Where our measures of health came from

Most of the tools we use every day arrived in our training already assembled. We learn what they measure and what the thresholds are. We rarely learn where they came from, or what question their inventor was actually trying to answer.

Body mass index is worth knowing the history of, because the answer is genuinely surprising.

A mathematician, an astronomer, and a population

The index originates with Adolphe Quetelet, a Belgian working in the 1830s. He was a mathematician, an astronomer and a statistician — not a physician, and he was not studying illness.

What he was doing was applying probability theory to human characteristics in pursuit of what he called l’homme moyen: the average man. From cross-sectional studies of population growth, he derived the observation that weight increases as the square of height.

That is the whole origin of the formula. It was a description of how a population is distributed, produced by someone interested in statistical averages rather than individual health.

The name came 140 years later

The term “body mass index” was coined in 1972, by the American physiologist Ancel Keys.

This part is often told carelessly, so it’s worth being precise: Keys was a health scientist. But what he was doing was evaluating which of several weight-for-height indices was most useful for comparing populations in epidemiological research. He was not proposing a diagnostic test for individuals.

So the honest version isn’t that no health professional was ever involved. It’s that at neither point — not with Quetelet, not with Keys — was this tool designed to tell you something about the health of the person in front of you.

Where the data came from

Both the original index and its twentieth-century revival drew on populations that were overwhelmingly white, European, and male.

That matters for how the thresholds travel. A recent review reports that BMI tends to overestimate fatness and health risk in Black people and to underestimate it in Asian populations.

A debate article in a behavioural nutrition journal makes the same point about representativeness — and it is worth noting that it is a debate article, with a genuine case on both sides. BMI is cheap, feasible and useful for comparing populations, which is exactly the job it was built for.

Why this matters for eating disorders specifically

Because a measure built to describe populations often sits inside decisions about individuals — including who gets referred, and when.

Earlier in this series we looked at a study of 35,404 adults in England and Northern Ireland with an eating disorder diagnosis recorded by their GP. Being underweight raised the odds of referral more than fourfold for anorexia nervosa — and lowered them for bulimia nervosa, for OSFED, and for generic eating disorder diagnoses.

UK GPs interviewed about managing these patients said the same thing directly: BMI-based referral criteria were, in their experience, excluding people who would have benefited from early intervention.

And when we asked professionals working here what made services inadequate, one answered without being prompted about measurement at all: “Too many services are dependent upon measures such as BMI.”

None of that is an argument against measuring anything. It is an argument for knowing what a threshold was built from before it shapes a referral.

Not an argument for abandoning anything

This isn’t a case for throwing out BMI, and the research isn’t unanimous either. It’s an argument for knowing what a tool was built for, so you know where it stops being informative.

A number designed to describe a population will tell you something useful about a population. Asked to describe the person sitting in front of you, it was never designed to answer.

That is the thread running through all four of these pieces. Most of what makes eating disorders hard to spot isn’t hidden. It sits just outside what we were trained to look for — in the presentations we don’t associate with food, in the behaviours we congratulate, and in the thresholds we inherited without asking where they came from.

You don’t have to be a specialist. You do need training — and we can help with that.

The training

Eating Disorders: What Every Professional Needs to Know is a one-day training for clinicians, therapists, dietitians and allied health professionals who don’t work in eating disorders and don’t intend to — but who want to recognise one when it’s in front of them, and know how to support a client who presents with signs and symptoms.

Saturday 28 November 2026, 10:00–16:00, at the Carmelite Centre, Dublin 2. Six hours, six CPD points (PSI). €169, or €139 for students. It runs on a Saturday, so it doesn’t cost you clinic time.

See the full programme → Eating Disorders: What Every Professional Needs to Know 

Frequently Asked Questions

Who invented BMI?

The underlying index originates with Adolphe Quetelet, a Belgian mathematician, astronomer and statistician working in the 1830s. He was studying population averages, not individual health, and derived the observation that weight increases as the square of height.

When did it become known as body mass index?

The term “body mass index” was coined in 1972 by the American physiologist Ancel Keys, who was evaluating which weight-for-height index was most useful for comparing populations in epidemiological research rather than proposing a diagnostic test for individuals.

Was BMI designed to assess individual health?

No. At neither stage of its development was it designed for individual assessment. Quetelet was describing population distributions and Keys was assessing indices for epidemiological comparison.

Is BMI equally accurate across ethnic groups?

No. A critical review documents that BMI tends to overestimate fatness and health risk in Black people and to underestimate it in Asian populations, reflecting the overwhelmingly white, European and male populations from which it was derived.

How does BMI affect eating disorder care?

Referral pathways built around BMI thresholds can exclude patients. In a study of 35,404 adults with an eating disorder diagnosis recorded by their GP, being underweight raised the odds of referral for anorexia nervosa but lowered them for bulimia nervosa and other specified feeding and eating disorders. UK GPs have described BMI-based referral criteria as excluding patients who would have benefited from early intervention.

Sources

• Eknoyan (2008), Nephrology Dialysis Transplantation 23(1):47–51. DOI: 10.1093/ndt/gfm517

• Bailey, Bessey, Lamarche & Griffin (2025), Body Image 54:101924. DOI: 10.1016/j.bodyim.2025.101924

• Byker Shanks, Bruening & Yaroch (2025), IJBNPA. DOI: 10.1186/s12966-025-01719-6

• Gao et al. (2026), The Lancet Primary Care 2(7):100179. DOI: 10.1016/j.lanprc.2026.100179

• Ashby & Ogden (2024), BJGP Open 8(3). DOI: 10.3399/BJGPO.2024.0014