Metabolically Healthy Obesity: Definition, Stability Over Time, and the Disagreement

A 45-year-old woman steps off the clinic scale: 92 kg, 1.65 m, BMI 33.8 kg/m², obesity class I. Her blood pressure is 118/76, fasting glucose 5.0 mmol/L, triglycerides 1.2 mmol/L, HDL cholesterol 1.6 mmol/L. She has no metabolic syndrome criteria by any standard definition. This is the puzzle that metabolically healthy obesity definition stability researchers have spent two decades trying to solve: does the absence of metabolic derangement mean she is safe, or is she simply not yet sick?

The term metabolically healthy obesity (MHO) describes a phenotype where an adult has a BMI at or above 30 kg/m² (WHO 1995 cut-off for obesity class I) but does not meet the threshold for metabolic syndrome. There is no single universally agreed definition. Different studies set different cut-offs for blood pressure (130/85 vs. 135/85 mmHg), fasting glucose (5.6 vs. 6.1 mmol/L), triglycerides (1.7 mmol/L), and HDL cholesterol (1.0 vs. 1.3 mmol/L in men). Some definitions allow zero metabolic syndrome components; others permit up to one. This variability alone makes cross-study comparison difficult and means the prevalence of MHO ranges from roughly 10% to 40% of adults with obesity depending on whose criteria you use, the age and sex of the sample, and the ethnic group under study.

Longitudinal studies consistently show that MHO is a transient state, not a fixed trait. The largest evidence comes from two prospective cohort studies that tracked adults over 5 to 10 year windows.

In the Australian Diabetes, Obesity and Lifestyle (AusDiab) study, Appleton and colleagues followed 4,051 adults for 5.5 to 10.3 years. Among those initially classified as MHO, 30.8% transitioned to metabolically unhealthy obesity during follow-up. The transition was not random: those who gained weight, lost cardiorespiratory fitness, or saw increases in waist circumference were far more likely to shift into the unhealthy phenotype.

The Multi-Ethnic Study of Atherosclerosis (MESA), reported by Mongraw-Chaffin and colleagues in 2018, followed 5,814 US adults for 12.2 years. The transition rate was higher: 48% of MHO individuals moved to metabolically unhealthy obesity. The difference between the two rates partly reflects the longer follow-up period and partly the stricter definition used in MESA. In both studies, the direction of travel was overwhelmingly one way: once someone lost metabolic health, they rarely regained it.

What Drives the Transition

Two factors most consistently predict the shift: accumulation of visceral adipose tissue and loss of cardiorespiratory fitness. Visceral fat, measured by CT or MRI at the L4-L5 level, is the metabolically active depot that secretes pro-inflammatory cytokines and drives insulin resistance. Waist circumference, the tape-measure proxy for visceral fat, predicts transition even in people whose weight is stable. Subcutaneous adipose tissue, by contrast, does not carry the same risk. Cardiorespiratory fitness, measured by maximal oxygen consumption during a treadmill test, independently predicts who stays metabolically healthy. In the AusDiab cohort, a one-standard-deviation drop in fitness doubled the odds of transitioning from MHO to unhealthy obesity within five years.

If MHO were genuinely healthy, people with the phenotype should have the same long-term risk of cardiovascular events and all-cause mortality as metabolically healthy normal-weight adults. The evidence says otherwise.

The two most cited meta-analyses both found elevated risk. Bell and colleagues (2015) pooled 14 prospective cohort studies covering 299,059 adults with a mean follow-up of 11.2 years. Compared with metabolically healthy normal-weight individuals, MHO adults had a 24% higher risk of all-cause mortality (RR 1.24, 95% CI 1.02-1.55) and a 45% higher risk of cardiovascular events (RR 1.45, 95% CI 1.20-1.70). Kramer and colleagues (2013) analyzed 8 prospective studies with 61,386 adults followed for 10 years or more. They found a 24% increase in all-cause mortality (RR 1.24, 95% CI 1.02-1.55). For cardiovascular events, the risk more than doubled (RR 2.11, 95% CI 1.46-3.06).

