Hi Andrew

You're right in that there is a wide range of reported values for sampling frequency to determine VO2max. A 10 second sample will give a greater chance of capturing a higher average, and that is why a large number of studies use 10 s averages. If you go up to 60 seconds, any peak VO2 may be inadvertently 'drowned out' by lower values earlier in the stage. A recent statement by Mezzani (2017) advocates 2-30 seconds. 

However, a recent statement by Poole & Jones (2017) has stated that ideally supra-maximal verification bouts should be conducted to ensure that VO2max has actually been obtained, otherwise the term VO2peak should be used.

Hopefully the following references may be of assistance:

Scheadler, C. M., et al. (2017). The Gas Sampling Interval Effect on VO2peak is Independent of Exercise Protocol. Medicine and Science in Sports and Exercise. ePub ahead of Print. 

Hill, DW et al (2003). Effect of sampling strategy on measures of VO2peak obtained using commercial breath-by-breath systems. Eur J Appl Physiol, 89(6):564-9.

Poole, D. C. and A. M. Jones (2017). Measurement of the maximum oxygen uptake Vo2max: Vo2peak is no longer acceptable. Journal of Applied Physiology 122(4): 997-1002.

Mezzani, A. (2017). Cardiopulmonary Exercise Testing: Basics of Methodology and Measurements. Annals of the American Thoracic Society. In Press. 

Many thanks Owen, that's extremely helpful. Will have a read.

Hi Andrew

No problem. I should add that the Mezzani (2017) paper advocates 20-30 seconds. Not 2-30, sorry about the typo!

Owen

Hi,

I'd say it depends what you want to measure. I would argue that with 3-min steps you are not trying to measure 'peak' values (especially in terns of power output). Similarly, averaging over 20-30 second will iron out artefacts, but also potentially miss meaningful data.

This being said, I'd agree with Owen's (comprehensive) feedback; 20-30 sec is a well-accepted timeframe.

If you've used automated machines such as a Medgraphic, you can choose to have the data displayed using a mean of X breaths or 5 out 7. For the latter, it is removing the highest and lowest data points, giving you an average of the remaining 5. I find it a good compromise, as it still removes the outliers and accounts for changes in breathing pattern; something the arbitrary 20-30 average doesn't offer (you would average more data points towards max as you hyperventilate). Something to think about if nothing else

We are using 15 or 20s averages, depending on the protocol and the type of variables co-measured with VO2 (cardiac output, blood flow, etc.). 

I'd advice you to check the following article:

Robergs RA, Dwyer D, Astorino T. Recommendations for improved data processing from expired gas analysis indirect calorimetry. Sports Med. 2010.

Briefly they suggest to use either 30s time average, 15-breath rolling average or a 0.04 Hz low cut-off digital filter (based upon the matabolic cart you have, and if it can record the expired gas breath-by-breath).

Hope it helps.

Regards