VO2 Max, Zone 2, and the Evidence on Dying Later

If you rank the things that predict when a person will die, aerobic fitness sits near the top — above smoking, above diabetes, above a diagnosis of coronary artery disease. That finding is real, replicated, and larger than most people expect.

It is also more slippery than the usual summary suggests. The evidence that fit people live longer is overwhelming. The evidence that getting fitter makes you live longer is thinner, and the one large trial that randomised people to years of supervised training did not find a significant mortality difference. Both things are true at once, and holding them together is the only honest way to read this literature.

The scale of the association

Start with the numbers, because they are genuinely striking.

A 2009 meta-analysis in JAMA pooled 33 cohort studies covering 102,980 people. Each additional MET of aerobic capacity — roughly the difference made by jogging 0.6 mph faster — was associated with 13% lower all-cause mortality. A MET is a unit of oxygen consumption, with one MET being roughly what you use sitting still.

A 2018 cohort study of 122,007 patients who had treadmill tests at the Cleveland Clinic found that patients in the top 2.3% of fitness for their age and sex had one-fifth the mortality risk of the least fit group (hazard ratio 0.20, 95% CI 0.16-0.24). For scale, in the same cohort and the same statistical model, smoking carried a hazard ratio of 1.41, diabetes 1.40, and coronary artery disease 1.29 — while being in the lowest fitness group carried 5.04.

The largest analysis, a 2022 study of 750,302 US veterans, reached the same conclusion across ages from 30 to 95, in men and women, and in every racial group examined. Its authors put it bluntly: being unfit carried greater risk than any traditional cardiac risk factor they measured.

Two details that change the interpretation

There is no observed ceiling. Both large cohorts looked specifically for a point where more fitness stops helping or starts hurting. Neither found one. The veterans study located its lowest risk around 14 METs and saw no penalty beyond it.

The steepest part of the curve is at the bottom. This is the finding most people miss. According to the American Heart Association’s 2016 scientific statement, more than half of the total reduction in all-cause mortality occurs between the least fit group (under 5 METs) and the merely next-to-least fit (5-7 METs). Moving off the floor matters far more than approaching the ceiling. If you are sedentary, the most valuable training you will ever do is the first few months of it.

What that means in years

Hazard ratios are easy to over-read. The veterans study also reported average age at death, which is more concrete:

Men around 10-12 METs died 4.5 years later on average than men in the bottom 20% of fitness (77.7 vs 73.2 years).
The most fit men died 6.0 years later (79.2 vs 73.2).
Among people already over 70, reaching about 7 METs was associated with roughly 2.7 more years (87.3 vs 84.6).

Read these as differences between groups of people in an observational study, not as a promise of what a training programme will buy you. Fit people differ from unfit people in many ways besides fitness.

The part usually left out

Almost everything above is observational. The people were not randomised to be fit; researchers measured fitness and waited. That design cannot separate three very different explanations:

Fitness protects against death.
Early, undiagnosed illness lowers fitness — so low fitness is a symptom rather than a cause. This is called reverse causality.
Something else, from genetics to income to healthcare access, drives both.

Serious researchers in this field say so explicitly. The veterans study states that its design “does not demonstrate causation.” Its authors did try to address reverse causality by excluding people diagnosed with cardiovascular disease or cancer in the year before testing and those who died in the first two years, and the association held — which is reassuring, but not the same as a trial.

The trial that did happen

In 2020, the Generation 100 study published five-year results in the BMJ. It randomised 1,567 Norwegians aged 70-77 to supervised high-intensity interval training twice a week, supervised moderate continuous training twice a week, or a control group told to follow national activity guidelines. Mortality was the primary outcome.

There was no statistically significant difference in all-cause mortality. The interval training group had the lowest death rate (3.0% vs 4.7% in controls, hazard ratio 0.63) but the confidence interval ran from 0.33 to 1.20 — comfortably including no effect. The moderate training group did marginally worse than controls, also not significantly.

This is the strongest test we have, and it is null. Anyone citing the observational fitness literature as proof that training extends life needs to explain this trial, not ignore it.

There are reasonable explanations, and they matter:

The control group exercised. They were told to follow national guidelines, and they did — in fact they chose to do more of their activity at high intensity than the moderate-training group did. The achieved control dose ended up between the two intervention arms. The trial largely tested intensity distribution among active older adults, not exercise against inactivity.
The participants were already healthy and active. About 80% reported medium or high activity at baseline, and 87.5% reported good health. There was little room to move.
It was underpowered. Seventy-two deaths across five years cannot detect modest differences.

But the possibility that the observational effect is simply inflated cannot be dismissed either. The most defensible summary: aerobic fitness is among the strongest known predictors of mortality, and the causal size of the effect remains genuinely uncertain.

