VO₂max and longevity: what ‘five times the risk’ actually means

A hazard ratio compares how often something happens in two groups over the same stretch of time. Here the event is death. A ratio of 1.0 means the same rate in both groups. A ratio of 0.20 means one group died at one fifth the rate of the other. A ratio of 1.41 means a rate 41% higher.

It is a relative measure. It tells you nothing about how many people died in each group. If the reference group’s risk over a decade is small, one fifth of it is smaller still, but the absolute gap depends on age and health.

"Adjusted" means a statistical model tried to account for other measured differences between the groups, such as diabetes and heart disease. It cannot account for differences nobody measured.

The Cleveland Clinic study followed adults referred for treadmill testing between 1991 and 2014, for a median of 8.4 years, during which 13,637 died. Their mean age was 53.4.

People were grouped by their fitness percentile for their age and sex: low was under the 25th percentile, elite was the 97.7th and above. Elite against low gave an adjusted hazard ratio of 0.20, with a 95% confidence interval of 0.16–0.24. The interval is the range of values the data are reasonably consistent with.

The comparison that made the news was a smaller one. Below-average fitness against above-average carried a hazard ratio of 1.41 (1.34–1.49). In the same model, smoking was 1.41, diabetes 1.40 and coronary artery disease 1.29. That puts the associations on a similar scale. It does not show that low fitness harms you the way smoking does.

Elite still beat high fitness (0.77, 0.63–0.95), including in people aged 70 and over (0.71, 0.52–0.98). The authors found no upper limit to the benefit. That means no observed harm at the top, in a small elite group. It does not prove that more fitness keeps adding benefit.

The elite thresholds, in METs:

• Men: 40–49 at least 14.7, 50–59 14.0, 60–69 13.0, 70–79 11.5, 80+ 10.0
• Women: 40–49 13.3, 50–59 13.0, 60–69 11.1, 70–79 10.0, 80+ 8.4

A study of 750,302 US veterans aged 30–95 found the same pattern at every age, including 110,637 people in their 70s and 26,989 in their 80s. The least fit fifth against the extremely fit had a hazard ratio of 4.09 (3.90–4.20). The lowest risk came at about 14.0 METs: 0.24 in men and 0.23 in women. There was no rise in risk at extreme fitness.

Meta-analyses put a figure on each step. Across 33 cohorts, each 1-MET higher fitness went with a relative risk of death of 0.87, that is 13% lower. The paper notes that 1 MET is about a 1 km/h faster running speed. A relative risk reads the same way as a hazard ratio here.

More recent reviews land in the same place. Han and colleagues, pooling 34 cohorts, found 0.88 per MET. Lang and colleagues, reviewing 26 systematic reviews with more than 20.9 million observations, report 0.89 and 0.83 per MET, or 11–17% lower risk, and about 0.47 for high against low fitness: roughly half the risk.

Every one of these studies is observational. Fit people differ from unfit people in many ways: illness they already have, genes, income, other habits. Adjustment removes some of that, never all of it. Illness can also lower fitness before it shows up as a diagnosis, which makes low fitness look more dangerous than it is.

The fitness numbers are not lab-measured VO₂max either. In both the Cleveland Clinic and the veterans studies, METs were estimated from treadmill speed, grade or workload. The Cleveland group was a referred clinical population, and only about 6% of the veterans were women.

Lang’s overview rated the certainty of the evidence very low to moderate. A per-MET figure describes how people with different fitness compare. It is not what one more MET would do for you, and none of these studies tested that.

What makes the finding hard to dismiss is its consistency: large, different populations, different decades, the same direction and a similar size.

To place yourself on these scales, divide your VO₂max by 3.5 to get METs. A VO₂max of 42 is 12 METs. For a man in his 60s, the Cleveland Clinic elite threshold of 13.0 METs is 45.5 mL/kg/min.

Longstanding puts VO₂max in its battery because of these studies, but it reports your result as a percentile against the Cooper Institute norms, never as a mortality multiplier or a number of years. The app’s line runs all the way to 85 because the association held in people in their 70s and 80s, and because VO₂max falls faster with age, by more than 20% per decade from the 70s in the same people.

You can see where a value sits for your age in the calculator. This is not medical advice, and a percentile is not a prognosis.