Medicine can treat obesity, but can’t agree on what obesity is

IN THE SPACE of less than a decade, medicine acquired a means of doing what millennia of diets and exhortation could not. A new class of drugs, GLP-1 receptor agonists, can strip a fifth or more from a patient’s body weight. This week Britain became the second large market, after America, to make one such treatment, Novo Nordisk’s Wegovy, available in tablet form as an alternative to injecting it.This pharmacological leap builds on the proven safety and durability of metabolic surgery, expanding a therapeutic arsenal that can now reach millions. And yet, as it deploys these extraordinary tools, medicine cannot answer a more fundamental question: what, exactly, is it treating? Astonishingly, obesity still lacks a universally accepted clinical definition.A growing view is that obesity should be defined as a chronic, progressive, relapsing disease. But not all of the one billion people living with obesity are ill, and distinguishing those who are sick because of obesity from those who are not is precisely what medicine’s definitions and diagnoses are meant to do.Medicine leans on body-mass index (BMI), a number that reflects little beyond body size and population-level risk, while saying nothing about whether the patient is actually ill. Without a clinically meaningful definition, therapies cannot be deployed rationally. Doses, durations and intensities cannot be matched to need; financial cost cannot be matched to medical benefit. Health systems already straining under demand are being asked to fund treatment for up to a third of all adults in some countries, with no reliable basis for deciding who truly requires it, and at what cost.The roots of the muddle lie in the fact that obesity has never yielded a clean clinical identity, like diabetes or cancer. That is because at the bedside it has never behaved as a single thing. Even Hippocrates viewed it as a spectrum—a warning of future illness in some, an active malady in others. Because obesity lacked a consistent pattern, for most of medical history doctors sensibly declined to characterise it as a uniform condition.That intuition was discarded in the 20th century—and, tellingly, not by doctors. Armed with actuarial tables linking weight to early death, the life-insurance industry recast all obesity as a risk state. BMI, a 19th-century measure of people’s average size never meant for clinical use, was pressed into service to quantify that risk, and became the very definition of the condition.Obesity came to be defined not by its direct toll—starving the lungs, straining the heart, wearing the joints—but by the other diseases it keeps company with. Yet in medicine a disease is defined by the damage it actually causes, not its statistical associations. This inversion of medical logic has had practical consequences. For decades, people whose excess weight was itself choking their breathing, burdening their hearts or crippling their mobility were refused treatment if they lacked the “right” comorbidity, because obesity was conceived as a risk, not a disease. No diabetes, no therapy. Active disease went unseen because the diagnostic apparatus could not see it.The current push to call obesity a disease is largely an effort to right this wrong. But the pendulum now threatens to swing to the opposite, equally implausible extreme. Population-level risks of obesity apply unevenly across individuals: many never develop adverse outcomes, even over decades. Declare obesity always a risk, as we did for much of the 20th century, and the genuinely sick are abandoned. Declare it always a disease, as a growing narrative now implies, and millions are marked ill for conditions they do not have and exposed to treatments they may not need.GLP-1 drugs are strikingly safe, yet carry side-effects such as muscle loss and, rarely, pancreatitis; metabolic surgery’s mortality can be as low as 0.1%, with serious complications in only 1-4% of patients. Even risks this small are disproportionate when the patient has little to gain.Last year, a Lancet commission of 56 experts, which I chaired, proposed a way out. It distinguished clinical obesity, a disease state where excess fat directly causes demonstrable organ dysfunction or impairs daily life, from pre-clinical obesity, a risk state in which organ function is preserved. This is how the rest of medicine works. A polyp is not cancer; prediabetes is not diabetes; high blood pressure is not heart disease. Such boundaries exist not to ration care but to make it proportionate: clinical obesity warrants disease-level treatment; pre-clinical obesity still warrants attention—lifestyle change, drugs, sometimes surgery—but at a dose and intensity calibrated to a risk, not to an established illness.The framework now carries the endorsement of 76 medical organisations, but also its share of criticism. The fear is that defining pre-clinical obesity as a risk state will hand insurers an excuse to withhold cover. The concern rests on a false premise: that medicine and insurance treat only disease and ignore risk. They do no such thing. Statins for high cholesterol, drugs for high blood pressure and pre-emptive surgery for inherited cancer risk are all routinely funded.Even drugmakers, who may be assumed to favour the broadest possible definition, stand to lose from a uniform-disease label. Developing metabolic therapies takes vast capital and precise, defensible indications. A world in which all obesity is a disease, yet many patients will never suffer the harms the drugs are designed to avert, is one in which the cost-effectiveness case collapses. Separating clinical from pre-clinical obesity lets dose, duration and price track the patient’s actual state—the only basis on which such investment earns sustainable returns, clinical and commercial alike.To those who spent decades fighting merely to have obesity taken seriously, a framework that declines to call every case a disease can feel like retreat. It is the opposite: the diagnostic discipline that aligns obesity, at last, with the rest of medicine. Francesco Rubino is a professor of metabolic surgery at King’s College London. He chaired a Lancet commission on clinical obesity.