Does lifelong hard training stop VO2max falling?
Not for a marathon world-record holder: 68.1 at 49 → 43.6 at 77 (−36%) on 10-15 h/wk; he stayed high by starting high. Enlarged atria, then AF (link unproven). n = 1.
Foulkes et al. report three maximal exercise tests across 50 years in Derek Clayton, the first man under 2:10 for the marathon and world-record holder until 1981, who reported up to 150-300 miles a week when racing (his July 1965 diary averages 175.2 miles a week) and 10-15 h a week of moderate-to-vigorous running, then cycling, after retiring in 1974.
VO2max barely changed from 27 to 49 years (5.09 → 5.13 L/min; 69.7 → 68.1 mL/kg/min) although HRmax fell 188 → 176. By 77 it had fallen to 3.26 L/min, 43.6 mL/kg/min (−36%), with HRmax 154 and oxygen pulse 21.2 mL/beat (−27% from 49).
That value is still above the 90th percentile (26.2) for men aged 70-79 in the FRIEND registry, equal to the median of healthy men aged 20-29, and 2.4 times the 18.0 mL/kg/min threshold for independent living.
Cardiac MRI at 77 showed large, symmetrically dilated ventricles (LV end-diastolic volume 272-273 mL; RV 337 mL, above the 95th percentile even for endurance athletes), proportional LV mass (174 g), no scar, low-normal resting function (echo LVEF 53%, GLS −14%) and a resting heart rate of 38; in supine exercise stroke volume reached 200 mL and cardiac output 22.2 L/min at a heart rate of only 111.
Both atria were severely enlarged (indexed volumes 93 and 89 mL/m², severe above 48) and left-atrial strain was 10% (reservoir) and 0% (contractile) despite sinus rhythm.
Six months after testing he presented with palpitations, fatigue and loss of exercise capacity and was in atrial fibrillation; he needed cardioversion, recurred within 4 months, had pulmonary vein isolation (June 2023), recurred after 4 weeks, had a second ablation including the posterior left atrial wall (September 2023) and was discharged on flecainide with advice to stop alcohol.
The authors judge the net balance favourable for health span; they also say the atrial remodelling "provides an example of the potential cost of lifelong endurance exercise", while age and sex alone could explain his AF.
Protocol
Design: longitudinal case report. Tests in 1970 (age 27, Ball State University, with David Costill), 1991 (age 49, Victoria University, Melbourne) and February 2020 (age 77, Baker Heart and Diabetes Institute, within a broader study of elite endurance athletes); ethics approval from the Alfred Human Research Ethics Committee for the 2020 study; consent to be named.
Who: man; 1967 world record 2:09:36.4 (Fukuoka), 1969 2:08:33.6 (Antwerp); Olympian 1968 and 1972; retired from competitive running in 1974. Ages 40-50: 10-15 h/week running; 50-60: running and cycling 10-15 h/week; from 70: cycling 10-15 h/week at moderate-to-vigorous intensity. No medications before or at the 2020 tests.
VO2max: treadmill to exhaustion with Douglas bags (1970, 1991); electronically braked cycle ergometer with breath-by-breath gas analysis (2020; highest 30-s rolling average). Maximal effort: plateau of at least 30 s; RER 1.21 and HRmax 101% of predicted in 2020. Age-predicted HRmax by "207 – 0.7 × age", cited to Tanaka.
Heart (2020 only): resting 3 T cardiac MRI (biventricular volumes, ejection fractions, LV mass, gadolinium scar imaging); resting echocardiography (LV global longitudinal strain, diastolic function, atrial volumes and strain); exercise cardiac MRI during progressive supine cycling to peak (LV volumes, stroke volume, cardiac output); whole-body DXA. Electrophysiological study and high-density left atrial voltage mapping during ablation (2023).
