IPM Take
BMI has enormous policy power for a measurement that tells clinicians relatively little about what is happening metabolically inside the body.
A new meta-analysis makes that limitation difficult to ignore.
Across 44 longitudinal studies, cardiovascular risk rose as both BMI and metabolic dysfunction worsened, with people who had obesity and an unhealthy metabolic profile facing the greatest risk. But the analysis also found something more provocative: people of normal weight who were metabolically unhealthy often had greater cardiovascular risk than people with obesity who were metabolically healthy.
That does not mean obesity is harmless when blood pressure, glucose and lipids look normal. Higher BMI still carried independent cardiovascular risk, and “metabolically healthy obesity” is neither guaranteed to remain metabolically healthy nor equivalent to metabolically healthy normal weight.
The point is different.
Thin does not automatically mean metabolically healthy, and obesity does not tell clinicians how much cardiovascular risk a person carries by itself.
That distinction matters as cardiology increasingly moves toward cardiovascular-kidney-metabolic medicine. Risk assessment built primarily around body size can miss a normal-weight patient with hypertension, dyslipidaemia or impaired glucose regulation while treating two patients with the same BMI as though their cardiovascular trajectories were interchangeable.
BMI remains useful at population scale.
It becomes much less useful when allowed to end the clinical conversation.
Executive Summary
A systematic review and meta-analysis published in the International Journal of Obesity examined how combinations of BMI and metabolic health relate to subsequent cardiovascular disease.
Researchers included 51 publications representing 44 longitudinal cohort studies and compared six broad phenotypes ranging from metabolically healthy normal weight to metabolically unhealthy obesity. Outcomes included overall cardiovascular disease, coronary heart disease, myocardial infarction, stroke and heart failure.
The highest risks consistently appeared when obesity and metabolic dysfunction occurred together.
Compared with metabolically healthy people of normal weight, metabolically unhealthy obesity was associated with approximately 2.1 times the risk of overall cardiovascular disease, 1.97 times the risk of coronary heart disease, 2.20 times the risk of myocardial infarction, 1.55 times the risk of stroke and 2.05 times the risk of heart failure.
However, risk did not follow BMI categories neatly.
Across several outcomes, metabolically unhealthy people of normal weight had greater risk than people classified as having metabolically healthy obesity. Stroke showed the clearest pattern, with metabolic health appearing to play the stronger role in determining risk.
The analysis nevertheless found a positive dose-response relationship between higher BMI and cardiovascular risk for cardiovascular disease, coronary heart disease, myocardial infarction and heart failure.
The implication is therefore not that metabolic health replaces obesity as a risk factor. Both matter, and their combination appears particularly harmful.
Instead, the findings support moving beyond BMI-only approaches toward assessment that considers the metabolic phenotype surrounding body size.
Why it matters
- HTA bodies: Cardiometabolic technologies and obesity interventions are increasingly evaluated in populations defined by both body size and metabolic comorbidity. The findings reinforce the importance of subgroup evidence based on blood pressure, glycaemia, lipids and cardiovascular risk rather than assuming that BMI alone identifies a clinically homogeneous population.
- Payers: Eligibility criteria built mainly around BMI may fail to distinguish patients with substantially different cardiovascular trajectories. As obesity medicines and prevention programmes create increasing budget pressure, payers may face stronger arguments for risk-based approaches that incorporate metabolic disease burden alongside body size, while avoiding the opposite mistake of withholding appropriate obesity care from patients who currently appear metabolically healthy.
- Industry / innovation partners: Obesity and cardiovascular drug development is increasingly moving toward phenotyping rather than simple weight classification. Biomarkers, longitudinal metabolic data and digital risk tools could help identify who is most likely to experience cardiovascular complications, but any move toward more sophisticated segmentation will need to demonstrate that it improves treatment selection rather than simply generating new labels.
BMI has survived for decades because it is simple. A person’s height and weight produce a number, the number produces a category, and the category offers an immediate way to describe obesity-related risk across large populations.
The problem begins when that simplicity is mistaken for precision.
Two people can have exactly the same BMI while having very different blood pressure, triglycerides, glucose regulation, fat distribution and cardiovascular histories. Another two can have very different body sizes while the apparently thinner person carries the more adverse metabolic profile.
A new meta-analysis published in the International Journal of Obesity brings those differences into sharper focus by asking what happens when body size and metabolic health are considered together rather than separately.
The researchers pooled longitudinal evidence across 44 studies, grouping participants according to both BMI and whether they were metabolically healthy or unhealthy. What emerged was not an argument that weight does not matter, but a much more complicated risk hierarchy than BMI alone can provide.
People with obesity and metabolic dysfunction consistently sat at the highest-risk end of that hierarchy. Compared with metabolically healthy people of normal weight, their risk was roughly doubled for overall cardiovascular disease, coronary heart disease, myocardial infarction and heart failure, while stroke risk was also substantially higher.
That result is unsurprising. Obesity occurring alongside hypertension, abnormal glucose metabolism, dyslipidaemia and other metabolic abnormalities creates several overlapping pathways toward cardiovascular disease.
What is more revealing is what happened elsewhere on the BMI spectrum.
Normal weight did not guarantee low cardiovascular risk.
Across several outcomes, people whose BMI fell within the normal range but whose metabolic profile was unhealthy had risks that exceeded those observed among people with obesity who were classified as metabolically healthy.
