“How old am I, really?” The question comes up often in our practice now that tests promise a biological age different from the one on your ID card. The same term covers very different things: the bathroom scale, a mail-order genetic test, a panel of blood biomarkers. This article sorts through them and proposes a reasonable use of metabolic age.
What is metabolic age?
Metabolic age is the age to which a person’s metabolic profile corresponds, on average, in a reference population: if your biomarkers resemble those of a younger person, your metabolic age is lower than your chronological age, and vice versa. It is a statistical estimate, not a direct measurement, and its value depends on what is measured and on the comparison population.
Today the term covers two uses that must be distinguished from the outset.
- The “metabolic age” of bioelectrical impedance scales. The device estimates resting energy expenditure (weight, height, age, sex, low-intensity electrical current) and compares it with an average by age. The number gives a rough indication of body composition; it is not based on any blood biomarker and has not been validated against health outcomes.
- Metabolomic age, derived from a blood panel. A statistical model, trained on tens of thousands of nuclear magnetic resonance (NMR) metabolomic profiles, estimates a person’s age from their biomarkers; the gap with chronological age is the useful information. This is the approach discussed here, known as MileAge in the literature.
“Metabolic” refers to metabolism, all the chemical reactions through which the body converts nutrients into energy and building blocks; a metabolic profile describes how a person handles sugars, fats and proteins at a given moment.
Chronological age, biological age, metabolic age: what are the differences?
Chronological age is the time elapsed since birth; biological age is an estimate of the body’s state of aging; metabolic age is one way of estimating that biological age, from metabolic biomarkers. There is no single biological age, but as many estimates as there are methods.
| Concept | What it measures | How | Main limitations |
|---|---|---|---|
| Chronological age | The time elapsed since birth | Civil records | Says nothing about health status |
| Biological age | The body’s state of aging, relative to a population | Several methods: epigenetic clocks, metabolomic age, composite clinical scores | No single definition; each method gives a different result |
| Epigenetic age | DNA methylation at sites that change with age | Blood or saliva sample, methylation analysis | Cost, availability; several clocks that are not interchangeable |
| Metabolomic age | The profile of blood metabolic biomarkers compared with a reference population | NMR metabolomics panel, statistical model (MileAge) | Sensitive to recent habits (fasting, diet); no individual standard |
| Scales’ “metabolic age” | An estimated resting energy expenditure, compared with averages by age | Bioelectrical impedance | No validation against health outcomes |
These concepts complement each other: epigenetic age reflects slow, partly programmed changes; metabolomic age reflects a more current state, influenced by lifestyle. The biomarkers behind metabolomic age are described in our guide to blood biomarkers.
How is biological age measured? Epigenetic clocks and metabolomic age
The two best-documented approaches rest on the same principle: train a statistical model to predict chronological age from biological measurements, then read the gap between predicted and actual age as an index of accelerated or slowed aging. They differ in what they measure.
Epigenetic clocks
Epigenetic clocks rely on DNA methylation, a chemical modification that regulates gene activity and whose level, at certain sites in the genome, changes predictably with age. According to the landmark review by Steve Horvath and Kenneth Raj (Nature Reviews Genetics, 2018), these clocks estimate age accurately in most tissues, and “epigenetic age acceleration” measured in blood is associated with lifespan, even after accounting for the usual risk factors. The first multi-tissue clock was published by Horvath in 2013.
Metabolomic age (MileAge)
Metabolomic age applies the same logic to blood biomarkers measured by NMR: a model trained on the profiles of UK Biobank participants predicts their age, and the gap between predicted and actual age is the result. This approach, referred to by the acronym MileAge, was developed by Julian Mutz and colleagues. It uses the same blood sample and the same panel of 249 biomarkers as the rest of the metabolomic analysis (see NMR metabolomics) and responds to lifestyle habits, which is why it is useful for tracking a trajectory.
This sensitivity is also its weakness: a profile measured after a large meal, an infection or a sedentary period will not yield the same age as one taken fasting on an ordinary morning. Metabolomic age is a snapshot, to be compared with a second measurement before drawing any conclusion.
What does the scientific literature show about metabolic age?
The literature shows that blood metabolomic profiles, of which metabolomic age is a summary, are associated with all-cause mortality and with the risk of several age-related diseases, including dementia, in large cohorts, without allowing any certain prediction for an individual.
Four findings shape this knowledge:
- Mortality. In 44,168 people from twelve cohorts, fourteen independent metabolic biomarkers were associated with all-cause mortality and predicted 5- and 10-year mortality better than conventional risk factors (Deelen et al., Nature Communications, 2019): the blood metabolome carries information about lifespan.
- Age-related diseases. In 117,981 UK Biobank participants, with validation in four cohorts, a metabolomic profile combined with age and sex matched or outperformed established predictors for the 10-year onset of fifteen of the 24 diseases studied, and added information to clinical variables for eight of them, including dementia, type 2 diabetes and heart failure (Buergel et al., Nature Medicine, 2022).