The risk is lower than for metabolically unhealthy obesity, where cardiovascular event risks are roughly three to five times higher than metabolically healthy normal weight. But it is not zero. The meta-analyses controlled for age, sex, smoking, and physical activity, so the excess risk is not explained by those confounders. The leading explanation is that MHO individuals still carry elevated levels of visceral adipose tissue even though their blood markers look normal. The fat is there; it has simply not yet triggered the systemic inflammation that drives metabolic syndrome.

What the Controversy Is Really About

The metabolically healthy obesity controversy turns on whether MHO is a stable protective subtype or a pre-disease state. One side argues that if metabolic health is defined strictly enough, the excess risk disappears. The other side points to the transition data and the meta-analysis effect sizes and argues that MHO is simply a lag phase. The research does not yet settle the argument because no study has followed a large MHO cohort for 20 to 30 years without treatment dropout. What is clear is that presenting MHO as reassurance to a patient is not supported by the evidence. A phenotype that carries a 24% to 124% higher risk of heart attack or stroke over a decade is not a clean bill of health.

Study	Cohort Size	Follow-Up	Key Finding
Bell et al. 2015	299,059	11.2 years	MHO: 24% higher all-cause mortality, 45% higher cardiovascular events vs. metabolically healthy normal weight
Kramer et al. 2013	61,386	≥10 years	MHO: 24% higher all-cause mortality, cardiovascular event risk more than doubled vs. metabolically healthy normal weight
Appleton et al. 2013 (AusDiab)	4,051	5.5-10.3 years	30.8% of MHO transitioned to metabolically unhealthy obesity
Mongraw-Chaffin et al. 2018 (MESA)	5,814	12.2 years	48% of MHO transitioned to metabolically unhealthy obesity

Obesity without metabolic syndrome is not evenly distributed across the population. It clusters in younger adults, premenopausal women, and people with higher cardiorespiratory fitness. In the NHANES database, women are roughly twice as likely as men to meet MHO criteria at the same BMI, partly because women carry a higher proportion of subcutaneous adipose tissue relative to visceral adipose tissue. Men, particularly after age 45, tend to accumulate visceral fat at lower total body weights, which makes it harder for them to maintain MHO status.

The ethnic dimension matters. For the same BMI and waist circumference, South Asian adults have higher visceral fat and are less likely to be MHO than White European adults. Black adults of African descent often have lower visceral fat and higher subcutaneous fat at the same BMI, which makes them more likely to meet MHO criteria. These differences mean that using a single BMI cut-off across ethnic groups, as the WHO standard cut-offs do, misclassifies risk. The WHO Asian cut-offs (23 kg/m² for increased risk, 27.5 kg/m² for high risk) were introduced in 2004 specifically because of this problem, but they are not universally adopted.

What Predicts MHO in a Clinical Setting

If a patient has a BMI above 30 kg/m² and normal blood pressure, lipids, and glucose, three tests tell you whether they are genuinely MHO or simply in the lag phase. First, a DXA scan for body fat percentage: MHO individuals have lower total body fat for their BMI than unhealthy obese individuals, though there is no universal cut-off by DXA. Second, a waist circumference measurement using the NIH/NHLBI protocol (≥102 cm for men, ≥88 cm for women): a high waist circumference despite normal blood markers is a strong predictor of future transition. Third, a cardiorespiratory fitness test: MHO individuals who can sustain a moderate-to-high exercise workload have a much lower probability of transitioning within five years than those who cannot.

For Clinicians

Measure waist circumference and cardiorespiratory fitness in every patient with obesity who has normal blood markers. A waist circumference above the NIH/NHLBI threshold (102 cm for men, 88 cm for women) or a low fitness level cancels any reassurance the blood numbers might give.

For Patients

If you are told you have MHO, do not treat the label as permission to ignore your weight. The transition data is clear: within a decade, roughly half of MHO individuals are no longer metabolically healthy. The phenotype was never designed as a clinical recommendation. It was a research classification to ask better questions. The question is not whether MHO exists but whether it lasts, and the answer from every prospective cohort study is that for most people it does not.