mortality ratio, opioid use disorder (14.7) is more than twice anorexia nervosa (5.9), and the two confidence intervals do not overlap. amphetamine use (6.2) is level with anorexia. both are psychiatric diagnoses. the 2014 authors did not rank anorexia first. they grouped it with substance use disorders. the ratio also shrinks in broader samples: a 2020 review reports a hazard ratio of 2.3 once primary-care patients are included, against 5-6.5 in specialist and inpatient samples. by absolute number of deaths, far more common conditions such as depression and substance use account for far more deaths than anorexia.
what this cannot say
standardized mortality ratios compare each group with the general population. they depend on the age and sex mix of each sample, on how the sample was drawn (specialist samples are sicker), and on which studies made it into each review. so ratios from different reviews are not strictly comparable. in the 2014 meta-review, 7 of the 20 underlying reviews were rated low quality and only one high. we could not open the full text of a large 2019 danish whole-population registry study (the publisher blocked access), so it is not cited here. that study is the most direct head-to-head test, and our ranking relies on meta-analytic data instead. "most deadly" could also mean absolute deaths, or crude rather than relative risk. the speaker named no source and no metric, so we checked the reading the claim usually carries.
sources, each one fetched and read
- Chesney E, Goodwin GM, Fazel S. Risks of all-cause and suicide mortality in mental disorders: a meta-review. World Psychiatry, 2014;13(2):153-160. PMC4102288 What we read there: "Those with the highest all-cause mortality ratios were substance use disorders and anorexia nervosa." Opioid use SMR 14.7 (12.8-16.5); amphetamine use SMR 6.2 (4.6-8.3); anorexia nervosa SMR 5.9 (4.2-8.3); suicide SMR 45.1 for borderline personality disorder and 31.0 (21.0-44.0) for anorexia nervosa. Only one of 20 included reviews was rated high quality. Fetched 2026-10-22 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC4102288/
- van Hoeken D, Hoek HW. Review of the burden of eating disorders: mortality, disability, costs, quality of life, and family burden. Current Opinion in Psychiatry, 2020. PMC7575017 What we read there: Reports anorexia nervosa SMR 5.2 (3.8-7.0) in secondary mental healthcare users and HR 6.5 (3.5-12.3) in tertiary care, but HR 2.3 (1.9-2.8) across primary and secondary care; compares these with a relative risk of 2.2 (2.1-2.3) for mental disorders overall. Fetched 2026-10-22 (HTTP 200). https://pmc.ncbi.nlm.nih.gov/articles/PMC7575017/
speakers are described by role, not named. one sentence is quoted under fair use, with attribution and a link to the original at the timestamp; we do not re-host audio or video. nothing on this page is a reason to start, stop or change a medication. if you have the file behind anything marked unverifiable, the research board is where people can answer. adam