In one sentence
In the card's own words: metabolic age compares your weight, height and waist with people of each age in a US reference population; it is a body-shape score expressed in years, and it describes a measure, not a diagnosis.
Why it matters
The number is S-AnthropoAge, the simplified form of AnthropoAge: sex-specific models of ten-year survival fitted on 18,794 adults in the US NHANES surveys, with body mass index and the waist-to-height ratio as the body terms, inverted so that the answer is the age in the reference population whose survival matches yours (Fermín-Martínez 2023). The full version separated those who died within ten years from those who did not with an area under the curve of 0.876, independently of ethnicity, sex and existing conditions, and the simplified version preserved every result (intraclass correlation 0.994 between the two). In 57,080 people aged 50 to 94 in the US, England, Mexico, Costa Rica and China, the gap between this age and calendar age added to calendar age's own discrimination of mortality (c-statistic 0.806 against 0.803), and a gap above zero was associated with about 37 percent higher mortality independent of age and covariates, and with new deficits in daily living (Fermín-Martínez 2025). It replaces the "metabolic age" most scales print, which compares a bioimpedance estimate of basal metabolic rate with an unpublished table of typical rates by age. That construction cannot be rescued: once fat-free mass and fat mass are held constant, expenditure is flat from 20 to 60 in 6,421 doubly-labeled-water measurements across the life course (Pontzer 2021), so any number built on it is body composition expressed in years, with a precision of ten to thirty years. VT's method note sets out the comparison and why the change was made.
What moves it
Waist, beyond what your weight would account for. At the median body, each centimeter of waist at a fixed weight reads about 0.35 years older for men and 0.29 for women; each kilogram at a fixed waist reads about 0.27 years younger for men and 0.15 for women, because the model's weight term is U-shaped and a heavier person at the same waist carries more lean mass; a typical five-kilogram gain, with the waist growing as it usually does, reads only 0.2 to 0.4 years older. Proportional fat loss, which shrinks weight and waist together, moves it little, and body fat is not an input once you have measured your waist, which is why body fat is shown beside it as a percentile. For a fixed body it rises one year per year of calendar age. Day-to-day weight moves it by about 0.2 years; a 1.5 cm error in the waist is half a year, so the tape, not the scale, is the error that matters.
How to measure it well
Measure the waist as the reference population was measured, at the top of the hip bone (the iliac crest), tape level, snug without pressing, at the end of a normal breath out, in the morning; the midway-between-rib-and-hip site many guides use differs by a few centimeters, and three centimeters is about a year (see How to measure your waist). Every two to four weeks, dated. Weight from the scale, any morning; the daily line uses that day's weight, or the nearest weigh-in within 30 days. Sex and date of birth from your profile.
How Vitality Tracker uses it
In the website calculator, as a whole number of years with the difference in words ("4 years younger than your age", "about your chronological age" when it is within two years); in the app, to one decimal place, as a line over time on the Metabolic Health page. No plus-or-minus range is printed, because the measurement noise is small and the population spread is not a measurement error. A measured waist gives the best result; until you log one, VT estimates your waist from your body fat and says so. The reference population's limits are respected: a weight-for-height or waist-for-height outside the 1st to 99th percentile of the reference uses the nearest reference value and says so, an age under 20 or over 90 is marked as an extrapolation, and a profile without a sex set shows the average of the men's and women's results with a note. Ethnicity is fixed at the reference level, because VT does not collect it. The waist is date-stamped; its date shows on the chart, and when it is more than about a month old VT asks for a fresh one, because a weight gain at a stale waist reads younger than it should. The number is calibrated so that the 2017 to 2020 US adult population reads its own age on average, within 0.1 years for men and 0.4 for women in every decade. Beside it, body fat is shown as a percentile for your age and sex against the NHANES DXA reference (Kelly 2009), which carries the signal this number does not.
How far to trust the number
Calculated, and honest about what it is. The spread of the difference from calendar age across the reference population is 4.5 years for men and 3.85 for women, so a difference inside two years is "about your age" and a difference of five is unusual; the measurement noise is 0.2 to 0.6 years. The reference is American; a New Zealand member is compared with US adults, and the authors themselves say the method may need recalibration in other populations. A member who would be recorded as "Other" in the US survey carries a value two to four years lower than the model would give them, and nothing in the calibration absorbs this. It uses no body fat, glucose, HbA1c or lipid, so it will feel undramatic beside a scale's figure, which can sit fifteen years from your age; the blood markers have their own numbers (PhenoAge, the Vitality Score's blood sugar area) and body fat has its percentile.
Sources
- Fermín-Martínez CA, et al. AnthropoAge, a novel approach to integrate body composition into the estimation of biological age. Aging Cell 2023;22:e13756. doi:10.1111/acel.13756
- Fermín-Martínez CA, et al. Multinational evaluation of AnthropoAge as a measure of biological age in the USA, England, Mexico, Costa Rica, and China: a population-based longitudinal study. npj Aging 2025;11. doi:10.1038/s41514-025-00232-1
- Pontzer H, et al. Daily energy expenditure through the human life course. Science 2021;373:808–12. doi:10.1126/science.abe5017
- Kelly TL, Wilson KE, Heymsfield SB. Dual energy X-ray absorptiometry body composition reference values from NHANES. PLoS One 2009;4:e7038. doi:10.1371/journal.pone.0007038
- Vitality Tracker, Metabolic age method: research and recommendation (October 2026).