How to evaluate a portable metabolic analyzer (including ours)

Stationary metabolic carts

A cart sits in the lab, next to a treadmill or bike. In an independent 2026 study, researchers used two stationary carts as their reference because those systems had previously shown relatively high accuracy and reliability in both simulation and human testing.² Carts are the usual choice when you need a clinical-grade reference, a controlled room and a fixed protocol. The same study listed one of those carts at about €32,000 in the Netherlands in 2025.²

Portable analyzers

Portable systems are worn on the body, so you can test outdoors, on court, or in the athlete's own sport. A review of portable gas analysis calls it the gold standard for most field studies of energy expenditure.³ The trade-off is that the device has to cope with movement, weather and battery limits, and quality varies. When researchers tested 15 cart and portable systems against a metabolic simulator, absolute VO₂ errors ranged from 1.10% to 13.3%, and only a few systems showed consistently acceptable error.¹

Many labs keep a cart as the reference and add a portable unit for field work. If you do, test the two side by side on your own athletes before you trust the portable numbers.

1. Published validation against a reference. Is there a peer-reviewed study comparing this exact model with a stationary cart or Douglas bags in people? Results differ between devices and studies. One portable system overestimated VO₂ and VCO₂ by about 10–17% during moderate and vigorous exercise compared with Douglas bags.⁹ Look for bias and limits of agreement, not just correlation.
2. Test–retest reliability. How much do results change between two days on the same person? In repeated human testing, day-to-day VO₂ variability ranged from 3% to 12% across systems, excluding one outlier, and roughly 60–70% of that came from the person rather than the device.² Ask for between-day data, not just within-session data.
3. Firmware and model version. Was the validation run on the version you will receive? One wearable showed high accuracy in a 2024 study but large deviations in human testing in a 2026 follow-up. The authors note its firmware had been updated in between, and a company representative observed the repeat tests.²
4. Calibration routine. What does daily calibration involve, and how long does it take? For example, the VO2 Master manual asks for gas recalibration every hour and a new flow calibration with a calibration syringe whenever the user piece or filter changes.¹¹ In one validation study, both systems were calibrated with reference gases and a 3 L syringe before measurements.⁴
5. Sampling: breath-by-breath vs mixing chamber. Breath-by-breath shows fast changes, such as ventilatory thresholds. A mixing chamber averages expired gas and smooths the signal. Which mode works better depends on the device: one study found the COSMED K5 mixing-chamber mode more accurate at high intensity,⁴ while another found both modes comparable against Douglas bags.⁵
6. Which gases are measured. VO₂ needs flow and oxygen. RER, fuel use and the V-slope threshold method also need CO₂. At least one portable analyzer measures oxygen only.¹¹
7. Dead space and breathing resistance. What is the mask and sensor dead space, and how much resistance does the athlete breathe against? Researchers have pointed to mask dead space and airflow resistance as possible reasons why two portable systems disagreed at maximal exercise.¹⁵ Mask dead space can also bias VO₂ at low intensities when the device assumes a fixed inspired oxygen fraction.⁴
8. Raw data export. Can you export every breath, with time, flow and gas fractions, to CSV or similar? Or only a report? You need raw data to apply your own averaging, pick thresholds and run audits.
9. Hygiene, cleaning and throughput. Which parts touch the breath, how are they cleaned, and how long does that take between athletes? How many spare masks and sensors do you need for a busy testing day?
10. Battery, software fees and regulatory status. How long does a charge last on a full protocol? Is the software a one-off cost or a subscription, and do reports cost extra? Is the device cleared as a medical device where you work, or sold for sport and wellness only? This matters if you test patients.

A few notes from the same sources. COSMED states the K5 is a medical device, and it received FDA 510(k) clearance in 2017.^7,8 The VO2 Master manual states it is not a medical device.¹¹ The 2012 MetaMax 3B study concluded it was stable and reliable but needed further correction of VO₂ and VCO₂ during moderate and vigorous exercise.⁹ In the 2026 follow-up, Calibre showed large deviations from the reference during human exercise.² As of October 2026, the Calibre website says SplendoHealth has acquired key assets of Calibre Biometrics.¹⁶ Approximate 2025 prices in the Netherlands, from one study: VO2 Master about €6,900 and Calibre about €465.²

Here are our answers to the checklist today. We have no published validation yet, so we make no accuracy claims. Prototypes are planned for 2027. VenTracer is not a medical device.

• Validation: planned against a laboratory metabolic cart on the same athletes, using rest and an incremental ramp to exhaustion.
• Statistics: Bland–Altman bias and limits of agreement, plus intraclass correlation. VT1 and VT2 compared with cart thresholds picked by experienced physiologists.
• Publication: full results published before any accuracy figure appears on our site.
• Data: raw breath-by-breath export for pilot partners.

Hold us to the same ten questions you ask everyone else.

If you run a lab or clinic and want to see how VenTracer compares with your own cart, read how we work with labs or apply to be a pilot partner.

1. Van Hooren B, Souren T, Bongers BC. Accuracy of respiratory gas variables, substrate, and energy use from 15 CPET systems during simulated and human exercise. Scand J Med Sci Sports, 2024.
2. Van Hooren B, Souren T, Miqueu F, Bongers BC. Technological and biological reliability, and validity of five different CPET systems during simulated and human exercise. Scand J Med Sci Sports, 2026.
3. Macfarlane DJ. Open-circuit respirometry: a historical review of portable gas analysis systems. Eur J Appl Physiol, 2017.
4. Perez-Suarez I et al. Accuracy and precision of the COSMED K5 portable analyser. Front Physiol, 2018.
5. Crouter SE et al. Accuracy of the Cosmed K5 portable calorimeter. PLoS One, 2019.
6. Guidetti L et al. Validity, reliability and minimum detectable change of COSMED K5 portable gas exchange system in breath-by-breath mode. PLoS One, 2018.
7. US FDA 510(k) K162515, COSMED K5 Wearable Metabolic Technology, decision 20 June 2017. Record via Innolitics FDA database.
8. COSMED. K5 product page. Accessed October 2026.
9. Macfarlane DJ, Wong P. Validity, reliability and stability of the portable Cortex Metamax 3B gas analysis system. Eur J Appl Physiol, 2012.
10. Woodway. Cortex MetaMax 3B package specification (PDF). Accessed October 2026.
11. VO2 Master. VO2 Master Analyzer user manual. Accessed October 2026.
12. Montoye AHK, Vondrasek JD, Hancock JB. Validity and reliability of the VO2 Master Pro for oxygen consumption and ventilation assessment. Int J Exerc Sci, 2020.
13. Thiessen JS et al. (Re)assessment of the COSMED Quark CPET and VO2Master Pro systems for measuring pulmonary gas exchange and ventilation. Scand J Med Sci Sports, 2025.
14. Tsekouras YE et al. Validity and reliability of the new portable metabolic analyzer PNOE. Front Sports Act Living, 2019.
15. Kwok MMY et al. Assessing the validity and reliability of the PNOĒ for measuring cardiometabolic outcomes during walking exercise. J Funct Morphol Kinesiol, 2025.
16. Calibre. Calibre product website. Accessed October 2026.