Power profiling and the power-duration relationship in cycling: CP + W' for individualised prescription
Key Findings
Field power data now inform training prescription, performance testing and race analysis; the review sets out how to build a rider's power profile and derive the power-duration relationship from it
Power-duration parameters come from the 2-parameter critical power model, P(t) = W'/t + CP (W' = work capacity above CP)
Field testing should follow pre-established guidelines: a set number of prediction trials, defined inter-trial recovery, road gradient and data analysis
The authors recommend avoiding single-effort prediction trials such as FTP — a single effort of one duration cannot fix both CP and W'
Laboratory and field testing should be combined and matched to rider specialisation and race demands (ecological validity)
Application
Inference from the model, not a number in the abstract: above CP the time to exhaustion at a constant power P is W'/(P − CP), so severe-domain interval length and recovery follow from the individual W', not from a fixed %FTP
After a CP/W' test, re-derive above-threshold interval durations from the individual W'
Practical Application
What the review recommends: profile the rider from several maximal efforts of different durations under standardised field-test rules, fit the 2-parameter CP model (P = W'/t + CP), and use CP and W' — not a single FTP effort — to individualise severe-domain prescription; the authors explicitly advise against single-effort prediction trials such as FTP.
Not from this paper; the base's own practice: re-test CP/W' roughly every 8-12 weeks or after a training block, and re-scale above-CP interval work from the new W'.
Not from this paper; the base's own practice: for repeated short efforts (under ~2 min), programme recovery so that most of W' has reconstituted — W' reconstitution kinetics under fatigue are held by Chorley 2022 REF-131 (τ-slow 388 s).
Limitations
Narrative review, not systematic—author selection of evidence may be biased
The 2-parameter CP model assumes a hyperbolic power-duration relationship; the abstract does not discuss its limits at very short or very long durations
CP is unstable across fatigue and prolonged exercise (see REF-145)—single-day measurement may not predict race-day CP
Inference, not in the abstract: W' estimates carry larger relative error than CP, so W'-based prescription is inherently noisier
How this entry was checked and changed
Content correction 2026-09-27
"FTP sufficient for sub-threshold zones" and "FTP typically sits 2-6% below CP" -> the abstract recommends avoiding single-effort prediction trials such as FTP and gives no CP-FTP percentage; removed/replaced.
"Every 8-12 weeks", "<2 min", "W' reconstitution ≥70%", "3-parameter behaviour", "CV ~10-15% vs ~3-5%" -> not in the abstract; removed or relabelled as the base's own practice / inference. "HIDIT and Billat" removed (not in the abstract).
Cross-reference "Jones 2025 REF-024 — CP declines after 2 h" removed (not verified against REF-024); REF-145 line aligned with what that entry states.
2026-09-28: heading "Key Finding" -> "Key Findings" so the section is recognised as the results section; content unchanged (checked against Europe PMC PMID 34708276).