One thing the observational data does support more directly is that fitness is not fixed. A 1995 study following 9,777 men across two exams found that men who moved from unfit to fit had 44% lower mortality (95% CI 25%-59%) than men who stayed unfit. That is still observational — men who improved may have differed in other ways, and improvement could reflect recovery from unrecognised illness — but it shifts the question from who you are to what you do.

Now, about “Zone 2”

Here the ground gets unstable, and not for the reason most people assume.

“Zone 2” means two opposite things depending on who is talking.

In the three-zone model used across the endurance research literature — as laid out in Stöggl and Sperlich’s 2015 review — the zones are:

Zone 1: easy work below the first ventilatory threshold, blood lactate at or under 2 mM.
Zone 2: threshold work, at or near the lactate threshold, around 4 mM.
Zone 3: hard work above the second threshold.

In popular longevity discussion, “Zone 2” almost always means the easy, conversational, could-do-this-for-an-hour intensity — which is zone 1 in that research model, not zone 2.

This is not pedantry. Polarized training, the approach most often invoked to justify easy aerobic work, puts roughly 75-80% of volume in zone 1 and 15-20% in zone 3, while deliberately keeping zone 2 under 10%. In the research model, zone 2 is the zone you avoid. Someone who reads a polarized training paper as an endorsement of popular “Zone 2” has inverted the recommendation.

For the rest of this article: easy aerobic work means below the first lactate threshold — what longevity writers call Zone 2 and researchers call zone 1.

What the training research actually found

A 2024 systematic review and meta-analysis pooled 17 studies of 437 athletes comparing polarized training with other intensity distributions. Polarized training was better for VO2peak, but the effect was small (standardised mean difference 0.24, 95% CI 0.01-0.48) and appeared only in interventions shorter than 12 weeks and only in highly trained athletes. For time trial performance, time to exhaustion, and power at threshold, there was no difference at all.

The mechanistic story — that easy volume drives calcium signalling while hard intervals drive AMPK, together maximising mitochondrial growth — is repeated far more confidently than the data warrant. A 2013 crossover trial in trained cyclists found meaningfully better performance gains from polarized training (peak power +8% vs +3%, lactate threshold +9% vs +2%) while measuring no change in mitochondrial enzyme activities after either programme. The performance moved; the proposed mechanism did not.

The gap nobody should paper over

No study has linked any training intensity distribution to all-cause mortality.

The argument for easy aerobic training as a longevity intervention is a chain:

easy aerobic volume → higher VO2 max → lower mortality

The first link rests on small, short trials in trained athletes — mostly male, often a dozen participants — measuring performance, not health. The second link is a strong association whose randomised confirmation is limited and, in Generation 100, absent. Neither link is worthless. Neither is proof. Any confident claim that a specific zone extends life is running well ahead of the evidence.

What the evidence does support is less exciting and more useful: total volume of aerobic activity is reliably associated with lower mortality, and easy intensities let you accumulate more of it without accumulating fatigue. Pooled data from 661,137 people found 20% lower mortality below the guideline minimum of 150 weekly minutes, 31% at one to two times the minimum, and a plateau around 39% at three to five times — with no evidence of harm even at ten times the minimum.

Notice where that curve flattens. Most of the available benefit arrives by the standard guideline. The returns past it are real but modest.

What follows from all this

Given evidence of this shape — strong association, uncertain causation, a terminology mess, and no mortality data on zones — a few things hold up:

Know roughly where you stand. The mortality gradient is steepest at the bottom, so the difference between under 5 METs and 7 METs matters more than anything you will read about elite athletes. Estimates from a treadmill test, a submaximal field test, or even a decent wearable are enough to tell you which end you are on. The AHA argues fitness deserves treatment as a clinical vital sign.

If you are sedentary, the specific protocol is close to irrelevant. Any sustainable increase in aerobic activity moves you along the steepest part of the curve. Zone debates are an optimisation problem for people who have already solved the volume problem.

Accumulate easy aerobic volume, mostly because it is sustainable. Every distribution studied in trained athletes puts the majority of volume at low intensity. That is a robust pattern, even though the reason is more plausibly “you can do more of it” than any zone-specific cellular magic.

Do not skip higher intensities. VO2 max responds to work near its ceiling, and in Generation 100 the interval group had the numerically lowest mortality even though the difference was not significant. The evidence does not support easy work alone.

Treat the total as the target. The guideline minimum — 150 minutes of moderate or 75 minutes of vigorous activity weekly — captures most of the measurable association. More is better up to a point, and there is no signal of harm from a lot more.

The genuinely defensible claim is narrower than the one usually made, and still worth acting on: aerobic fitness is among the most powerful predictors of how long people live, it is substantially modifiable, and the returns are largest for those with the least. Whether it is precisely 4.5 years, and whether any particular zone is responsible, remains unsettled.