People with the same body mass index can have radically different health statuses. Credit: Hiraman/Getty
Nearly one billion people worldwide live with obesity1. Yet the scientific community is divided over what obesity means for health.
For decades, obesity was viewed mainly as a health risk, increasing the likelihood of diabetes, cardiovascular disease and premature death2. But in the past few years, it is being increasingly described as a chronic disease, a shift that reflects efforts to reduce stigma and improve access to care long denied to people living with obesity3,4.
If it were to be adopted formally by policymakers and clinicians, a uniform disease label would classify one in three adults in many high-income countries1 as having the same chronic illness. Each would be potentially entitled to lifelong treatments, from weight-loss drugs to bariatric surgery and specialist follow-up.
Why BMI is flawed — and how to redefine obesity
Yet even people with obesity who have the same body mass index (BMI) — a measurement of weight divided by height squared, used to assess whether someone has obesity — can have radically different health statuses and trajectories2,5. Some people might struggle with heart failure, breathlessness and limited mobility. Others might remain in good health long term or even throughout their lives. The experiences, prognosis and treatment needs of these two groups differ profoundly.
To address this problem, an international commission of 56 global experts (chaired by me) was convened by The Lancet Diabetes & Endocrinology in June 2022. In January 2025, after reviewing the evidence, we concluded that a single, uniform disease label is incompatible with the varied manifestations of obesity6. The commission proposed a distinction to reflect this nuance: clinical and preclinical obesity.
Clinical obesity refers to the presence of excess fat tissue that directly impairs daily activities or causes demonstrable organ dysfunction — such as heart failure, breathing disorders, metabolic dysfunction and limited mobility6. That is unequivocally disease. Preclinical obesity, by contrast, describes a condition of increased body weight and excess fat, but with preserved organ function. Risk is elevated, but there is no established disease. Crucially, both types are defined independently of obesity-associated diseases, such as type 2 diabetes, cancer and mental-health disorders, which might coexist with either state but do not define them. The distinction between preclinical and clinical obesity allows risk to be treated as risk, and disease as disease.
Despite endorsement by 76 medical organizations worldwide6, this framework has been contested by some scientific societies and other specialists7,8. Critics argue that it sets an impractical diagnostic threshold, because of the challenges of demonstrating that excess fat directly causes a person’s organ dysfunction. They worry that the existence of the preclinical category might lead to restricted access to treatment for some people. And they contend that diseases that commonly occur alongside obesity — such as type 2 diabetes — should be included in the diagnosis of clinical obesity7.
Much of this pushback stems from the view that obesity should be treated as a disease because, on a population level, it increases the risk of many health conditions. This viewpoint has long held sway in the field, but is at odds with the variable manifestations of obesity in individuals (see ‘A diverse landscape’). And it runs counter to the way diseases are typically diagnosed in the rest of medicine9 — in a person, not a population.
Source: Ref. 5
The controversy is sowing confusion among professionals and the public7,8. Division over something as fundamental as how to diagnose disease could distort clinical care, policy, insurance and public-health decisions.
Here, I outline why the commission was right to rule that obesity can be a disease — but not in every case. This view, I argue, better serves patients, policy, advocacy and scientific inquiry.
A modern mistake
For millennia, physicians have recognized that excess body weight can have health consequences. Obesity, however, has been understood as a spectrum of condition, not as a uniform disease state. Classical medical scholar Hippocrates, for instance, noted that ‘corpulence’ could signal future illness in some, constitute disease in others, and seem to protect people from the ill effects of other maladies10.
How to make America healthy: the real problems — and best fixes
Views of obesity narrowed in the twentieth century. Population-level observational data examining statistical associations between body weight and health outcomes — particularly the ‘ideal weight’ tables developed in the 1940s by the New York City-based Metropolitan Life Insurance Company — linked excess weight to increased risk of early death. Research over the following decades showed that this risk is associated not with body weight itself, but with conditions such as diabetes, high blood pressure and cancer that frequently occur alongside obesity2–4. This reinforced the view of obesity as a risk factor or risk state, rather than as a disease entity in itself.
This oversimplified idea has shaped medical practices and public opinion. Treatment access is now often tied to comorbidities. People with obesity who have limited mobility, for instance, are often denied care by doctors and insurers if they don’t also have ‘qualifying conditions’ such as diabetes or high blood pressure.
The current push to declare all obesity a disease seeks to correct this inequity. But shifting from ‘always a risk, never a disease’ to ‘always a disease’ simply exchanges one oversimplification for another — imposing a uniform definition on a condition that is not uniform.
Three factors support the need for nuance. Most importantly, obesity does not always show the clinical manifestations — signs and symptoms of organ dysfunction — through which diseases are typically defined. It also has no single unifying biological basis. In addition, a blanket label risks distorting care, science and health policy.
