REF-1530 · Caen et al., 2024 · Sports Medicine · Paper · PubMed

Are critical power and MLSS the same number?

Authors close to the CP model argue strict testing leaves a few watts; n, individual limits unread. A meta-analysis: +12 W, limits −19 to +44 W. Contested; for one rider, do not assume.

Main caveat: Abstract and public page only. Closed access; Europe PMC holds no full text, OpenAlex lists it as closed, the UGent file is restricted to UGent, and the VUB…
limited, mixed evidenceMeasured on a mixed group

Caen, Poole, Vanhatalo and Jones review how critical power (CP) and the maximal lactate steady state (MLSS) are determined and why earlier studies consistently found them apart in power output. They argue that inappropriate methods and inconsistency between studies have contributed to the documented differences.

Through a critical look at the literature and by integrating their laboratory data, they conclude that differences may be reconciled to only a few watts with strict criteria, so both indices may be used in practice, and they set out good-practice guidelines.

Our reading: the abstract says "we demonstrate", then uses "may" twice, gives no number and does not say the two are equivalent. It describes a mean-level claim under strict conditions. Whether two strictly measured numbers agree in one rider is a separate question that the abstract does not answer, and we could not read the part of the paper that might.

Protocol

Design: narrative review with integrated laboratory data; search method not stated in the abstract.
Who: n = not given in the abstract; the sample behind "our laboratory data" (sex, age, training status) is unread.
Not read: the criteria for CP and MLSS, the guidelines, how studies were judged, the data, any statistics.
Funding and interests: on the publisher's public page, no funding was used for the article and the four authors state no conflicts of interest directly relevant to the review.

Key Results

ItemWhat was readStatus
Earlier studiesConsistently reported significant discrepancies in power outputAuthors' summary; no figure in the abstract
Under strict criteriaMay be reconciled to "only a few Watts"No number, no interval
In practiceBoth indices may be used to estimate the thresholdAuthors' conclusion
n, who was measuredNot in the abstractUnread
Mean bias, SD of the difference, limits of agreementNot in the abstractUnread
Any correction for the step between MLSS trialsNot in the abstractUnread
Agreement: no limits of agreement or individual spread could be reported from this paper. A mean bias says nothing of agreement in one person.
Nearest figures we hold, not this paper: the same first author's 2022 study, as tabulated in REF-1316's full text from author-provided data: CP minus MLSS 3.0 ± 6.0 W in 10 women and 6.0 ± 14.1 W in 11 men. That is about −9 to +15 W and −22 to +34 W as mean ± 1.96 SD (our arithmetic, small samples).The same full text says only three of 42 comparisons in that study differed by more than 10 W, which sits oddly with an SD of 14.1 W. Whether these data are part of this review's is unread.

Mechanisms

Measured: nothing could be read here.
Authors' interpretation: CP and MLSS both estimate one threshold, and how each was determined has contributed to the gap in the literature, enough that strict criteria may reconcile it to a few watts.
Inference: better methods could shrink a mean gap by removing a systematic error; that alone does not show the two numbers land together in each rider. Any adjustment for the step size between MLSS trials, if the paper makes one, would be the authors' argument and not a measurement.
The tension with the base, one way: REF-1316 (A-; 10 small crossovers, 111 adults) found CP 12.42 W higher (95% CI 4.69 to 20.16; prediction interval −11.62 to 36.47) with limits of −19.23 to +44.08 W, and its own authors attribute much of that to method, cite this review for reconciliation under "very stringent criteria" and write that with more rigorous methods it is reasonable to expect "minor and practically irrelevant differences", while still warning against interchangeable use because of the spread between people.If the method argument is right, +12 W and those limits may overstate what a strict protocol would show. REF-1326 (B, abstract with statistical errors; Poole is a co-author) puts MLSS about 11% below CP, a larger gap from 9 small studies whose methods that abstract does not give.
The tension with the base, the other way: REF-1316 is systematic and reports individual limits; this is a narrative review whose abstract reports neither.The first author's 2022 samples are already inside REF-1316's pool, beside studies whose methods this review may fault (which ones is unread), so that pool as published does not test the claim either way (our inference).Three of the authors wrote REF-168 (B+), which reports CP about 7% above MLSS on average and asks that the two stop being treated as synonymous, while this abstract says both may be used.
Not resolved here. Reading the paper could shift this entry's level and wording either way; the abstract cannot settle it.

Practical Application

Are critical power and MLSS the same number? Authors close to the CP model argue strict testing leaves a few watts; n, individual limits unread. A meta-analysis: +12 W, limits −19 to +44 W. Contested; for one rider, do not assume.
Do not quote as "equivalent": the abstract says may be reconciled and may be used, under strict criteria, without a number.
Not good for: moving a rider's CP onto a lab MLSS, or the reverse, without testing both. Not from this paper; the base's own practice: pick one anchor and keep it through a season (SYNTH-27, SYNTH-53).
Good for: a reason to ask how each number was determined before reading a gap as physiology. The guidelines themselves are unread.

Limitations

Abstract and public page only. Closed access; Europe PMC holds no full text, OpenAlex lists it as closed, the UGent file is restricted to UGent, and the VUB listing led to a search page with no file. We did not try another route.
No data read. No n, no population, no bias, no spread, no criteria.
Narrative. No systematic method stated; what counts as "appropriate and strict" criteria is the authors' own call, and how they weighed the studies is unread.
Position. No conflict is declared. Poole, Vanhatalo and Jones are among the authors of the base's accounts of the CP model (REF-168, REF-024, REF-751), and the reference list opens with their 2019 paper; a stance in the debate, not a financial interest. The base also lists Dotan 2022 as a critical review from the other side (PHYS-07).
Funding and interests as printed: no funding for the article; no conflicts directly relevant to the review.
How this entry was checked and changed
Why this level

the rubric gives B+ to an authoritative narrative review from the leading group; held two steps lower because this entry carries one sentence of conclusion and no data: n, participants, mean bias, spread and methods are unread, and the conclusion comes from authors with a stated position on the question