It is five measurements, not a feeling
Metabolic health has a working definition, and it is unglamorous: your blood sugar, blood pressure, blood fats, and waist are all in an optimal range, without medication holding any of them there.
An analysis of 8,721 adults in the US National Health and Nutrition Examination Survey used exactly that definition, requiring all five at once: waist circumference below 102 cm in men and 88 cm in women; fasting glucose below 100 mg/dL together with an HbA1c below 5.7%; blood pressure below 120 systolic and 80 diastolic; triglycerides below 150 mg/dL; and HDL cholesterol at or above 40 mg/dL in men and 50 mg/dL in women, with nobody taking medication for any of them.
On that definition, 12.2% of American adults were metabolically healthy (95% CI 10.9 to 13.6). The figure fell sharply with body size: fewer than a third of adults in the normal-weight range met it, 8.0% of those in the overweight range, and 0.5% of those in the obese range.
Two things follow. The first is that most people reading this will fail the definition on one component, and that is closer to normal than to alarming. The second is that “metabolically healthy” is a strict, all-five standard set by researchers, not a marketing term, and that when a product tells you your metabolic health is poor, it is almost certainly using a different and looser definition it has not published.
The five, one at a time
Fasting glucose and HbA1c describe how your body is handling sugar over a morning and over about three months respectively. Blood pressure and blood lipids describe the vascular consequences. Waist circumference stands in for where fat is being stored, which turns out to matter more than how much of it there is.
These are all ordinary clinical measurements. Your doctor probably already has four of them on file, and the fifth takes a tape measure. None of this requires a subscription.
Waist circumference earns its place
The most useful single change most people can make to how they measure themselves is to stop relying on BMI alone.
The consensus statement from the International Atherosclerosis Society and the ICCR Working Group on Visceral Obesity puts it directly: “the combination of BMI and waist circumference identifies a high-risk obesity phenotype better than either measure alone”. The panel recommends waist circumference be treated as a routine clinical measurement and describes it as a vital sign.
The evidence behind that is not subtle. Waist circumference is positively associated with mortality within every BMI category examined, from 20 kg/m² to 50 kg/m². When the two are analysed together as continuous variables, waist circumference remains a positive predictor of death while BMI is unrelated or even negatively related to it.
Two caveats travel with the measurement. There is no agreed protocol for taking it. The panel notes that no consensus exists on the optimal measurement method and that little scientific rationale is offered for any of the ones in use, so your own consistency matters more than which landmark you choose. And the threshold that counts as high risk varies substantially by ethnicity, spanning 80 to 98 cm in men and 80 to 96 cm in women across populations. A single number printed on a chart is not a universal cut-off.
What the panel does support is tracking it. Waist circumference is recommended as a simple way to judge whether a lifestyle change is working, which makes your own trend the most useful thing about it.
What one glucose reading tells you
Not much, and the diagnostic thresholds show why.
| Test | Normal | Prediabetes | Diabetes |
|---|---|---|---|
| A1C | below 5.7% | 5.7% to 6.4% | 6.5% or above |
| Fasting plasma glucose | 99 mg/dL or below | 100 to 125 mg/dL | 126 mg/dL or above |
| Oral glucose tolerance test, 2 hours | 139 mg/dL or below | 140 to 199 mg/dL | 200 mg/dL or above |
| Random plasma glucose | n/a | n/a | 200 mg/dL or above |
Those are the National Institute of Diabetes and Digestive and Kidney Diseases figures, and the important line sits underneath them: a doctor will usually use a second test to confirm a diagnosis. If one reading were sufficient, that sentence would not be there.
The A1C test has its own conditions. It is unreliable in the second and third trimesters of pregnancy, in certain anaemias and other blood disorders, and in people with some haemoglobin variants. If an A1C result and a glucose result disagree, that mismatch is information, and it belongs with the clinician who ordered the tests.
