REF-192 · Mateo-March et al., 2022 · International Journal of Sports Physiology and Performance · Paper · PubMed

The Record Power Profile of Male Professional Cyclists: Fatigue Matters

B+moderate evidenceMeasured on elite to world-class athletesWritten before October 2026, being re-checked

Participants and Design

n=112 professional male cyclists; age 29 (6) y, 8 (5) y of professional experience
46 ProTeam (PT) level
66 WorldTour (WT) level
Period: 2013-2021, 8 (5) seasons per cyclist (historical data)

Protocol

Observational analysis of the mean maximal power (MMP) values of 112 professional male cyclists during 2013-2021, with ProTeam and WorldTour categories compared (from the abstract).
Protocol: record power profile — mean maximal power (MMP) for efforts lasting 10 seconds to 120 minutes, computed with no prior fatigue (after 0 kJ·kg⁻¹ of work) and after increasing prior work: 15, 25, 35 and 45 kJ·kg⁻¹
Inference, not tested: at typical professional endurance intensities 15 kJ·kg⁻¹ corresponds to very roughly three-quarters of an hour of riding and 45 kJ·kg⁻¹ to a few hours; the paper reports work done, not time

Key Results

> ⚠️ FIGURES WITHDRAWN 2026-08-20 — they were not from this paper. This section used to carry a

> table of absolute watts for nine durations together with a decline at 35 kJ·kg⁻¹ larger than the

> published decline at 45 kJ·kg⁻¹ and a WT-vs-PT split that appears nowhere in the publication.

> Not from this paper: the withdrawn absolute watts also failed an external sanity check against

> the published 10-second MMP of professional cohorts in DOI 10.3390/sports13110406 — a mean-maximal

> curve must decrease with duration, and the withdrawn 30-second value exceeded that 10-second value.

>

> **What the abstract actually publishes (PMID 35240578, verified against PubMed 2026-08-24 and

> re-verified 2026-09-27): PERCENTAGES ONLY.** MMP declined significantly and progressively across all

> durations from the lowest fatigue level: −1.6% to −3.0% after 15 kJ·kg⁻¹ and −6.0% to −9.7% after

> 45 kJ·kg⁻¹ (P < .001). PT cyclists decayed more than WT (P < .001) and the gap widened with fatigue:

> after 15 kJ·kg⁻¹ WT −1.8% to −2.9% vs PT −1.1% to −4.4%; after 45 kJ·kg⁻¹ WT −4.7% to −8.8% vs

> PT −7.6% to −11.6%. With no fatigue (0 kJ·kg⁻¹) there were no consistent WT-PT differences in MMP;

> with accumulating fatigue WT attained significantly higher MMP values, particularly for long efforts (≥5 min).

>

> What survives from the paper: the qualitative finding — resistance to MMP decline under

> accumulated work separates WorldTour from ProTeam, and it is independent of fresh values — and the

> percentages above. The watts may not be cited.

>

> Where to get professional watts instead (not from this paper): DOI 10.3390/app122111020

> (Mateo-March 2022, Appl Sci, CC-BY) publishes W/kg across a range of durations for a WorldTour cohort — one cohort, one method, open licence.

Mechanism ("Durability" Concept)

"Durability" = the ability to maintain power output as work accumulates. The paper measures it (MMP after prior kJ·kg⁻¹) and shows it separates the two professional tiers; it does not identify the physiological cause.
Inference, not tested: candidate mechanisms are the ones the durability literature proposes — glycogen sparing and substrate use, fatigue resistance of the recruited fibres, and pacing skill (REF-195). None was measured here.

Application

A power profile measured fresh is not the power that gets raced. Across 112 professionals, every MMP duration fell progressively with prior work: −1.6% to −3.0% after 15 kJ·kg⁻¹ and −6.0% to −9.7% after 45 kJ·kg⁻¹. Test rested, prescribe for the late race, and expect the gap.
Fatigue resistance, not fresh power, is what separated the two professional tiers. At 0 kJ·kg⁻¹ there were no consistent differences between WorldTour and ProTeam. Under load there were: after 45 kJ·kg⁻¹, WT declined −4.7% to −8.8% against PT −7.6% to −11.6%, and the WT advantage was clearest for efforts of 5 min and longer.
This is the argument for testing after work, not only when fresh. If two athletes have the same 5-minute power rested and one holds it after 45 kJ·kg⁻¹, they are not the same athlete, and a fresh-state profile cannot tell them apart.
Do not quote absolute professional watts from this entry. The paper reports percentages. For W/kg reference values use the open-access dataset named in Key Results (not from this paper).

Limitations

Observational, from historical training and racing files — not a controlled protocol. "Fatigue" is accumulated kJ·kg⁻¹ before the effort, not a standardised preload, and the efforts are whatever these riders happened to produce in training and competition.
The best available effort at a given fatigue level is a FLOOR, not a maximum. MMP mining records what was done, not what could have been done; a rider may simply never have gone deep after 45 kJ·kg⁻¹.
Professional men only (Tier 4-5). Nothing here establishes the size of the decline in amateurs, in women, or in any other sport.
The published numbers are percentage ranges across durations, not a single figure. Quoting "about 9%" as if it applied to every duration overstates the precision; the range is what the paper supports.
This entry previously carried fabricated absolute watts for nine months (withdrawn 2026-08-20, see Key Results). Treat any number in an older note, screenshot or downstream summary derived from this entry before that date as suspect.
How this entry was checked and changed
Content correction 2026-09-27
All percentages re-verified against the abstract (unchanged). Added age 29 (6) y, 8 (5) y experience, 8 (5) seasons per rider.
"~45 min / ~75 min / ~2+ h / ~3 h" time equivalents of the kJ·kg⁻¹ levels and the nine-duration list (10 s … 120 min) removed as stated facts; the abstract gives only "10 seconds to 120 minutes" and work levels. Time equivalence kept only as labelled inference.
Mechanism block (Z2-Z3 time, mitochondrial efficiency, pacing experience) rewritten as labelled inference; the paper measured the decline, not its cause.
External-source figures in the withdrawal note (other cohorts' watts, mass, n) reduced to the DOIs and labelled "not from this paper". Level A -> B+ (observational cohort, n=112); Sport "cycling, sprint" -> cycling.
Cross-references to duplicate stubs repointed to their canonical entries (REF-162->REF-306); the stubs were emptied on 2026-09-27.
Cross-reference labels pointed at the wrong entry or named the wrong author (checked against each cited entry's byline); repointed to the entry the label describes, or removed where the base holds no such paper.
2026-09-28: added Protocol heading and design bullet from the abstract (Europe PMC PMID 35240578).