- Atlas of associations. In 118,461 UK Biobank participants, the 249 biomarkers of the NMR panel were associated with more than 700 diseases, well beyond the cardiometabolic field (Julkunen et al., Nature Communications, 2023).
- Metabolomic age and dementia. Mutz and colleagues examined the link between the metabolomic age gap (MileAge) and the risk of dementia (Alzheimer’s & Dementia, 2026), extending to neurodegenerative diseases an indicator first validated against mortality.
As for epigenetic clocks, Horvath and Raj (2018) report associations with lifespan, physical and cognitive fitness, certain neurodegenerative diseases and cancer risk.
What these studies do not show deserves equal clarity: none establishes that a given metabolic age corresponds to a given life expectancy for an individual, or that a decrease in that age automatically leads to a decrease in risk. The age gap is one marker among others, and its place relative to validated scores such as SCORE2 remains to be determined (see 10-year cardiovascular risk).
What are the limits of biological age tests?
Biological age tests have four main limitations: measurement variability, the lack of a standard across methods, dependence on the training population, and the absence of any norm or “ideal” age. These limitations do not invalidate the tool; they define its reasonable use.
- Variability. Every biological measurement fluctuates from day to day (fasting, sleep, activity, a minor infection), from one laboratory to another, and through the randomness of sampling. A difference of a few years between two samples taken close together may be nothing more than noise: a single measurement is not enough to draw a conclusion.
- Lack of a standard. There is no official definition of biological age, no reference method, and no body that certifies these tests for this use. Two clocks or two models applied to the same person can produce different ages; comparing the results of two methods is meaningless.
- Dependence on the reference population. A model trained on UK Biobank (British volunteers aged 40 to 69 at recruitment) applies with more confidence to adults with that profile than to very young people, very old people or people of other ancestries.
- No ideal age. A metabolic age lower than chronological age is, on average, associated with a lower risk in cohorts; that does not make “ten years younger” a health goal. Presenting biological age as a score to optimize turns a risk marker into a performance target.
As for free online “biological age tests,” they are questionnaires that assign points to lifestyle habits, with no biological measurement: sometimes educational, they measure no age at all.
Can you influence your metabolic age?
The biomarkers that make up metabolomic age respond to lifestyle habits; the profile can therefore change, but no individual result can be promised, and we do not use the word “rejuvenation.” The value of the indicator lies in tracking a trajectory, not in reaching a number.
The levers are those of cardiovascular and metabolic prevention, recommended regardless of any test: quitting smoking, regular physical activity, a balanced diet, sufficient sleep, limited alcohol, and monitoring weight, blood pressure and blood sugar with your doctor. Metabolomic age does not add a lever, but a way of documenting, several months apart, how the profile has changed.
In our practice, we use the age gap in three ways: as a starting point for discussing habits without blame; as a point of comparison for a later sample; and, when the gap is clearly unfavorable, as a signal to check validated risk scores and, if needed, refer for targeted tests. Never as a verdict.
How do we, as doctors, view metabolic age?
Metabolic age is a useful educational and monitoring indicator when it is measured with a documented method, interpreted by a doctor and placed within a complete check-up; it becomes misleading when presented on its own, as a score to improve or as an individual prediction. The difference lies in the framework, not in the number.
That framework is personalized health: start from the covered care pathway, first and foremost the prevention check-up, fully covered (100%) by French National Health Insurance (Assurance Maladie) at four key ages; estimate risks with validated tools; then, if it could change a decision, refine with extended biomarkers. Metabolic age is a summary within that framework, not an entry point.
Key takeaways
- Metabolic age is a statistical estimate: the age to which your biomarker profile corresponds, on average, in a reference population; it has nothing to do with the number on a scale.
- Epigenetic clocks and metabolomic age (MileAge) are the two documented approaches to biological age; they are not interchangeable and no standard exists.
- In large cohorts, metabolomic profiles are associated with mortality and with the risk of several age-related diseases, including dementia: population associations, not individual predictions.
- There is no ideal metabolic age; the indicator is used to track a trajectory, never to promise rejuvenation.
- Its interpretation is the doctor’s job, within a complete check-up and the care pathway.
What Sokrate lets you do
The Sokrate service prepares your prevention check-up online with an adaptive questionnaire (Sokrate 360, nine dimensions), after which a doctor writes and signs your Personalized Prevention Plan, sent to your regular doctor (médecin traitant) unless you object. As an option, an NMR metabolomic analysis of 249 biomarkers, performed on the Nightingale Health platform, is interpreted by a doctor: metabolomic age is presented as a trajectory indicator, with its limitations, alongside validated risk scores. A coordinating nurse handles follow-up. See how the Sokrate pathway works or join the waitlist.