Comparators: FRIEND registry cycle-ergometer norms; UK Biobank values for men 65-74; published reference ranges for elite endurance athletes; one young national-level endurance athlete (23 y, VO2max 5.14 L/min, matched to DC at 27) and one young healthy control (23 y, 3.21 L/min, matched to DC at 77) for the supine exercise MRI.
Funding: NIA grant AG-10576 (USA) and NHMRC project grant APP1130353 (Australia). Conflicts: P.M. Kistler has received consultancy and speaking funding from Abbott Medical and served on an advisory board with fellowship support from Biosense Webster (both make ablation equipment); no other conflicts. Costill, a co-author, is pictured with DC at the 1970 evaluation, and the 1970 test is his published work.
Key Results
| Measure | 1970, 27 y | 1991, 49 y | 2020, 77 y | 77 vs 49 |
|---|---|---|---|---|
| VO2max, mL/kg/min | 69.7 | 68.1 | 43.6 | −36% |
| VO2max, L/min | 5.09 | 5.13 | 3.26 | −36% |
| HRmax, beats/min | 188 | 176 | 154 | −13% |
| Max oxygen pulse, mL/beat | 27.1 | 29.2 | 21.2 | −27% |
| Body mass, kg | 73.0 | 75.3 | 72.9 | −3% |
| Haemoglobin, g/dL | 14.4 | 14.8 | 14.0 | −5% |
| Resting blood pressure, mmHg | not measured | 116/69 | 134/67 | systolic +16% |
| Heart and body at 77 | DC | Reference (healthy men unless stated) |
|---|---|---|
| LV end-diastolic volume (MRI) | 272 mL (text: 273), indexed 139 mL/m² | 109-218 mL; UK Biobank 95th percentile 110 mL/m² |
| RV end-diastolic volume | 337 mL, indexed 172 mL/m² | 124-248 mL; above the 95th percentile for endurance athletes (301 mL) |
| LV mass; mass/volume ratio | 174 g; 0.64 g/mL | 64-141 g; 0.42-0.84 (eccentric, proportional) |
| LVEF echo / MRI; RVEF | 53% / 57%; 47% | 53-72% / 48-69%; 45-65% |
| LV global longitudinal strain | −14% | −17 to −24% (reduced) |
| Left / right atrial volume index | 93 / 89 mL/m² | below 36; severe dilation above 48 |
| Left atrial reservoir / contractile strain | 10% / 0% (in sinus rhythm) | 23-68% / 6-28% |
| Myocardial scar (gadolinium) | none | — |
| Resting heart rate | 38 beats/min | — |
| Peak supine exercise | HR 111, stroke volume 200 mL, cardiac output 22.2 L/min (11.4 L/min/m²), LVEF +16% | young control 159 beats/min and 128 mL; young athlete's output 3.4 L/min higher |
| Peak upright systolic pressure | 245 mmHg | authors: proportionate to his high power output |
| Appendicular lean mass; body fat | 25.0 kg (7.46 kg/m²); 13.9% | above sarcopenia cut-offs; body fat at the 3rd percentile for age |
| Clinical course after the 2020 tests | Event |
|---|---|
| ~6 months later | palpitations, fatigue, could not keep up on group rides; atrial fibrillation, ventricular rate 62 |
| then | anticoagulation, cardioversion, symptoms resolved |
| 4 months later | symptomatic AF again |
| 1 June 2023 | pulmonary vein isolation; atrial voltages preserved despite severe enlargement |
| 4 weeks later | AF recurrence; cardioversion |
| 21 September 2023 | re-isolation of the left pulmonary veins plus posterior left atrial wall isolation; discharged on flecainide 50 mg daily, advised to abstain from alcohol |
The decline is age-typical, by the authors' own reading. "DC's superior V̇o2max at the age of 77 is not dictated by slowing the age-related declines, but by the development of a large V̇o2max reserve in early-to-midlife".Over 49 → 77 the fall was 1.3% a year; the authors call it "comparable to the expected rate of 15–25% decrease per decade" in people over 50 (by simple division 36% over 28 years is about 13% a decade, our calculation).He reported 10-15 h a week at ages 40-60 and from 70 (ages 60-70 are not itemised), far below his racing volume; the authors note that this reduction after retirement "potentially contributed" to the decline.
