Heat monitoring
Two sessions in TT position to locate the constraint and establish a testable hypothesis.
Reference longitudinal case · anonymised
VST put to the field test
From monitoring to racing, from racing to the next test
A longitudinal sequence documented over seven weeks in the follow-up of a WorldTour climber. Two heat monitoring sessions, a national time-trial championship, a reference test, eighteen days of altitude camp with eight instrumented sessions, then a WorldTour classic and a professional stage race. Every figure is rebuilt from the second-by-second files.
MEASURE→ANTICIPATE→STEER→REASSESS
Key message for the staff
VST turns an invisible ventilatory pattern into markers the staff can use.
In this rider, follow-up identified when ventilatory organisation became costly with duration, built individual targets, then checked whether they held in training and competition. This is the value of the VST intervention: making the constraint measurable, anticipatable and steerable without replacing the coach.
01 · IDENTIFYThe individual boundary and the context in which it emerges02 · STEERRf, Tv, ventilation and strategies according to the effort03 · VERIFYTransfer and robustness under real-world conditions
See exactly what was managed ↓What this sequence represents
A seven-week window in a follow-up that started well before it.
What this is
- Hypotheses stated and dated before the races.
- Their confrontation with training and race files.
- An intervention run during an altitude camp.
- A check across five consecutive days of racing.
What this is not
- Causal proof — n=1, no control condition.
- A promise of watts.
- A generalisation to every rider.
- A replacement for the coach, nor a medical reading.
Every claim carries its level of evidence. Mechanistic hypotheses are labelled as such — and so is one conclusion that was corrected along the way.
Reference test
Thresholds, upstream ventilatory function and entry into the altitude block.
Altitude camp
Eighteen days and eight instrumented sessions to confront, then steer, the pattern.
WorldTour classic
Observe what transfers when the rider’s attention is absorbed by racing.
Stage race
Assess tactical cost and day-to-day repeatability.
New monitoring
Turn race observations into the next hypothesis, test and decision.
Starting point
No shortage of available air. An individual boundary emerges with duration.
4.82 LFEV1
5.68 LFIV Full
1.18ICIF — not discriminant here
123 L/minVE at the top
55 L/minMargin
Predicted MVV = FEV1 4.82 × 35 = 168.7 L/min. The reserve moves from 56% at the first threshold to 26% at the top: no mechanical ceiling is reached. Looking only at the stage data would suggest there is nothing to address on the ventilatory side.
The expiratory lever
Time, not force.
With a PEmax of 200 cmH₂O and flows that still plateau, pushing harder will not move the air out faster. Te = (1 − 0.42) × 60 / Rf.
Rf 24> 1.4 sRf 321.10 sRf 400.87 s
The ≈32 boundary is an individual marker, not a norm.
Base observation
Above roughly 32 cycles/min, tidal volume breaks down; below it, tidal volume is stable and full-emptying manoeuvres work.
Over-under session
Only the under segment lengthens: 90 → 120 → 240 s. Terminal Rf 39.7 → 63.3; Tv 297 → 194 AU.
11 intervals at 317 W
Same power, duration and climb. Rf rises from 25 to 38; the switch occurs on the eighth interval.
Duration, not intensity
The constraint emerges with time spent above the boundary.
Six efforts of comparable power, decreasing then increasing in duration — 3′, 2′, 1′, 1′, 2′, 3′ — all above 400 W and therefore entirely on the oral route.
Ventilation barely moves. The pattern swings completely.
14%Spread in ventilation
32–41Ventilatory frequency
254–278 AUVolume per cycle
The degradation is not a loss of ventilated quantity: it is a shift along an iso-ventilation curve, towards a fast and shallow configuration.
The best pattern sits at the highest power — 459 W on the one-minute effort against 401 W on the three-minute. This observation does not support an explanation driven by intensity alone.
Longitudinal intervention
What was managed throughout the longitudinal follow-up.
Longitudinal VST did not consist of imposing a permanent ventilatory pattern or replacing the staff’s work. It consisted of identifying an individual boundary, building ventilatory targets adapted to the different intensity domains, then checking their adoption and robustness under the rider’s real-world conditions.
What was managed in practice
Individual targets
Target ventilatory frequency according to intensity and duration.
Ventilatory organisation
Tidal volume and ventilation associated with those targets.
Switch point
The moment at which the spontaneous pattern became more costly.
Stochastic efforts
Ventilatory strategies to use during and after those efforts.
Field transfer
The transition from the reference test to training and then competition.
Reassessment
The evolution of targets through successive monitoring blocks.