Continuous glucose monitors, without the marketing
A continuous glucose monitor is a genuinely good instrument being sold well ahead of its evidence for people without diabetes.
A 2025 review of CGM in people at high risk of diabetes and dysglycaemia is blunt about the state of play. CGM-derived data are not currently validated for the diagnosis of diabetes. There is no CGM-derived definition of normoglycaemia in people without diabetes, meaning nobody has established what a normal trace looks like, so the app comparing you against one is comparing you against an assumption. And the authors conclude that there is no strong evidence to support global implementation of CGM even in people who are at risk of developing diabetes.
Their position is that the current evidence supports research use and selective monitoring in defined high-risk situations, and that diagnostic use would require validated thresholds and a demonstration of improved outcomes that matter to patients before widespread adoption. That is a reasonable summary of where a healthy adult sits: interesting device, unvalidated interpretation.
If you wear one anyway, treat the spikes as feedback about a specific meal rather than as a verdict, and do not let a fourteen-day trace lead you into cutting out foods your diet depends on.
Reading a trend rather than a point
Every measurement in this guide moves. Glucose moves with sleep, illness, stress, and what you ate yesterday. Blood pressure moves within a single morning. Waist circumference moves with the tape and with the day.
The approach that survives that noise is boring and effective. Measure the same way, at the same time, on a fixed schedule. Compare each result to your own previous results, not to a population average. Look at quarters, not days. And write down in advance what you would change at what value: a number that would not change a decision is a number you do not need to take.
Most of what improves these five measurements is covered elsewhere on this site: it is resistance training, aerobic work, sleep, and what you eat, applied for long enough to show up. There is no metabolic-specific protocol that bypasses those.
When to take this to a clinician
Some of these numbers are yours to track. The decisions they lead to are not.
- A fasting glucose of 100 mg/dL or above, or an A1C of 5.7% or above. Both fall in the prediabetes range and belong in a conversation with your doctor, who will usually repeat the test.
- Any result in the diabetes range. This is not a “watch it for a quarter” situation. Book an appointment.
- Symptoms that go with high blood sugar: unusual thirst, passing urine much more often, unexplained weight loss, blurred vision, or unusual fatigue. Same week, regardless of what a home device says.
- Repeatedly high home blood pressure readings. Take the log to your doctor rather than acting on it yourself.
- A waist that is climbing steadily while your weight looks stable. That pattern is worth raising even when the scale is not moving.
- Before you start any medication, supplement, or restrictive diet aimed at glucose. This matters most if you already take a glucose-lowering medication, where changing your eating without adjusting the medication can cause a hypoglycaemic episode.
If you are pregnant, have a blood disorder, or are being treated for kidney or liver disease, the interpretation of several of these tests changes. Do not apply the general thresholds to yourself in those situations.
What the evidence does not show
- That 12.2% is your risk of anything. It is a prevalence estimate from a US survey against a strict all-five definition. It describes how common a standard is, not what happens to the people who miss it.
- That waist circumference causes the outcomes it predicts. The evidence behind the vital-sign recommendation is mostly observational. Waist circumference is a strong marker of risk; the studies do not establish that a change in the tape measure is what changes the risk.
- That there is one correct waist threshold. High-risk values range from 80 to 98 cm in men and 80 to 96 cm in women across ethnic groups, and the measurement protocol itself is not standardised. Use your own trend.
- That a continuous glucose monitor can tell a healthy adult anything actionable. No validated definition of a normal trace exists in people without diabetes, and no trial has shown that wearing one improves an outcome in this group.
- Which body-composition method is most reliable. We have not found a head-to-head comparison good enough to recommend one over another for home or clinic use, so this guide does not rank them. Waist circumference is recommended because it is measurable, tracked over time, and supported by the consensus statement above, not because scans do not work.
- That any of these five measurements can be interpreted on its own. The definition is deliberately all-five-at-once. A good glucose result does not offset a rising blood pressure, and no single number summarises the set.