The test
Heart rate recovery
How quickly your heart rate falls in the minute after hard exercise. The threshold is solid and the percentile beside it is not, and this page is careful about which is which.
What you need. A heart rate monitor or watch, and somewhere you can keep walking for a minute afterwards. Any exercise you like to get there — bike, brisk walk uphill, rower, stairs, whatever you have.
How to do it
Doing it properly
This test raises your heart rate substantially. If you have a heart condition, chest pain on exertion, or have been told to avoid vigorous exercise, do not do this test — leave it blank and your other results will still produce a score.
- The screen shows your target heart rate.
- Exercise in any way you choose until you reach it, then hold it for one minute.
- Record your heart rate — that is your peak.
- Then keep walking at a slow, easy pace. Do not stop and do not sit down.
- Record your heart rate again exactly one minute later, still walking.
What counts
Your peak reading minus your reading one minute later. Slow means roughly half your normal walking pace — slow enough that going slower would feel awkward. The pace is what makes one test comparable with the next.
What most people get wrong
- Stopping or sitting down, which is a different test with different thresholds.
- Walking at your normal pace rather than an easy one.
- Stopping before the target is reached.
- Guessing rather than measuring.
The evidence
What it actually shows
Cole and colleagues established the threshold in 1999: recovery measured during an active cool-down, taken as peak heart rate minus heart rate one minute later, with fewer than twelve beats marking abnormal recovery.
Sydo and colleagues confirmed it in a group much closer to this audience — 19,551 adults aged 30 to 79 with no cardiovascular disease, followed for a median of twelve years. Recovery below thirteen beats carried an adjusted hazard ratio of 1.56 for death from any cause, 1.95 for cardiovascular death and 1.41 for non-cardiovascular death.
You keep walking through the recovery minute rather than stopping. That is not a detail: abnormal-recovery thresholds are specific to position and activity — twelve beats upright with an active cool-down, eighteen lying down, twenty-two seated at two minutes. Walking makes the twelve-beat threshold native to the protocol rather than borrowed from a different one.
The threshold is not a score. There is no published table of heart rate recovery by age and sex anywhere — a clinical review says so explicitly, and the age tables circulating on commercial fitness sites are labelled approximate guidelines and are not traceable to a study.
So the 0–100 score beside the flag is scored against the only distribution measured under this protocol: Cole's cohort, mean age 57 and referred for exercise testing, which is older and less healthy than the people reading this. It reads high for most of them. It counts towards the Index and it is marked provisional wherever it appears.
Sydo found abnormal recovery was not a significant predictor in people with normal cardiorespiratory fitness, nor in those taking beta-blockers, nor in current smokers. This test discriminates hardest at the deconditioned end and is close to inert at the fit end.
"Doubles your risk of dying" is wrong for this test. It is accurate for cardiovascular death and overstated for death from any cause, and the two figures are not interchangeable.
Cole CR, Blackstone EH, Pashkow FJ, Snader CE, Lauer MS. Heart-rate recovery immediately after exercise as a predictor of mortality. New England Journal of Medicine 1999;341:1351–1357.
Sydo N, Sydo T, Gonzalez Carta KA and colleagues. Journal of the American Heart Association 2018;7(7):e008143. 19,551 adults aged 30–79 without cardiovascular disease.
The next step
One test is one test.
This measures one thing. The Evolution Index measures nine, across the three domains that decide how long you stay useful, and reports your weakest rather than your average — a strong heart does not buy you out of a floor you cannot get up from.
Every score is published with the error around it, because a change smaller than the measurement error is not a change.