Medical Disclaimer

What the score is — and isn't

In plain terms: MarkerScore takes a handful of everyday health markers you enter — things like step count, sleep, resting heart rate, and blood pressure — and compares them to associations reported in large published studies of other adults. The result is a weighted index, on a 0–100 scale, that says how your inputs line up with patterns linked to lower all-cause mortality risk in those studies. That is the whole product.

It is not a diagnosis. It is not treatment advice. It does not predict your biological age or how long you will live — we deliberately don't offer a "biological age" number or a lifespan estimate, because no combination of a handful of self-reported markers has been clinically validated to predict an individual's aging rate or longevity. If you've seen a questionnaire elsewhere give you a biological age in years, ask what it was validated against before trusting the number.

The composite is a bounded weighted index of single-marker associations. It has not been validated as a personal mortality, biological-age, or lifespan result. No study has measured all eight markers on the same people and counted deaths. A one-point move on the score is not a quantified drop in death risk. A death finding stays on its source study and on that study’s kind of death. Each underlying marker-to-risk relationship comes from its own study population. And everything here reflects association, not causation: people whose habits look like the "lower risk" end of a study weren't necessarily made lower-risk by that habit — healthier people are also more likely to have that habit for other reasons the studies can't fully separate out.

How to interpret your score

Your score compares your eight numbers with a typical 45-year-old's. 50 means you match them on all eight; above 50, your numbers look more like those of people with lower death rates in large studies. The scale goes to 100; that is a cap, not a test score. It's a comparison, not a prediction.At or below 120/80 the blood-pressure curve is flat. A lower reading in that range is not a score gain. Lewington et al. 2002 reports vascular death, not all-cause mortality, and that finding is not a score payoff below 120/80. Zero days of strength training is the reference. It does not lower the score. The marker can only raise the score. Counting each day as 30 minutes is an assumption. Your score and percentile are a directional summary of how your entered metrics compare to published, population-level associations. A higher score means your inputs line up, on average, with patterns reported in those studies. It does not mean you personally will or will not experience any particular health outcome. Small changes in the score should not be over-interpreted; treat it as a rough compass for where to focus attention, not a diagnosis or a guarantee. For the full mechanics of how the score is built — including where curve shapes are directly from the literature versus interpolated — see Methodology & Sources.

Known limitations of the model

• The eight scored inputs are correlated with one another in ways the model does not fully account for.
• Grip strength, sitting time, and extra aerobic minutes are not scored. They do not move the number.
• The modeled ranking is an age- and sex-group comparison shown as a range. It combines U.S. survey and exercise-testing data, proxy-measurement and assumed-dependence sensitivity runs, and core survey sampling uncertainty. It is not a medical result or a fully validated population percentile. The 70+ group has fewer data and a wider range.
• Some scored shapes are constructed or assumed: VO2 max knots past the cited range, the blood-pressure joint rule above 120/80, interpolated interiors on the fiber curve, and the 30 minutes counted for each strength-training day. Grip strength’s curve is also constructed, and grip is not one of the eight scored inputs.
• Self-reported and home-measured scored inputs — blood pressure, VO2 max, and sleep — carry real measurement error depending on your equipment and technique. Grip strength is not a scored input.
• The model estimates relative, population-level mortality association. It does not model any specific disease, and it does not account for family history, genetics, current medications, or conditions you have not entered.
• Every user is scored against the same fixed ~45-year-old reference profile, regardless of the age you enter. Several underlying studies report that a marker's health association changes with age (for example, daily step count and resting heart rate), and that age-adjustment is not currently applied — your score reflects your distance from a 45-year-old reference, not from someone your own age.

See Methodology & Sources for the complete, current list of limitations and the primary study behind every metric.

Who should not rely on this tool

The underlying studies enrolled general adult populations. This calculator was not validated as a combined predictor of death, biological age, or lifespan. It is not appropriate for, and should not be used by or relied upon for:

• Pregnant users — pregnancy substantially changes normal ranges for blood pressure, heart rate, weight, and other metrics used here, and none of the underlying studies were conducted on pregnant populations.
• Users under 18 — the reference cohorts and hazard ratios are derived from adult populations and do not apply to children or adolescents.
• People with a diagnosed chronic disease already managed by a clinician — for example diabetes, cardiovascular disease, chronic kidney disease, or cancer. Your treating clinician's guidance, informed by your full history and clinical exam, always takes precedence over a general population estimate.

Ways this gets misused — please don't

• Don't start, stop, or change a medication because your score went up or down. Medication decisions belong to you and your prescriber, based on your full history and exam — not a population-average estimate.
• Don't use it to rule a condition in or out. A high score is not evidence you don't have a disease, and a low score is not evidence you do. It has no diagnostic accuracy for any specific condition, because it was never designed or tested to have one.
• A good score doesn't mean you're healthy, and a low one doesn't mean you're sick. It reflects eight markers you chose to enter, nothing about the rest of your health, and nothing your clinician would find on an exam or in labs the calculator doesn't ask for.
• Don't skip or delay a screening or checkup because your score looks reassuring. Routine screening exists to catch things a self-reported score structurally cannot.
• Treat wearable-derived inputs as estimates, not measurements. Step counts, resting heart rate, sleep stages, and VO2 max from a phone or watch vary by device and algorithm — sometimes substantially — so your score can shift when you change devices even though nothing about your health has changed.

Clinical tools your clinician can use alongside this

MarkerScore is education and self-tracking, not a clinical risk calculator, and it is not trying to compete with the tools clinicians actually use for cardiovascular risk. If you want a clinically validated estimate, talk to your clinician about running one of these — they take inputs like cholesterol and kidney function that MarkerScore doesn't collect:

• The American Heart Association's PREVENT™ calculator — a contemporary tool clinicians use to estimate 10- and 30-year cardiovascular risk.
• The ACC/AHA CVD Risk Estimator Plus — the current version of the ASCVD Risk Estimator, now combined with PREVENT, built for clinician-patient conversations about prevention.

PREVENT comes from the American Heart Association and the Risk Estimator Plus from the American College of Cardiology. Both are built for clinicians, not for self-diagnosis — bring the result to your clinician rather than acting on it alone.