The Relationship Between Physiological Characteristics and Durability in Male Professional Cyclists
Participants
n=10 male professional cyclists: VO2max 74.4 ± 4.8 mL·kg⁻¹·min⁻¹, W' 23.7 ± 5.4 kJ, CP (critical power) 5.6 ± 0.6 W·kg⁻¹; aged 19.2 ± 0.8 yr, 182.9 ± 4.0 cm tall, weighing 70.4 ± 5.5 kg, BMI 21.0 ± 1.3 kg·m⁻².
Protocol
A laboratory test (VO2max, ventilatory thresholds, gross efficiency and carbohydrate oxidation at fixed 200 W and 300 W loads) and a CP test, each on two occasions
The second CP test was preceded by a novel fatiguing protocol: five bouts of 8 min at 105-110% of CP
Durability metric: CP in the fatigued state expressed as a percentage of fresh CP, coined "delta CP" (ΔCP)
Pearson correlations between laboratory measures and ΔCP; multiple linear regression
Key Findings
ΔCP (the share of fresh CP retained after fatigue) was significantly and positively related to: relative peak power output (r = 0.891, P < 0.001), relative VO2max (r = 0.835, P = 0.003), relative power at the second ventilatory threshold (r = 0.738, P = 0.015), power at the first ventilatory threshold (absolute r = 0.748, P = 0.013; relative r = 0.826, P = 0.003), gross efficiency at 300 W (r = 0.869, P = 0.001) and at 200 W (r = 0.792, P = 0.006)
ΔCP was negatively related to carbohydrate oxidation at 200 W (r = -0.702, P = 0.024): riders burning more carbohydrate at 200 W lost more CP
A multiple linear regression predicted ΔCP from laboratory measures with R² = 0.96-0.98 (P < 0.001)
Authors' conclusion: the results show which physiological factors determine durability in professional cyclists
The abstract does not report the mean size of the CP decline
Practical Application
In 10 professional cyclists (age 19.2, VO2max 74.4), the share of critical power retained after five 8-min bouts at 105-110% of CP was strongly and positively correlated with the rested lab profile — relative peak power r = 0.891, relative VO2max r = 0.835, gross efficiency at 300 W r = 0.869, relative power at VT1 r = 0.826 — and negatively with carbohydrate oxidation at 200 W (r = -0.702); a regression on lab measures explained the variance with R² = 0.96-0.98.In this cohort and with this high-intensity fatiguing protocol, the better the rider's aerobic profile and efficiency, the smaller the loss.
Direction matters: higher VO2max, higher peak power, higher thresholds and better efficiency all went WITH better durability here — the opposite of the "durability is independent of the lab profile" reading. That reading comes from REF-196 Valenzuela, where the fatiguing ride was ~4 h submaximal and no associations were found; the two studies differ in fatiguing protocol, durability metric and analysed variables.
Inference, not tested: the negative correlation with carbohydrate oxidation at a fixed 200 W is consistent with fuel economy (lower CHO use at a given power) protecting late power, and gross efficiency was among the strongest correlates — so efficiency and substrate use, not only VO2max, belong in a durability profile.
Correlation in n=10 is not causation and not a training prescription; the abstract offers no intervention.
Limitations
n=10, correlational, single fatiguing protocol (5 × 8 min at 105-110% CP — high intensity, unlike the long submaximal rides of REF-196); results may not transfer to other protocols
Very young professional cohort (19.2 ± 0.8 yr), all male
R² of 0.96-0.98 from a regression on 10 subjects is at high risk of overfitting
Cross-sectional — does not inform causal direction
The abstract gives no mean ΔCP, so the size of the decline is not reported here
How this entry was checked and changed
Why this level
regraded 2026-09-27 from A-: correlational study, n=10
Content correction 2026-09-27
"n=30, VO2max ~73-78, FTP ~5.8-6.4 W/kg" -> n=10, age 19.2, VO2max 74.4 ± 4.8, CP 5.6 ± 0.6 W/kg, W' 23.7 kJ.
Fatiguing protocol "prolonged steady-state ride >30 kJ/kg" -> five 8-min bouts at 105-110% of CP.
"CP declined 4-7% (range -2% to -15%)" -> deleted; the abstract gives no mean decline.
"VO2max correlation small to none; GE small-to-moderate; LT1/LT2 weak" -> ΔCP correlated r = 0.835 with relative VO2max, r = 0.891 with relative peak power, r = 0.869/0.792 with GE at 300/200 W, r = 0.738-0.826 with threshold powers, r = -0.702 with CHO oxidation at 200 W; regression R² = 0.96-0.98. Direction and magnitude in the entry were wrong.
"Among 8 studies reviewed (KIW Note #154)", the range-restriction mechanism and "~3-5% CP test noise" -> deleted; not from this paper. Byline "Spragg, Levi, Sweet" -> Spragg, Leo & Swart. Level A- -> B; Tier and Sport lines added.
Resync 2026-09-28: Listed REF-426 as a durability cohort without noting it is untraced -> flagged as untraced C+.
2026-10-01: cross-reference repointed from a merged framework to the one that replaced it (SYNTH-01→31, 05→34, 09→70, 12/15→32, 19→26).