Anorexia and bulimia cases forecast to hit 21.8 million by 2050

What happened

A new global analysis estimates that 18.1 million people were living with anorexia nervosa or bulimia nervosa in 2023. It forecasts that the number will reach 21.8 million by 2050, about 20% higher. The work was published in JAMA Psychiatry, as first reported by Medical Xpress (Medical Xpress).

The coverage frames social media thinspiration and GLP-1 misuse as context for the numbers. That framing is the news outlet's, not a finding of the study, as far as we could read it.

What the study actually found

The paper, by Bai and colleagues with the GBD Eating Disorders Collaborators, uses Global Burden of Disease 2023 methods. It draws on systematic literature reviews, population surveys and vital registration systems, and covers 204 countries and territories from 1990 to 2023, with forecasts to 2050 (JAMA Psychiatry; PubMed).

There is no participant sample here. This is a statistical model, not a head count.

The regional picture varies a lot. Case counts rose most in South Asia (+170.7%), North Africa and the Middle East (+135.9%) and Andean Latin America (+113.9%). High-income regions saw a much smaller rise in counts (+18.9%) but have higher rates. The high-income rate was 424.8 per 100,000, and Australasia's was 946.4 per 100,000.

The authors conclude that eating disorders should be built into mental health surveillance, prevention and service planning worldwide.

What it does not show

• It does not count diagnoses. Medical Xpress says more than 18 million people were "diagnosed," but the study estimates people living with the disorders through modeling, not a tally of clinical diagnoses.
• It does not offer precision. The 2023 interval runs from 13.9 to 24.0 million, so the 2050 figure of 21.8 million is a point estimate with substantial uncertainty. We did not retrieve the interval for the forecast itself.
• It does not show a surge in risk per person. The rate rose 20.6% since 1990 while cases rose 71.0%. That gap suggests most of the growth reflects population growth and ageing. That is our reading of the table, not a statement from the authors, and we inferred the column labels from a jumbled layout.
• It does not link the forecast to social media or GLP-1 drugs. We found no such analysis in the portions we could read, though we could not read the full Discussion. We also could not confirm the forecast does anything beyond extending past trends, because the forecasting methods sit in a supplement we did not see.
• It does not, in what we read, tell us how the burden splits by age. The paper has an age-specific figure, but we did not see its values. So we cannot say the forecast hits teens and twenty-somethings in particular.

What it means for you

Do not read 21.8 million as a prediction about your own odds. It is a global, modeled estimate, and it says nothing about any one person.

The practical takeaways are more modest. Eating disorders are common enough that surveillance and services are being called for, and they are not only a female or only a high-income-country issue: 5.5 million of the 2023 cases were male, and counts are rising fastest outside high-income regions.

If your feed is pushing extreme thinness, treat that as a reason to curate it. If you are thinking about a weight-loss injection while also restricting food, or you have a history of disordered eating, raise it with a clinician first. A doctor is also the right person to ask about any change in eating that worries you or a friend.

Context: GLP-1 misuse and social media

Both concerns have some support, but neither has been shown to drive these projections.

On GLP-1 drugs, a JAMA Psychiatry cross-sectional study examined use and misuse among people with eating disorders in the US. Its stated rationale was that no earlier study had estimated this. A JAMA Network release says use and misuse were common in a targeted sample, with lifetime use above the 15% seen in the general adult population. It adds that some people may use the drugs in contraindicated ways to maintain restriction. We did not retrieve the sample size or exact prevalence, and a targeted sample is not representative of everyone with an eating disorder.

A scoping review found no clear clinical guidelines. Most included studies did not report eating disorders emerging from treatment or misuse driven by an eating disorder. The authors still flag people with restrictive presentations, such as anorexia, as high risk for misuse, partly because the drugs are easy to get. The review used no formal synthesis method.

On social media, a review of food-related content states that appearance-focused posts such as thinspiration and fitspiration have been shown to contribute to poor body image and disordered eating. A study of adolescents and young adults looked at thin-ideal content exposure and probable eating-disorder diagnoses. We saw only its stated purpose, not its results. We found no evidence that social media exposure caused changes in the GBD estimates.

For comparison, an earlier GBD 2021-based analysis projected bulimia burden to 2030 (PMC). We saw only its title, so we cannot compare its numbers with the new ones.