REF-1455 · Spragg et al., 2023 · European Journal of Sport Science · Paper · PubMed

Does training mix relate to power when fatigued?

In 30 U23 pros, a shift toward polarized went with better 12-min power after 2,000 kJ (r = 0.41). A correlation across a season, not a test of tempo. Abstract only.

Main caveat: Correlation only; no intervention or control.
limited evidenceMeasured on highly trained to elite athletes

Spragg, Leo and Swart collected data from 30 under-23 professional cyclists across a competitive season, collated into early, mid and late periods. Two power profiles were built from maximal mean power over 2, 5 and 12 min: fresh (before 2000 kJ of work) and fatigued (after 2000 kJ).

Absolute 5- and 12-min and relative 12-min fatigued power were significantly lower late in the season than early and mid-season, and the gap between fresh and fatigued 12-min power grew.

Time below the first ventilatory threshold related to improvements in absolute and relative 2-min fatigued power (r = 0.43, p = 0.018; r = 0.376, p = 0.04), and a shift toward a polarized distribution related to improvements in absolute and relative 12-min fatigued power (r = 0.414, p = 0.023 for both), between subsequent periods.

The authors conclude there is greater variability across a season in the fatigued than the fresh profile, and highlight that a tendency toward polarized is associated with a better fatigued profile. Our reading: a rare season-level link between training mix and durability, and a useful reminder that fatigued power moves.

It is correlation in a selected professional cohort, where racing load and calendar drive both training mix and power; a rider peaking may simply train and race differently in those weeks.

Protocol

Participants: n = 30 under-23 professional cyclists, 20.1 ± 1.1 y, 69 ± 6.9 kg, VO2max 73.8 ± 2.5 mL/kg/min, CP 5.48 ± 0.38 W/kg, W′ 17.83 ± 3.57 kJ.
Data: a competitive season divided into early, mid and late periods; fresh profile before 2000 kJ and fatigued profile after 2000 kJ, from maximal mean power at 2, 5 and 12 min; training characteristics analysed against the profiles.
Not in the abstract: data source and whether efforts were truly maximal, how the distribution was quantified, number of correlations tested, correction for multiple tests.
Funding and interests: none stated in the PubMed or Crossref record.

Key Results

Relation (change between periods)Outcomerp
Time below first ventilatory thresholdabsolute fatigued 2-min power0.430.018
Time below first ventilatory thresholdrelative fatigued 2-min power0.3760.04
Shift toward polarizedabsolute and relative fatigued 12-min power0.4140.023
Season pattern: absolute 5- and 12-min and relative 12-min fatigued power significantly lower late in the season than early and mid-season (p < 0.05); the gap between fresh and fatigued 12-min power significantly greater late.

Mechanisms

Measured: seasonal decline in fatigued power and correlations with training characteristics.
Authors' interpretation: the fatigued profile varies more than the fresh one, and a tendency toward polarized is associated with improvement in it.
Inference: reverse causation and confounding by racing schedule and fatigue are not excluded; maximal efforts after 2000 kJ arise in racing, not under test control.

Practical Application

Does training mix relate to power when fatigued? In 30 U23 pros, a shift toward polarized went with better 12-min power after 2,000 kJ (r = 0.41). A correlation across a season, not a test of tempo. Abstract only.
Durability: not from this paper; the base's own practice for building it is long rides with hard efforts late in them, with heavy strength among what may help, on data its sources call scant (SYNTH-32, B-). The direct experiments are REF-1099 (B, untrained, durability rose alike with low or high intensity) and REF-193 (B, intensity of prior work).
What it does not show: that polarized training builds durability, that between-threshold work harms it, anything in amateurs or women. It tested no sweet-spot format.

Limitations

Correlation only; no intervention or control.
Abstract only; closed access. Methods and the number of tests unknown.
Maximal mean powers from a season are not shown to be standardised efforts, and the opportunity to produce them may change with racing (our reading; the abstract gives no data source).
Elite young men only, all professional, one season.
Modest r values with p values near 0.02-0.04, uncorrected as far as the abstract shows.
Funding and interests: not stated in the record.
How this entry was checked and changed
Why this level

(from the abstract) — the rubric's B for a small cohort or correlational design; r values around 0.4 in 30 riders explain about a sixth of the variance (our arithmetic), and the abstract does not say how many relations were tested