No bottom-up definition
Disease definitions enable physicians to predict a person’s prognosis and anticipate how an individual will respond to treatments, and to help researchers to study how biological processes go awry. The conventional framing of obesity does not fit this brief. To understand why, consider how a definition of disease is typically reached, through a ‘bottom-up’ approach.
Typically, diseases are first recognized when clinicians spot patterns of symptoms and outcomes that are seen again and again in many different patients. Take diabetes, for instance. It was defined when clinicians linked symptoms such as excessive urination, persistent hunger and fatigue to a consistent downward health spiral involving cardiac, neurological and vascular complications, blindness and death11.
Definitions can then be refined through mechanistic research and population studies. For diabetes, this helped to distinguish between distinct subtypes of the disease and the underlying biological mechanisms, and led to increasingly precise diagnostic criteria based on biological markers such as elevated blood glucose levels11.
A disease definition grounded in consistent clinical manifestations in this way makes diagnosis a meaningful prediction about what is likely to happen and what should be done about it. It also ensures that all biological, clinical and epidemiological work rests on consistent foundations.
Clinical approaches including drugs and bariatric surgery can help people with health consequences of obesity. Credit: BSIP/Universal Images Group via Getty
The benefits of this consistency were clear when it came to COVID-19. Clinicians recognized a recurring pattern of atypical acute respiratory illness in affected individuals. This clinical syndrome was then linked to infection with SARS-CoV-2, and specific mechanisms of contagion and population-prevalence data. This bottom-up definition enabled scientists and public-health experts to organize a coherent response that ultimately accelerated vaccine development and deployment.
Obesity’s health impact was not defined through a bottom-up clinical approach, but through its population-level associations with various health conditions. Yet these group-level associations do not directly translate into illness in individuals. Although all people with obesity share a high BMI, the associated clinical manifestations are highly variable2,5. One person with a BMI of 35 might be mostly healthy, another might have severe organ dysfunction. BMI can’t be used to predict an individual’s clinical status, prognosis or response to treatment, or to understand the biological underpinnings of their weight. All that can be concluded is that, at a population level, excess weight is associated with adverse outcomes.
By contrast, the subset of obesity that the Lancet Commission dubbed clinical obesity can be defined as a disease. This term describes individuals for whom excess fat tissue objectively and directly causes organ dysfunction — pulmonary, cardiorespiratory, metabolic or musculoskeletal — independently of co-occurring conditions such as diabetes6. By anchoring definition and diagnosis in organ dysfunction in an individual, the commission’s framework realigns obesity with the rest of medicine.
No consistent biological basis
Some researchers have argued that evidence of biological and genetic influences on body-weight regulation justify classifying all obesity as disease. Doing so, proponents argue, could help to counter the persistent stigma of obesity as a lifestyle choice or personal failing. These arguments have merit, but the available evidence does not fully support them.
There are no known genetic abnormalities that are shared across all people with obesity. Rare disorders involving mutations in individual genes — such as leptin deficiency — can be understood as genetic diseases, of which obesity is a common manifestation. For most other people who have obesity, a constellation of gene variants predicts a risk of excess body fat, but not whether that will lead to clinical disease12.
No ‘easy’ weight loss: don’t overlook the social cost of anti-obesity drugs
The efficacy of modern obesity medications such as GLP-1 receptor agonists is also used to support classifying obesity as disease. Yet these drugs demonstrate that body weight is biologically modifiable — not that obesity uniformly represents disease. Some therapies directly correct a pathological defect, such as insulin injections for insulin deficiency in type 1 diabetes or cancer drugs that counteract biological signalling pathways gone awry. By contrast, GLP-1 drugs act by modulating physiological systems governing appetite and satiety pathways that are broadly shared across individuals13 and are not always pathologically altered in people with obesity.
Although body-fat mass is higher in people with obesity, it should be understood as a regulated physiological state, analogous to other tightly controlled variables such as body temperature or haemoglobin concentration. An expansion of fat mass can result from many factors, including environmental pressures such as ultra-processed foods, ageing, chronic sleep deprivation, medications and disease processes such as an underactive thyroid. Like fever or high red-blood-cell count, changes can be either pathological or adaptive, depending on cause and individual context.
Fever, for example, can be a physiological response that helps fend off infections or a pathological manifestation of inflammatory disease. High red-blood-cell count can represent physiological adaptation to high altitude or pathological conditions such as polycythaemia vera. Likewise, increased body fat might reflect physiological responses to environmental pressures — it’s not always representative of an intrinsically pathological process. Treating all obesity as disease would group together biologically and clinically diverse states, potentially misleading efforts to understand the biological underpinnings of each manifestation of the condition.
