REF-489 · Bossi, Timmerman & Hopker, 2020 · International Journal of Sports Physiology and Performance, · Paper · PubMed
Energy expenditure equation choice: effects on cycling efficiency and its reliability
Participants
n=14 cyclists (11 men, 3 women; age 33 ± 10 yr; 76.0 ± 15.1 kg; VO₂max 51.4 ± 5.1 mL/kg/min; peak power 4.69 ± 0.45 W/kg).
Protocol
Five lab visits: a maximal ramp test, then 4 identical submaximal trials of three 7-min bouts at 60, 70 and 80% of the gas exchange threshold. Energy expenditure computed with 4 equations: Péronnet & Massicotte, Lusk, Brouwer, Garby & Astrup.
Key Findings
All 4 EE equations produced GE estimates that differed from each other (all P < .001). The size of the shift is not given in the abstract.
Changing the equation did NOT affect GE reliability: reliability parameters differed only when typical error was expressed in absolute GE units, a negligible effect tied to the magnitude of GE each equation yields.
Mean coefficient of variation for GE across intensities and calculation methods: 4.2%.
Authors' advice: share raw gas-exchange data so GE can be recalculated and compared across studies.
Practical Application
Bossi, Timmerman & Hopker 2020 (n=14 cyclists, 4 repeat trials): four common energy-expenditure equations give different absolute GE values (P < .001) but the same reliability; test-retest CV for GE was 4.2%. Compare GE only within one equation and one protocol; an absolute GE from another lab or paper is not on the same scale.
With a 4.2% CV, a change in GE of about one percentage point on a ~20% baseline (Inference, not tested: arithmetic, roughly 5% relative) sits near the noise of a single test — repeat measurements before reading a trend.
When a lab reports GE, record which EE equation it used; if you track an athlete longitudinally, keep the equation fixed.
Limitations
Methodological paper — does not change underlying physiology
n=14, 3 women; the abstract gives no magnitude for the between-equation GE difference, so the entry cannot say how big it is
How this entry was checked and changed
Why this level
regraded 2026-09-27 from B+: test/methods validation study, n=14
Content correction 2026-09-27
Level B+ -> B: methods/reliability study, n=14. Tier "N/A" -> Tier 2 (VO₂max 51.4).
Removed "shift of 0.3-1.2%", "CV ~1.5-2.5%", "training changes of 1-2%" and "Weir" (not among the 4 equations): none is in the abstract. Added the abstract's CV of 4.2%, the 4 equations, the protocol and n=14.
"Reliability across equations is not good" reversed: the abstract says changing the equation does not affect GE reliability, only the absolute estimate. "Trisystems-style lab tests" reference removed.