In partnership with
Create work your clients can rely on
Your client doesn't care how their site was built. They care that it launches on time, runs on their own domain, and keeps working long after you hand it over.
With Lovable, you describe what your client needs and get a working version you can show them in the first meeting. Then you take it all the way to a live site on your client's domain, built to be found on Google. Lovable's project monitoring feature spots bugs and suggests fixes, so your name stays on work you're proud of.
Start today and take $5 off your first month.
Discount of $5 applies to first month of any paid plan. You need to sign up via this webpage in order to claim the discount. The discount applies to new and returning users without an active paid plan. Terms apply.
LONGEVITY LATEST ISSUE 35 · 4 NOVEMBER 2026
LONGEVITY LATEST · DEEP DIVE
The Shrinking Engine
Why VO2 max falls with age, where the ceiling sits, and what a two-year Dallas trial says about changing it in middle age.
By Christian Thomsen · Companion to Issue 35 · 4 November 2026 · ~7-minute read
I assumed exercise would flatten the curve. Stay active, and your fitness holds more or less steady until something breaks. A study that retested the same 810 adults for a median of almost eight years suggests otherwise.
In that study, part of the Baltimore Longitudinal Study of Aging, Jerome Fleg and colleagues measured the peak oxygen uptake of healthy adults aged 21 to 87 every few years. In people’s 20s and 30s it fell by 3–6% a decade; after 70, by more than 20%. The speeding-up showed in every band of self-reported activity, and the more active started each decade higher.
That’s an observational finding, with activity measured by questionnaire, so it can’t tell us whether training changes one person’s slope. What it does suggest is a more useful goal than stopping the decline: starting it from higher up.
Where the ceiling sits
Follow one breath. Oxygen crosses from the lungs into the blood, rides on haemoglobin to the working muscles, passes into the muscle cells and is used by the mitochondria. VO2 max is the most that whole chain can handle at full effort. Physiologists write it as delivery multiplied by extraction: how much blood the heart pumps each minute, times how much oxygen the muscles pull out of it.
In healthy people the weak link is usually delivery. David Bassett and Edward Howley’s classic review sets out three lines of evidence. Change oxygen delivery, by blood doping, thin air or beta-blockers, and VO2 max moves with it. Training raises VO2 max mainly by raising the heart’s maximum output, not extraction. And a small muscle given more blood than usual can use far more oxygen than it does during whole-body exercise. The muscles have spare capacity. The pump doesn’t.
The heart’s output is heart rate times stroke volume, and the two age differently:
Maximum heart rate falls by roughly 0.7 beats a year (208 minus 0.7 × age, on average), and training barely changes that.
Stroke volume, the blood pushed out per beat, depends on how well the left ventricle fills. Years of sitting stiffen it, and a stiffer chamber holds less.
Extraction tends to slip too, as muscle and its blood supply shrink. It matters, but it’s the smaller lever.
Maximum heart rate falls whatever you do. Stroke volume is the part you can still train.
A stiff heart can loosen
The best evidence that stroke volume can still be trained in middle age comes from Dallas. In a trial published in 2018, Erin Howden, Benjamin Levine and colleagues randomised 61 healthy but sedentary adults aged 45 to 64, average 53, to two years of training or an attention-control programme.
Each person in the training group had an individual plan that built up through the first year, peaking at two interval sessions a week, then eased in the second year into a maintenance pattern: four or five sessions a week of at least 30 minutes, including one interval session, plus strength work. Early sessions were supervised. After that, most training was done on their own, with heart-rate monitors, activity logs and regular consultations with the team. The controls did yoga, balance and strength sessions three times a week, so this was exercise against exercise, not exercise against the sofa. Adherence was 88%, and 53 people finished.
VO2 max in the training group rose 18%, from 29.0 to 34.4 ml/kg/min. The controls slipped from 29.5 to 28.7. Using right-heart catheterisation and 3D echocardiography, the team found the left ventricle had become measurably less stiff: it filled more at the same pressure, giving a bigger stroke. In the controls, stiffness didn’t change.
What Dallas shows, and what it doesn’t
It shows that in healthy, sedentary people in late middle age, two years of progressive, closely supported endurance training can raise VO2 max by about five points and loosen a stiffening heart, beyond what a gentler exercise programme achieves. It also shows people in their 50s can stick with a demanding plan for two years.
What it can’t tell us:
Whether it prevents heart failure. Stiffness is a risk marker, not an outcome, and 61 people is far too few to count events.
Which ingredient mattered. Intervals, total volume and strength work were bundled together.
Whether it works later. The authors argue late middle age is the window while the heart is still adaptable, but they didn’t test older people.
What happens without the support, or after stopping. Participants trained largely on their own, but with individual plans, monitoring and regular check-ins. Nobody tested the programme without that scaffolding, or followed people after it ended.
