The research team categorized study participants into six distinct phenotypes by crossing three body mass index categories with two metabolic states. People could be metabolically healthy or unhealthy at normal weight, overweight, or obese. Metabolic health in these studies was typically defined by the presence or absence of clustered abnormalities such as elevated blood pressure, dyslipidemia, high fasting glucose, and related markers of the metabolic syndrome. The reference group for all comparisons was the metabolically healthy normal-weight individual, the profile conventionally assumed to carry the lowest cardiovascular risk. Using random-effects models to pool hazard ratios and their 95 percent confidence intervals, the investigators quantified how each phenotype fared across five outcomes: overall cardiovascular disease, coronary heart disease, myocardial infarction, stroke, and heart failure.

The headline numbers are striking. Compared with metabolically healthy people of normal weight, those with metabolically unhealthy obesity faced roughly double the risk across nearly every cardiovascular endpoint examined. Their hazard ratios reached 2.10 for overall cardiovascular disease, 1.97 for coronary heart disease, 2.20 for myocardial infarction, and 2.05 for heart failure, with stroke at 1.55. Every confidence interval sat comfortably above unity, signaling that these elevations are unlikely to be statistical noise. On its own, that finding would surprise no cardiologist. The twist lies in the comparisons that follow.

When the researchers arranged the phenotypes along a gradient of body size and metabolic status, a clear dose-response pattern emerged for cardiovascular disease, coronary heart disease, myocardial infarction, and heart failure: risk climbed steadily with both higher body mass index and deteriorating metabolic health, and the combination of the two conferred the greatest danger of all. Yet the analysis also revealed that people who were metabolically unhealthy but normal weight frequently faced risks that exceeded those of individuals who were obese but metabolically healthy. In other words, a lean person with high blood pressure, adverse blood lipids, or elevated glucose may be in more immediate cardiovascular danger than a heavier person whose metabolic panel remains clean. Body size, in this framing, is not the primary engine of risk; the metabolic machinery underneath it is.

Stroke emerged as the most dramatic illustration of this principle. For this outcome, the data pointed to poor metabolic health as the dominant determinant, with body mass index contributing comparatively little once metabolic status was accounted for. The finding carries mechanistic plausibility. Stroke risk is tightly linked to hypertension, atrial fibrillation, and diabetes-related vascular damage, all of which are metabolic disturbances rather than consequences of body mass per se. Elevated blood pressure damages cerebral arteries over decades, and impaired glucose metabolism accelerates atherosclerosis in the vessels supplying the brain. If excess weight raises stroke risk mainly by degrading metabolic health, then a normal-weight individual with undiagnosed hypertension or insulin resistance would be expected to lose the protective advantage that leanness is supposed to confer, and that is precisely what the pooled data show.

The study’s methodology reflects the current standards for evidence synthesis in observational epidemiology. The authors searched PubMed and Scopus for literature published through January 2025, followed PRISMA reporting guidelines, and assessed the methodological quality of included cohorts using the ROBINS-I tool, which is designed to evaluate bias in non-randomized studies. Statistical pooling employed random-effects models, an appropriate choice given the inevitable heterogeneity among cohorts that differ in geography, age structure, follow-up duration, and the specific criteria used to define metabolic health. The authors also applied techniques for deriving comparable effect estimates across studies that reported their results in different formats, and they examined publication bias using the classic funnel-plot asymmetry test developed by Egger and colleagues.

That heterogeneity deserves attention, because it cuts to the heart of a long-running controversy in this field. The definition of metabolically healthy obesity has never been standardized, and studies have used anywhere from one to five metabolic criteria, different thresholds for each, and varying follow-up windows. Prior work has shown that shifting the definition can materially change the estimated risk, and the new analysis inherits that ambiguity. There is also a deeper biological complication: metabolic health is not a fixed trait. Longitudinal studies, including large analyses of the Nurses’ Health Study and the UK Biobank, have demonstrated that a substantial share of people classified as metabolically healthy at baseline convert to metabolically unhealthy over follow-up, and that this transition sharply raises cardiovascular risk. A single baseline snapshot may therefore understate the true hazard carried by people who appear healthy on the day they enroll.

Even with those caveats, the consistency of the signal across 44 independent cohorts, multiple continents, and five distinct cardiovascular endpoints lends the conclusion considerable weight. The findings align with a growing body of evidence that has steadily eroded the notion of benign obesity. Earlier meta-analyses found that even metabolically healthy overweight and obesity were associated with elevated cardiovascular risk, and large single-cohort studies of millions of participants reached similar conclusions. What the new analysis adds is a head-to-head quantification of the two risk dimensions, showing that the metabolic axis generally dominates the adiposity axis, and that the two interact multiplicatively rather than redundantly. Carrying excess weight and having poor metabolic health together is worse than either alone.

The clinical and public health implications are substantial. First, the results argue against using body mass index as a stand-alone triage tool. A normal-weight patient with elevated blood pressure, triglycerides, or fasting glucose should not be waved through on the strength of a favorable BMI, particularly given the doubled stroke risk observed in metabolically unhealthy normal-weight individuals. Second, the findings do not give people with obesity a free pass. Metabolically healthy obesity still carried elevated risk in this analysis, and the transition studies suggest that this state is often a way station rather than a stable destination. Weight management remains relevant, but the data suggest that interventions targeting blood pressure, lipid profiles, and glucose metabolism, through lifestyle change or pharmacotherapy, may deliver cardiovascular benefits that are at least as large as those from weight loss alone.

The authors conclude that worsening metabolic health and higher body mass index each independently increase the incidence of cardiovascular disease, coronary heart disease, myocardial infarction, and heart failure, with their combination posing the greatest threat, while stroke risk is driven mainly by metabolic status. As obesity rates continue to climb globally and health systems grapple with where to direct finite screening and prevention resources, this analysis offers a reframing that is both humbling and actionable: the scale tells only part of the story, and the blood panel may tell the more important half. For clinicians, the message is to look past body size and interrogate metabolic function in every patient; for researchers, it is to refine the definitions and track the transitions that determine who converts from apparent health to genuine risk.

Subject of Research: The relationship between body mass index, metabolic health phenotypes, and cardiovascular disease incidence

Article Title: Metabolic health is a stronger determinant than obesity for cardiovascular disease incidence: a meta-analysis of longitudinal cohort studies

Article References: Ghorbani, Z., Soltani, S., Aune, D., Khandouzi, N., Noormohammadi, M., & Kazemi, A. (2026). Metabolic health is a stronger determinant than obesity for cardiovascular disease incidence: a meta-analysis of longitudinal cohort studies. International Journal of Obesity. https://doi.org/10.1038/s41366-026-02237-6

Image Credits: AI Generated

DOI: 10.1038/s41366-026-02237-6

Keywords: metabolic health, obesity, cardiovascular disease, meta-analysis, body mass index, stroke, coronary heart disease, heart failure, myocardial infarction, metabolic syndrome, cohort studies, epidemiology