That finding exposes one of the weaknesses of using body size as a shorthand for metabolic status.
A person can look relatively lean while carrying visceral adiposity, insulin resistance, high blood pressure or abnormal lipids. Those abnormalities do not become less consequential simply because BMI remains below an obesity threshold.
Conversely, some people living with obesity can maintain relatively favourable blood pressure, glucose and lipid profiles, at least for a period of time.
The temptation is to turn that observation into the idea that “metabolically healthy obesity” is benign.
The new analysis does not support that conclusion either.
Higher BMI and poorer metabolic health were both independently associated with increasing cardiovascular risk for several outcomes, and the combination of the two was consistently the most adverse. Even when obesity occurred without the metabolic abnormalities used to define an unhealthy phenotype, risk was not necessarily equivalent to that of metabolically healthy people in the normal-weight category.
The term “metabolically healthy obesity” therefore describes relative risk, not absence of risk.
That distinction has become increasingly important because metabolic health is not fixed.
A patient who has obesity but normal glucose, blood pressure and lipids at 35 may not remain in the same metabolic category at 45. Longitudinal studies have repeatedly shown that some people transition from metabolically healthy to unhealthy states over time, particularly as age, visceral adiposity and other risk factors accumulate.
A snapshot can therefore be useful without being permanent.
This is where the new analysis becomes relevant to cardiovascular prevention rather than simply to debates over obesity terminology.
Risk assessment often begins with visible or easily measured factors, and BMI is among the easiest. It requires no blood test, no imaging and almost no cost. At population level, it remains useful for identifying broad relationships between excess body weight and disease.
But cardiovascular prevention increasingly depends on what BMI cannot reveal.
It cannot determine where fat is stored. It does not separate adipose tissue from muscle. It does not show whether blood pressure is elevated, whether triglycerides are high, whether insulin sensitivity is deteriorating or whether kidney disease is emerging.
The same BMI can therefore conceal very different cardiovascular biology.
Waist circumference begins to address part of that limitation because central adiposity more closely reflects visceral fat, but even waist measurement remains only one component of a much broader metabolic picture.
The direction of cardiovascular guidelines is already moving accordingly.
The 2026 US cardiovascular-kidney-metabolic guideline explicitly treats obesity, diabetes, kidney disease and cardiovascular risk as interconnected rather than independent conditions. That framework makes it increasingly difficult to justify evaluating excess weight without simultaneously asking what is happening to glucose, blood pressure, lipids, kidney function and overall cardiovascular risk.
The meta-analysis adds a useful warning from the opposite direction.
Clinicians should not reserve that metabolic scrutiny only for people whose BMI is high.
A normal-weight patient with an adverse metabolic profile may have substantially greater cardiovascular risk than appearance or BMI alone suggests.
That has implications for screening. If blood pressure, glycaemia and lipids are checked less aggressively because a patient looks lean, the system may miss precisely the phenotype the analysis identifies as potentially high risk.
It also has implications for the increasingly political debate around obesity treatment.
The arrival of highly effective anti-obesity medicines has intensified questions about who should receive therapy, how health systems should define medical need and whether BMI thresholds are sufficient for allocating expensive long-term treatment.
The new evidence does not provide a ready-made reimbursement formula, and it would be a mistake to use an epidemiological meta-analysis to redraw treatment eligibility by itself.
What it does show is why simple thresholds become increasingly uncomfortable as treatment becomes more personalised.
Two patients with the same BMI may have very different immediate cardiovascular risks. A third patient below an obesity threshold may already have hypertension, dyslipidaemia and impaired glucose regulation that place them on a worse cardiovascular trajectory than either.
A prevention system interested in outcomes rather than labels has to recognise those differences.
There is also a danger in taking the argument too far.
If metabolic health becomes the dominant focus, clinicians could reassure patients with obesity simply because current laboratory results are normal. That would ignore evidence that obesity itself can contribute to heart failure, atrial fibrillation, sleep apnoea, osteoarthritis and other conditions, while metabolic status may worsen over time.
The more useful interpretation is not metabolic health versus obesity.
It is metabolic health and obesity.
The meta-analysis found a dose-response relationship in which worsening metabolic health and increasing BMI both contributed to cardiovascular risk, with the greatest burden appearing when the two converged. Stroke was somewhat different, with metabolic health appearing to play the stronger role, but the overall pattern remained one of interacting rather than competing risks.
That makes cardiovascular prevention more complicated than sorting patients into a few BMI categories.
It also makes it more clinically realistic.
People do not develop cardiovascular disease because they crossed an arbitrary body-mass threshold. Risk accumulates through blood pressure, glucose, lipids, inflammation, adipose distribution, kidney disease, smoking, physical activity, genetics and many other interacting factors.
BMI captures one part of that story.
The problem arises when it is allowed to represent the whole story.
The new analysis therefore adds to a broader move away from treating obesity as a uniform cardiovascular state. The future of risk assessment is likely to involve more attention to phenotype, including where fat is distributed, which metabolic abnormalities are present and whether those abnormalities are stable or worsening over time.
That approach should not minimise obesity.
It should make obesity care more precise.
The cardiovascular lesson is not that weight no longer matters. It is that body size and metabolic health tell different parts of the risk story, and clinicians lose information when either one is allowed to speak for the other.