How much bias crept in. Participants knew their group, and the analysis covered completers only, though the assessors were blinded.
Still, five points of VO2 max is about one and a half METs, on a scale where large cohorts link each MET to roughly 11–17% lower mortality. That’s a link between groups of people, not a promise to any one of them.
Howden et al. Circulation (2018) · PEDro. Trial quality rating (2018) · Lang et al. Br J Sports Med (2024)
Does raising it buy time?
This is the uncomfortable part. Almost all the evidence linking VO2 max to lifespan is observational. A 2024 umbrella review covered 199 cohort studies and more than 20.9 million observations: each extra MET of fitness was linked to 11–17% lower all-cause mortality, and high versus low fitness to roughly half the risk. Among 750,302 US veterans, the least fit fifth had about four times the death rate of the fittest. But fit people differ in genes, illness and habits, and one treadmill test can’t untangle all of that.
Randomised evidence is thinner and messier. Generation 100 gave 1,567 Norwegians in their seventies five years of supervised training or national activity advice and found no overall difference in deaths: 4.5% against 4.7%. The catch is that the control group stayed active and kept its fitness too, so the trial ended up comparing active with active.
Then, last year, CHALLENGE. 889 people who’d had surgery and chemotherapy for colon cancer were randomised to a three-year coached exercise programme or health-education material, and followed for a median of almost eight years. The primary outcome, disease-free survival, improved: a 28% lower risk of recurrence, new cancer or death. Eight-year overall survival was 90% against 83%, a secondary result the authors describe as consistent with a survival benefit. It’s one cancer, in survivors, with an exercise programme rather than VO2 max as the intervention. Within that setting, it’s direct randomised evidence that an exercise programme can improve survival.
The UK Biobank analysis in this week’s issue fills in the middle: when people’s estimated fitness rose between two tests about four years apart, their later cardiovascular risk was lower, and when it fell, risk was higher. Associations again, but about change rather than a single snapshot.
None of this proves that raising VO2 max extends life in healthy people. But the direction of evidence is consistent.
Lang et al. Br J Sports Med (2024) · Kokkinos et al. J Am Coll Cardiol (2022) · Stensvold et al. BMJ (2020) · Courneya et al. N Engl J Med (2025) · López-Bueno et al. Eur J Prev Cardiol (2026)
The frontier
Two threads worth watching. First, measurement. Watches now put a VO2 max estimate on millions of wrists. The 2022 INTERLIVE meta-analysis found exercise-based estimates reasonable for populations but too loose for individuals, and two Apple Watch studies, in 2025 and on the Series 10 in 2026, found it underestimated by about 6 ml/kg/min. The 2026 authors note that nobody has yet independently tested whether repeated readings are reliable enough to monitor change. If that holds up, researchers could follow fitness cheaply in very large groups.
Second, the late-life heart. Whether the Dallas result holds after 65 is still open. A small Boston trial registered to start this year (NCT06247774) will look at the protein changes behind VO2 max gains in older adults at risk of heart failure. Early days, and only 42 people planned. Right question, though.
Molina-Garcia et al. Sports Med (2022) · Lambe et al. PLOS ONE (2025) · Lambe et al. Mayo Clin Proc Digit Health (2026) · ClinicalTrials.gov. NCT06247774
What this means for you
If I were designing a plan from this evidence, it would be boringly simple, and it’s a blend of trials rather than a tested programme. One hard interval session a week, the 4×4. Two or three easy sessions you could chat through. Two strength sessions. It’s close to the Dallas maintenance year, minus the build-up phase, the professional support and the catheters. If you want a reliable number, a lab test before you start and again a few months in is the way to get it; a single watch reading is too loose, and its ability to track change hasn’t been tested.
The training needs little more than a heart-rate strap. The measuring is where the money goes, and it’s optional.
The ceiling tends to fall with age. Training changes where it falls from.
Longevity Latest grades the evidence on interventions like these every Wednesday: what works, what’s premature, and what’s just marketing. The weekly issue carries the full grades and the practical plan.
Sources and further reading
Evidence reviewed through 31 October 2026. Fleg’s figures are longitudinal peak VO2 from the Baltimore study, measured between 1978 and 1998, with activity self-reported. The delivery-versus-extraction argument applies to healthy people exercising at sea level; in some diseases the limit sits elsewhere. Howden’s training was individually prescribed, supervised at first and then largely self-directed with monitoring and consultations; stiffness was measured by right-heart catheterisation; 53 of 61 completed and the analysis was not intention-to-treat. Cohort figures are associations. Generation 100 controls kept their fitness. CHALLENGE studied colon cancer survivors; overall survival was a secondary outcome. UK Biobank fitness values were estimated from submaximal tests. Both Apple Watch studies were cross-sectional. Educational content only; check with your GP before starting high-intensity training if you have heart disease or symptoms.