At 77 still far above his age group: above the 90th percentile for men 70-79 (26.2 mL/kg/min) and equal to the median for healthy men 20-29.
The cost: severe bi-atrial enlargement with absent left-atrial contraction, then AF needing two ablation procedures. The paper cites a 2.5- to 14-fold higher AF prevalence in middle-aged and older endurance athletes than in non-athletes.
Mechanisms
Measured: his cardiac output at 77 came from size, not rate: a large LV end-diastolic volume meant a peak supine stroke volume of 200 mL (72 mL above a young control of similar VO2max) at a heart rate of 111. The authors calculate he would exceed 15 L/min with a rise in heart rate alone.
Measured: the 49 → 77 decline came with lower maximal heart rate (−22 beats/min) and lower oxygen pulse (−8 mL/beat); the authors infer reductions in stroke volume and/or oxygen extraction as well, from an estimated upright output 11% below that estimated in 1970.
Measured: atrial remodelling was extreme (left atrial index 93 mL/m², contractile strain 0%), while voltage mapping at ablation showed preserved atrial tissue voltage despite the enlargement.
Authors' interpretation: the enlarged, poorly contracting atria "could be consistent with" atrial myopathy, a substrate for AF; in non-athletes such atria predict AF, but the same association "has not been identified in athletes", and his AF may instead come from age and sex. The arrhythmia is "not definitively linked as a root cause in his case".
Authors' interpretation: the net benefits (functional reserve, body composition, expected longevity) outweigh the cost in this man; whether that holds for most athletes training at such levels "requires further evaluation".
Authors' interpretation: his high VO2max and remodelling more likely reflect "the combined effects of lifelong exercise and a favorable genetic profile" than training alone.
Inference: reduced resting LVEF and GLS overlap with cardiomyopathy ranges in big athlete's hearts; here exercise MRI showed normal augmentation, which is how the authors separate adaptation from disease. Resting numbers alone would have been ambiguous.
Practical Application
Does lifelong hard training stop VO2max falling? Not for a marathon world-record holder: 68.1 at 49 → 43.6 at 77 (−36%) on 10-15 h/wk; he stayed high by starting high. Enlarged atria, then AF (link unproven). n = 1.
AF in an older endurance athlete is a real cost to name, not a myth (one case plus the prevalence data it cites): palpitations, unexplained fatigue or a sudden loss of exercise capacity in a veteran endurance athlete are for a doctor and an ECG, not a lighter week. Here the first sign was being dropped on a group ride.
Do not read athlete's-heart imaging alone (the base's reading of this case): low-normal resting ejection fraction and strain sat beside excellent exercise reserve; interpreting an older athlete's echo is a sports-cardiology question.
No dose rule follows. One man at an extreme training volume; this case cannot say how much training raises AF risk, or whether less would have avoided it. The base holds no trial or cohort on AF dose in athletes (gap).
Limitations
One man; three VO2max points 21 and 29 years apart, so the shape of the decline between them is unknown; cardiac imaging only at 77.
Test mode changed from treadmill (1970, 1991) to cycle ergometer (2020); the authors judge the effect "likely negligible" because he was a cyclist by then, but part of the 36% fall could be mode.
Training history is self-reported apart from the 1965 diary; no intensity detail across life.
Supine exercise MRI underestimates upright heart rate and output; the comparisons with the two young men are single individuals.
The causal link between his remodelling and his AF is unproven, as the authors say; age and sex are AF risk factors on their own.
Internal inconsistencies: the Results give the 49 → 77 decline as "1.3% per year (0.57 mL/kg/min)", but 24.5 mL/kg/min over 28 years is about 0.88 per year (our calculation); LV end-diastolic volume is 272 mL in Table 3 and 273 mL in the text; the prediction formula is printed as "207 – 0.7 × age" while Tanaka's published equation is 208 − 0.7 × age (REF-1082).
