REF-031 · Jamnick et al., 2018 · PLOS ONE, Vol. · Paper

The lactate threshold and VO₂max cannot both be taken validly from one graded test: longer stages under-read VO₂peak by 1–5 mL/kg/min, and most traditional LT methods missed the maximal lactate steady state (Jamnick 2018, 17 trained cyclists)

Blimited evidenceMeasured on trained to highly trained athletesWritten before October 2026, being re-checked

Protocol

Seventeen trained male cyclists completed five graded exercise tests with stage lengths of 1, 3, 4, 7 and 10 min, plus a series of 30-min constant-power rides to find the maximal lactate steady state (MLSS). Fourteen LT methods were applied to the 3-, 4-, 7- and 10-min tests — 56 LT estimates — and compared with MLSS; VO₂ was measured in every test and in a verification bout.

Key Findings

VO₂peak on the 1-min-stage test 61.0 ± 5.3 mL·kg⁻¹·min⁻¹, peak power 420 ± 55 W; power at MLSS 264 ± 39 W.
The 3-, 4-, 7- and 10-min tests UNDER-estimated VO₂peak by about 1–5 mL·kg⁻¹·min⁻¹.
Many traditional LT methods were not valid estimates of MLSS. A newly developed modified Dmax on the 4-min-stage test came closest (mean difference 1.1 W, ICC 0.96).
The authors conclude that VO₂max and the lactate threshold cannot be determined in a single graded test, even with a verification bout.

Practical Application

Test VO₂max and the threshold separately: a short-stage ramp for VO₂max, a longer-stage step test for the threshold. Reading VO₂max off a lactate step test under-reads it by 1–5 mL/kg/min.
The method matters as much as the protocol. Only some LT methods landed on MLSS; the modified Dmax from 4-min stages was best in this group. A "lactate threshold" without the method named cannot be compared with anything.
Keep protocol and method identical across retests — otherwise a change in the number may be a change in the test.

Limitations

17 trained men, one laboratory; women and other tiers untested.
Validity is judged against MLSS, which carries its own measurement error.
Group agreement (mean difference, ICC) hides individual error, which the abstract does not report.
How this entry was checked and changed
Why this level

regraded 2026-09-26 from B+: measurement study, n=17

Content correction 2026-09-26

This entry used to say that 7–10-min stages give reliable thresholds (CV < 5%) while 3–5-min stages are invalid (CV > 15%), that the same cyclists produced "56 different protocols", and that VO₂peak stabilises with stages over 7 min. None of those numbers are in the paper and the direction was wrong: the 56 are LT ESTIMATES (14 methods × 4 tests), longer stages UNDER-estimated VO₂peak, and the most valid threshold came from a 4-min-stage test.