Feedback
Sharing with the rider and sharing useful observations with the staff.
The aim was not to systematically reduce ventilatory frequency, but to extend the time during which the rider could maintain a ventilatory organisation compatible with the required power and the constraints of racing.
Within-athlete comparison
The same effort, spontaneous then steered.
Two sessions two days apart on the same climb. Shared 30-minute window between 800 and 1,450 m, at 300 W, without a formal control condition.
Spontaneous day
Rf: 29 → 33 → 36 → 39
The ≈32 boundary is crossed in the second quarter.
Steered day
Rf: 26 → 27 → 28 → 29
The boundary is never crossed despite more load, heat and altitude across the full block.
Off-task change
What actually improved: the spontaneous pattern.
Three days after the synthesis report was shared, matched easy climbs at 150–260 W show median frequency moving from 39 to 27 cycles/min and time below the boundary moving from 20% to 72%.
Training content remained comparable between the two sessions. The main documented change was that the rider had been shown his own pattern. This temporal proximity is an interesting signal, but it does not allow the change to be attributed causally to the feedback alone.
Median Rf39 → 27−12 cycles/minTime controlled20 → 72%+52 points
Side effect
Ventilatory function degrades less across the camp.
Two altitude camps in the same season, one without a ventilatory sensor, one with a steered pattern. Daily morning spirometry over 19 days in both.
| No sensor | Steered pattern | |
|---|---|---|
| Regression slope | −44 mL/day | −9 mL/day |
| Variance explained (R²) | 79% | 24% |
| Lowest value reached | 4.08 L | 4.47 L |
| Gap to pre-camp baseline | −13.9% | −7.8% |
With the steered pattern, the slope is markedly reduced and day-to-day variability becomes proportionally more prominent.
Hypothesis stated before the race
A hypothesis formulated before the race becomes testable.
Expected at D−4
Hypothesis at D−4: the terminal constraint should emerge on the ventilatory side before the available metabolic power is exhausted. Targets: 360–365 W in normal conditions; 340–345 W in the heat with effective cooling. Identified risk: the climbs break the pattern.
Observed at D+0
38 min 43 s, averaging ≈338 W. Full pattern ≈11 min: Rf 42, Tv 2.29 L, 359 W. Final block: Rf 62, Tv 1.87 L, 329 W.
359 → 329 WPower
95 → 116 L/minVentilation
175 → 180 bpmHeart rate
≈11 minFull pattern
At a similar heart rate, the second half requires 21 L/min more ventilation for 30 W less power, including an estimated ≈5 L/min linked to the increase in dead-space ventilation. Core temperature: 37.8 → 39.7 °C.
Transfer in competition
The pattern is observed when attention is elsewhere.
In a 223 km WorldTour classic, control holds until km 170. Median Rf: 33 before the switch, 50 after; time below the boundary: 31% then 4%. On tempo climbs at 290–335 W, Rf remains between 26 and 31.
At matched watts: 270–300 W, Rf 27 in the race versus 27 at camp; 320–340 W, 27 versus 28.
Before / after33 → 50Median RfBelow boundary31 → 4%Time controlled
The stage race tests approach and repeatability.
Four climbs at matched power — 309, 309, 311 and 314 W. Everything differs except the watts.
| Climb | C2 | C4 | C10 | C9 |
|---|---|---|---|---|
| Rf carried into the foot | 31 | 30 | 33 | 44 |
| Rf held within the climb | 28 | 28 | 32 | 42 |
| Median tidal volume (AU) | 252 | 255 | 228 | 192 |
On all four climbs, the frequency held is lower than the one carried into the foot — by two cycles on average. The order of the climbs indicates that accumulated fatigue is not sufficient, on its own, to explain the difference: the climb entered at Rf 33 comes later than the one entered at Rf 44, at matched power, and its pattern is better, not worse. Duration is not sufficient either: the longest of the four is the most economical.
Below VT1 (335 W), correlation 0.73 and mean gap −2 cycles; above VT2 (375 W), mean gap +5.
| Repeatability | Day 1 | Day 3 | Day 5 |
|---|---|---|---|
| Median Rf | 33 | 35 | 38 |
| Time below 32 | 41% | 30% | 14% |
| Tidal volume when climbing | 245 | 229 | 225 |
Central finding
The volume ceiling closes in as the hours pass.