Conflicts: one author (Kistler) has consultancy and advisory ties to Abbott Medical and Biosense Webster; a co-author (Costill) is pictured with him at the 1970 evaluation; a decades-long research relationship with the subject is evident from the paper but is not discussed as a source of bias.
Funding: NIA and NHMRC grants.
How this entry was checked and changed
Why this level
single case study (the rubric's row for n = 1), read from the full text (PMC, author-accessible version). Three VO2max points over 50 years with a change of test mode (treadmill with Douglas bags in 1970 and 1991, cycle ergometer with breath-by-breath gas analysis in 2020); cardiac imaging exists only at 77, so remodelling cannot be dated
Full text (checked 2026-10-22)
DC's V̇o2max barely changed between 27 yr (69.7 mL/kg/min) and 49 yr (68.1 mL/kg/min), then fell by 36% by 77 yr (43.6 mL/kg/min) (full text).
"he reported completing up to 150–300 miles per week in the lead-up to his world record-breaking marathons"; Table 1 average 175.2 miles/wk for July 1965; from the age of 70 yr cycling "for 10–15 h/wk" (full text).
1970 (5.09 L/min; 69.7 mL/kg/min) to 1991 (5.13 L/min; 68.1 mL/kg/min); maximal HR from 188 b/min to 176 b/min; at 77 yr (3.26 L/min; 43.6 mL/kg/min), "a 1.3% per year (0.57 mL/kg/min) decline"; peak heart rate −22 beats/min and oxygen pulse −8 mL/beat, 27% reduction (full text).
V̇o2max at 77 "is equivalent to that predicted for a healthy 20- to 29-yr-old male"; "2.4 times the value considered necessary for maintaining functional independence [18.0 mL/kg/min"; above the 90th percentile (26.2 mL/kg/min) for 70- to 79-yr-old males (full text).
"DC's superior V̇o2max at the age of 77 is not dictated by slowing the age-related declines, but by the development of a large V̇o2max reserve in early-to-midlife"; decline "comparable to the expected rate of 15–25% decrease per decade" (full text).
LV EDVi: 139 mL/m2; RV EDVi: 172 mL/m2; RV EDV = 337mL exceeds the athletes' upper reference (>95th percentile of 301 mL); resting HR: 38 beats/min; LVEF: 53%; GLS: −14%; LV mass (174 g); LV mass: LV EDV (0.64 g/mL) (full text).
LAVi = 93 mL/m2 and RAVi = 89 mL/m2, above the cut-offs of 48 mL/m2 for severe dilation; LASreservoir and LAScontractile of 10% and 0% (full text).
Peak supine exercise: HR 111 b/min, Qc of 22.2 L/min (indexed 11.4 L/min/m2); LV EDV (273 mL vs. 216 mL); peak SV 200 vs. 128 mL; supine HR 111 vs. 159 beats/min (full text).
"Six months after the comprehensive exercise testing and imaging, DC presented with palpitations and fatigue and was found to be in atrial fibrillation (AF) with a controlled ventricular response rate (62 bpm)" (full text).
Pulmonary vein isolation "On June 1, 2023"; cardioversion "4 wk later for AF recurrence", on September 21, 2023; discharged on flecainide 50 mg daily and advised to abstain from alcohol (full text).
"The prevalence of AF among middle-aged and older endurance athletes has been observed to be 2.5- to 14-fold higher than in non-athletic referent populations" (full text).
"it is also possible that DC's AF and atrial enlargement were not directly related to his previous exercise habits but were a consequence of other AF risk factors such as his age and sex" (full text).
Age-predicted maximal HR "using the Tanaka et al. formula (207 – 0.7 × age)"; grants AG-10576 and APP1130353; P.M.K. "has received funding from Abbott Medical" and support "from Biosense Webster" (full text).