REF-1081 · Mentzoni et al., 2024 · European Journal of Applied Physiology, · Paper · PubMed · Free full text

Precision and accuracy of four handheld blood lactate analyzers across low to high exercise intensities

Against a Biosen reference between 0.88 and 4.89 mM, the four handheld meters (two units each of Lactate Plus, Lactate Pro2, Lactate Scout 4 and TaiDoc TD-4289) read on average −7% (Plus), 7% (Pro), −10% (Scout) and 42% (Tai), with 95% prediction intervals 0.60-0.87 mM wide at 3 mM; precision was poor for all four, and even the second stationary analyzer (YSI) read −32%: lactate values from different meters are not comparable

Blimited evidenceMeasured on a mixed groupWritten before October 2026, being re-checked

The Norwegian Olympic Sports Center compared measured blood lactate concentrations from four commonly used handheld lactate analyzers with two stationary analyzers, on venous samples taken at exercise intensities from low to high.

All four handheld meters had poor precision compared with the stationary analyzers; their mean offsets from the Biosen differed by brand and direction, and the most precise handheld meter (TaiDoc) was the least accurate. Content from the abstract only.

Participants

Not described in the abstract (venous blood samples taken at exercise intensities ranging from low to high). The unit of comparison is the paired measurement, not the athlete.

Protocol

Venous blood samples taken at exercise intensities ranging from low to high.
Each sample measured simultaneously with four pairs of handheld analyzers (two new units of each brand): Lactate Plus, Lactate Pro2, Lactate Scout 4, TaiDoc TD-4289.
Compared with two stationary analyzers: Biosen C-Line (the reference) and YSI Sport 1500.
Measurements repeated across Biosen-measured concentrations from 0.88 to 4.89 mM, with a median difference between measurements of 0.10 mM.
Outcomes: mean relative difference to Biosen (accuracy); residual standard error after linear regression against Biosen and the resulting 95% prediction interval (precision).

Key Results

AnalyzerMean relative difference to BiosenResidual standard error vs Biosen95% prediction interval width at 3 mM (Biosen)
Lactate Plus−7%0.18 mM0.72 mM
Lactate Pro2 ("Pro")7%0.20 mM0.80 mM
Lactate Scout 4−10%0.22 mM0.87 mM
TaiDoc TD-4289 ("Tai")42%0.15 mM0.60 mM
YSI Sport 1500 (stationary)−32%0.06 mM0.23 mM
Compared with the two stationary analyzers, the precision of all four handheld analyzers was poor.
TaiDoc was the most precise of the four handheld meters but had low accuracy, with a substantial mean difference to the reference analyzer.
The direction of the offset differed between brands: Plus and Scout read below Biosen, Pro2 and TaiDoc above.

Practical Application

"Can I compare my handheld lactate reading with a lab value, a published zone table or another meter?" In a device study at the Norwegian Olympic Sports Center (simultaneous venous samples, 0.88-4.89 mM), four handheld meters differed from the Biosen reference by −7% (Lactate Plus), 7% (Lactate Pro2), −10% (Lactate Scout 4) and 42% (TaiDoc), and a true 3 mM gave a 95% prediction interval 0.60-0.87 mM wide on the handheld meters; precision was poor on all four. Not across devices: a number is only comparable with numbers from the same meter.
Inference, not tested here: a fixed-concentration threshold ("2 mM", "4 mM") is device-specific — with offsets of 7% and −10% in opposite directions, the same rider's 3 mM on a Lactate Pro2 and on a Lactate Scout 4 are not the same intensity. Not from this paper: Olympiatoppen's intensity-scale guide states that its lactate values are Biosen values and cites this study for the larger variation between handheld devices.
A single reading near a zone boundary cannot place a rider in one zone or the next: at 3 mM the 95% prediction interval on a handheld meter was 0.72-0.87 mM wide for three of the four models.
Even the two stationary analyzers disagreed (YSI −32% vs Biosen): a lab-to-lab comparison has the same problem.
Not from this paper; the base's own practice: keep one meter (and one strip lot within a test block) for all of an athlete's tests, use the same sampling site and technique, and read a change in lactate at a fixed power or pace only against that athlete's own earlier tests on that meter (REF-030 makes the same "one brand" point from a 2015 device study).

Limitations

Only the four named handheld models and their strip generation; results do not transfer automatically to other meters or newer strips.
Venous samples; capillary (finger, earlobe) sampling, as used in the field, was not the subject of this comparison in the abstract.
Concentrations 0.88-4.89 mM only — nothing about peak post-exercise values.
The abstract does not describe the people who gave samples, the number of samples, or the exercise protocol.
Biosen is the chosen reference; the −32% difference of the YSI shows the reference itself is a convention, so "accuracy" here means agreement with Biosen.
How this entry was checked and changed
Why this level

test validation study