REF-246 · Chorley & Lamb, 2023 · Sport Sciences for Health, Vol. 19, pp. 1199-1207 · Paper · PubMed

Effect of varying recovery intensities on power outputs during severe intensity intervals in trained cyclists during the Covid-19 pandemic

B+moderate evidenceMeasured on trained to highly trained athletesWritten before October 2026, being re-checked

Participants

n=18 trained cyclists (FTP 258.0 ± 42.7 W, weekly training 8.6 ± 1.7 h/week), familiar with interval training and with the Zwift platform through the Covid-19 pandemic; FTP established as 95% of 20-min mean power. The abstract does not say whether ERG mode was used.

Protocol

5 × 3-min severe-intensity efforts with 2-min recoveries. Three recovery conditions: LOW (50 W), MOD (50% FTP), SELF (self-selected power).

Key Results

Recovery ConditionMean Interval Power
LOW (50 W)300.1 ± 48.1 W
MOD (50% FTP)296.9 ± 50.4 W
SELF (self-selected)298.8 ± 53.3 W
No difference in severe-interval power across recovery conditions; power declined across the session in all of them.
Self-selected recovery averaged 121.7 ± 26.2 W, but was not flat: the last 15 s of each recovery was harder than the first 15 s (p<0.001), and recovery power fell across the session (128.7 ± 25.4 W to 113.9 ± 29.3 W).
Authors' conclusion: dropping recovery intensity below 50% FTP did not enhance the next interval, suggesting a lower limit for optimal W′ reconstitution had been reached; because self-selected recoveries adapted to maintain interval power, that strategy may be preferable.

Practical Application

n=18 trained cyclists (FTP 258 W), 5 × 3 min severe with 2-min recoveries: interval power was the same whether recovery was 50 W, 50% FTP or self-selected (300.1 vs 296.9 vs 298.8 W); riders left to themselves recovered at ~122 W on average, eased off more as the session went on (128.7 to 113.9 W) and ramped up in the last 15 s before each effort — within a 2-min rest, going softer than 50% FTP buys nothing.
For 2-min recoveries between hard 3-min efforts, prescribing recovery power is not critical — anything from near-zero to 50% FTP gave the same next interval. Let the athlete self-select; the paper's authors lean that way.
Inference, not tested: W′ reconstitution over 2 min appears to have a floor — once recovery is easy enough, going easier does not restore more. Longer recoveries or higher recovery intensities (above 50% FTP) were not tested.
Inference, not tested: this study does not address ERG mode (the abstract does not mention it); the base's earlier reading of it as "evidence for ERG standardization" was an assumption.

Limitations

Zwift-only platform. Whether ERG mode was used is not stated in the abstract; no non-ERG comparison either way. 2-min recovery may be too short to detect recovery-intensity effects that emerge over longer rest periods. Covid-19 context may have affected training status. Recovery intensities above 50% FTP not tested.

How this entry was checked and changed
Why this level

regraded 2026-09-27 from B: controlled crossover, n=18

Content correction 2026-09-27
Level B -> B+ (controlled crossover, n=18 per rubric). Sport line "cycling, swimming" -> cycling.
Removed "All training on Zwift using ERG mode for interval control" and the "ERG mode as research tool" lead -> the abstract never mentions ERG mode; lead now the study's finding with the abstract's values.
Added the self-selected recovery data the entry omitted (121.7 W mean; 128.7 -> 113.9 W across the session; last 15 s harder than first 15 s) and the authors' conclusion that self-selected recovery may be preferable.
Cross-references to duplicate stubs repointed to their canonical entries (REF-200->REF-024); the stubs were emptied on 2026-09-27.