A ventilation band is fixed, and the volume produced inside it is tracked hour by hour. Within the same ventilation band, volume per cycle declines across the race on all three observed days.
| Band 75–90 L/min | Start of race | End of race |
|---|---|---|
| Day 1 | 296 AU | 183 AU |
| Day 3 | 243 AU | 199 AU |
| Day 5 | 265 AU | 198 AU |
In the 120–140 L/min band, where only two days carry enough data: day 3 from 330 to 262 over four hours, day 5 from 277 to 232 over three.
Day 1 is the one that matters.
It was the easiest stage of the week — 2,434 kJ, half the classic, at low altitude. The fact that the ceiling closes there too shows that stage difficulty alone is not sufficient to explain the phenomenon.
What this ceiling represents, in litres.
Everything above is in strap units. What follows uses none: frequency is a real measurement, and if ventilation is estimated in litres per minute, volume per cycle follows by simple division.
Over the last six minutes of the final climb, at Rf 60–69, volume per cycle sits between 1.9 and 2.5 L. At 3.4 L — a volume this rider does produce, from his own earlier measurements — the same ventilation would require a frequency of 38 to 51. He was at 69. The deficit is of the order of one litre per cycle.
Even assuming ventilation at 175 L/min, the matching frequency would still be 51.
Show the full estimation range
| Assumed true ventilation | Tv at Rf 69 | Tv at Rf 50 | Rf required for 3.4 L |
|---|---|---|---|
| 130 L/min | 1.88 L | 2.60 L | 38 |
| 140 L/min | 2.03 L | 2.80 L | 41 |
| 150 L/min | 2.17 L | 3.00 L | 44 |
| 160 L/min | 2.32 L | 3.20 L | 47 |
| 175 L/min | 2.54 L | 3.50 L | 51 |
From racing to the next test
The race produces the next question.
RACE→DEBRIEF→HYPOTHESIS→MONITORING→DECISION
Race observations become strategies to test during the next monitoring block, inside the sessions already prescribed by the coach.
Methodological boundaries
A serious signal, with explicit boundaries.
What the case shows
- A prediction dated before the event, then confronted with the file.
- A modifiable pattern at matched power, altitude and duration.
- Transfer observed in a classic, then across five racing days.
- An individual boundary identified on three independent readouts.
What it does not establish
- Exclusive causality.
- The number of watts or seconds recoverable.
- Dynamic hyperinflation, inferred but never measured.
- Calibrated tidal volume: ventilation is a derived product.
- Automatic generalisation.
One conclusion in the dossier was corrected along the way: a volume ceiling reported as “barely closing” had been computed without matching ventilation. At matched ventilation, it was closing by 20%. The correction is dated in the following report.
Robust enough to justify a shared evaluation — not to demand belief.
Value for the staff
An additional reading layer.
MEASURE
Rf, Tv, ventilation, power, duration, HR, NIRS, heat, altitude and perceived effort — second by second.
INTERPRET
Separate what is observed, what is merely compatible, and what requires a new measurement.
TRANSLATE
Reduce the analysis to a few markers usable inside the session or the race.
SHARE
Coach and staff keep the decision; ventilatory data enriches their model.
Specialised expertise is only worth something if it integrates without friction, stays auditable, and improves a decision. VST acts as an additional reading layer: the coach, performance staff and medical staff retain their scopes and the final decision.
Conclusion for the staff · observed VST contribution
What the VST intervention contributed to changing in this case.
A previously invisible constraint became measurable, steerable and verifiable in the field.
Fast spontaneous pattern; only 20% of time below the individual boundary on matched easy climbs.
Identification of the switch point, Rf–Tv targets, emptying strategies and successive reassessments.
Three days after feedback: −12 cycles/min and +52 points of time below the boundary, with intensity and terrain matched.
For the staff, the value is concrete: no longer merely documenting the decline after racing, but knowing where the rider’s boundary lies, when the pattern is likely to give way, which targets to use and whether they truly transfer to the field.
Within-athlete change observed across two nearby sessions, without a control condition: a strong operational longitudinal signal in this follow-up, without claiming isolated causal proof of VST.
A continuous decision loop
After every key session and every race day.
I review the file: what was executed, what held, what held less well, and when the ventilatory pattern began to lose efficiency.
The data are then considered alongside the rider’s own perception and the demands of the race. We look for the best possible fit between the targets set, the physiological response measured, and what the rider can actually apply when racing.
This loop — observe, compare, adjust, decide, verify again — allows the strategies to evolve with as little friction as possible for the rider and the staff. It makes ventilatory pattern efficiency usable as one of the levers of performance.
For the staff: VST does more than deliver targets. It provides a decision loop after every key session and every race day — data, rider feedback, adjustment and renewed verification. What emerges is not a pattern that works only in theory, but one that holds under race conditions.