Hey Alon

Thanks for posting this. As you mention me directly in your posting I hope you don't ming I reply directly to your two questions:

1) VO2max
To determine an accurate VO2max from power only data isn't only depending on gross efficiency (GE). But as your question is about this, let me comment on this. The review of James Hopkers group you linked here is great (side note: James Hopker and me discussed about this GE a few years ago). There is a widespread misunderstanding / miscommunication when it comes to GE, which unfortunately can also be found in this review: GE is not energy measured with a metabolic cart vs. mechanical power output. When GE is determined only using a met-cart (James Hopker has much more precise methods then standard lab carts), it often falls short in two areas: 1) met-cart reflects full body VO2 (and VCO2). To determine GE you are only interested in the O2 related to the actual muscle work which is involved in the mechanical power output (at least this is how we define GE - see comment below). So therefore one need to deduct an (estimated!) amount of VCO2 and VO2 which is not derived from the actual muscle work involved in energy production for the power you measure at the crank. 2) Energy derived from non-aerobic sources, or more precisely glycolysis isn't represented in any met-cart measurement. This means: the higher the anaerobic energy contribution the "better" the GE seems to be.
In case you want dive a bit deeper into this I recommend one recent blog posts which gives a vivid example in the case of measuring running economy: https://www.inscyd.com/blog/running-economy
When all those factors are taken into account the picture of GE changes. This said I recommend the work of Jim Martins group: https://www.physiology.org/...plphysiol.00982.2001.
In an inter-individual comparison even manipulating biomechanical factors like crank length didn’t change GE significantly (within the tolerance of a typical met cart).

In the case of INSCYD we always make it very clear that all metrics and data given relate only to the energy production for the mechanical power that is measured (and for example not full body). Firstly this also fixes any minor tolerances – like the supposed alterations of UCP-3 by nitrate supplementation as mentioned in the review you linked. Secondly it might happen that one sees a slightly higher VO2max in a lab measurement then you see in INSCYD, for example from upper body movement. Anecdotal information: we work with more than half a dozen WT cycling teams now, they usually compare INSCYD results with conventional lab tests before they roll it out to the team. 1.5y ago we had a case in which the lab test VO2max was significantly higher than in INSCYD. After double checking the data it turned out that INSCYD vs. measured VO2 was nicely correlated until the last 90s of the ramp test: here the athletes started to show abnormal – not to say quite extreme – upper body movement. This was the moment when the two VO2 data started to not match anymore.
Finally let me comment: assuming one is using mechanical power on the bike to estimate energy consumption or anything related to this – like usually done – this is always based on an estimated gross efficiency, which in INSCYD, you as a user can change in case you are not happy with the default value.



2) FatMax
First let me make clear that the 20min effort in the protocol you did, does not have any direct or more important connection to obtaining FatMax than any other data point of the whole data set. There are several (infinite to be precise) compositions of datapoints that could be used to create a metabolic profile in INSCYD.

Secondly I would like to clarify an obvious misunderstanding: I have not stated that there is a correlation between VLamax and %CHO at threshold. This must be a misunderstanding – sorry for this! At anaerobic threshold the energy derived from carbohydrates trends to 100%. You can see this also in the report that you linked: the fat combustion curves decreases to zero at AT.
So this said, at threshold efforts the carbohydrate combustion will trend towards 100% and fat combustion therefore towards 0% (again, looking at working muscle not at full body). This is not linked to the type of athletes (masters, U23, elite, etc as mentioned), but just basic physiology.
(In a lab test scenario like an incremental load test the percentages will look a little bit different - a tiny bit of remaining fat combustion at threshold - but discussing the physiological reason behind this maybes goes a bit to far here.).

Concluding it is important to point out that the FatMax is neither derived from the 20min test only nor from obtaining VLamax. It is a combination of ALL metrics / datapoints obtained. Regulation of fat metabolism is to complex to just getting estimated from one single effort or physiological metric. The fact that a combination of several datapoints is used to determine fat & CHO combustion rates also rules out your suspicion that bell curves are used, because this would be far too many bell curves we would need to do so. In point of fact, studies like the one linked above are used to do this.

I hope this helped to shed some light on those subjects

Sebastian,
Thanks so much for replying!
There’s still a few things I would like to understand better. Also, if you have any empirical evidence of these tests (determination of VO2max and fatmax vs. values of ‘gold standard’ tests), I haven’t seen any and that would give me much more confidence in the INSCYD tests.
Regarding GE, basically what I was getting at is that by measuring VO2 and RER at a sustainable (below anaerobic threshold) power, we can calculate the total calories burned, right? For the same amount of calories burned, different people will be able to produce different mechanical power (What many call GE). How does INSCYD account for these changes? If INSCYD does not account for basal metabolic rate, should I expect to see a slightly lower VO2max than I would measure from gas exchange?
If one is able to adjust the GE rate in INSCYD, where does one get this number from? Previous met-cart tests?

Regarding threshold, the quote I misunderstood is
"First building up the engine is something also very important because... the influence on something like the VLamax increases with VO2 max, so just like you say, if your FTP is 200W... then you don't need to worry that much about the VLamax, because the influence of VLamax on endurance performance increases with VO2max... so the higher the VO2max, once you are above 55, 60, 65, whatever, physiologically, the influence of the VLamax increases. So whatever you do there in terms of going up or down with VLamax by training, the effect is has on your endurance performance increases with increasing VO2max."
For others interested, the best explanation I found is:
http://www.lactate.com/...obic_gatekeeper.html

Still, can you point me to a paper that shows the correlation (and the reasons for it) for the assertion that high VLamax triggers higher lactate production rates at all sub maximum intensities? How well does INSCYD’s model hold up on less highly trained athletes?
Thanks so much for taking the time to answer these questions,
Alon