VO2 Max: The Strongest Single Predictor We Have

If you could track exactly one number about your body over the next thirty years, the evidence says it should probably be cardiorespiratory fitness.

Not weight. Not blood pressure. Not cholesterol. Those all matter, and two of them appear elsewhere on this site. But none of them produces effect sizes like the ones below.

The study

In 2018, a Cleveland Clinic team published an analysis of 122,007 consecutive adults referred for symptom-limited exercise treadmill testing between 1991 and 2014, followed for a median of 8.4 years — about 1.1 million person-years (Mandsager et al., 2018).

Fitness was quantified as peak estimated metabolic equivalents, or METs, achieved on the treadmill, using sex-specific formulas. Participants were sorted into five bands by age- and sex-matched percentile (full text):

• Low: below the 25th percentile
• Below average: 25th–49th
• Above average: 50th–74th
• High: 75th–97.6th
• Elite: at or above the 97.7th percentile, meaning two standard deviations above the mean for their age and sex

The adjusted hazard ratio for all-cause mortality, elite versus low, was 0.20 (95% CI 0.16–0.24). An 80% lower risk of death.

The comparison that reframes everything

Effect sizes are hard to feel in the abstract. The same paper put low fitness alongside conditions that get far more clinical attention. Adjusted hazard ratios for all-cause mortality:

Risk factor	Hazard ratio
Coronary artery disease	1.29
Smoking	1.41
Diabetes	1.40
Low fitness (vs elite)	5.04

Being in the bottom quartile of fitness was associated with a larger mortality hazard than being a smoker. Than having diabetes. Than having coronary artery disease.

No physician would look at a smoking habit and say nothing. Almost none measure fitness at all.

Even the modest steps mattered. Below average versus above average — one band, a difference most people could close in a year — carried a hazard ratio of 1.41 (95% CI 1.34–1.49). That is roughly the mortality hazard of smoking, moved by one fitness band.

No ceiling

The finding that surprised the field most: there was no upper limit of benefit.

Elite versus high — the top 2.3% against the next 22% — still favored elite (HR 0.77, 95% CI 0.63–0.95). The authors concluded that cardiorespiratory fitness is inversely associated with long-term mortality with no observed upper limit of benefit, and that extremely high fitness was associated with the greatest survival, including in older patients and those with hypertension.

This matters because a widespread belief holds that extreme endurance training is harmful past some threshold. In this cohort of 122,007 people, that threshold did not appear.

Two honest caveats

This is observational. Everyone in the cohort was referred for a treadmill test, meaning they had a clinical reason. People who can walk fast on a treadmill are healthier in ways beyond fitness. The authors adjusted extensively, but adjustment is not randomization. Some of that hazard ratio is fitness reflecting health rather than producing it.

Elite is not a realistic target for most people. Two standard deviations above the age- and sex-adjusted mean is a serious athlete. The useful finding is not “become elite.” It is the shape of the curve at the bottom, where the largest gains sit.

Can you actually change it?

This is where fitness separates from most risk markers. A meta-analysis of 37 studies covering 334 participants in 40 training groups, published between 1965 and 2012, examined VO2 max trainability with high-intensity interval training (Bacon et al., 2013).

Mean improvement with HIIT was 0.51 L·min⁻¹ (95% CI 0.43–0.60), a standardized effect of 0.86 SDs (95% CI 0.72–0.99). A subset of nine studies using longer intervals of three to five minutes produced larger gains, roughly 0.8–0.9 L·min⁻¹. For reference, the authors note that traditional continuous endurance training in the Heritage Study produced mean increases of about 0.4 L·min⁻¹.

The nine highest-performing studies shared three features: longer programs (9.7 ± 1.8 weeks versus 6.9 ± 1.4), greater weekly volume (209 ± 90 minutes versus 123 ± 67), and longer intervals.

The authors made a further claim worth noting, since it contradicts a common belief: they challenged the assumption that some individuals are low responders, reporting marked responses in all subjects when optimized protocols with longer intervals were used. That is one meta-analysis, not settled fact, but it argues against writing yourself off as a non-responder before you have tried a properly structured program.

What to do with this

• Get a number. A proper VO2 max test on a metabolic cart is best. A symptom-limited treadmill test through a cardiologist gives you estimated METs, which is exactly what the Mandsager cohort used. Even a validated field test — a 12-minute run, a submaximal step test, or your watch’s estimate used consistently over time — beats no measurement.
• Find your percentile, not your absolute value. The study’s bands were age- and sex-adjusted. A 42 mL/kg/min means something very different at 30 than at 65.
• If you are below average, that is the highest-value information in this article. The largest mortality differences in the data sit between the bottom band and the middle. You do not need to become an athlete; you need to leave the bottom quartile.
• Build volume before intensity. The trainability data favored longer programs with more weekly minutes and three-to-five-minute intervals. The best-performing protocols averaged over 200 minutes per week — that is total training, not interval time.
• Retest annually. Fitness declines with age by default. A stable number over a decade means you are gaining ground relative to your age band.
• Do not treat this as the only number. Fitness in this cohort is partly a marker of underlying health. Train it, and also manage the things it cannot fix — lipids, glucose, blood pressure, sleep.

The reason to prioritize this measurement over almost any other is not that fitness is magic. It is that fitness is the rare risk factor that is both strongly associated with the outcome you care about and demonstrably responsive to something you can start doing this week.