Saturday, August 24, 2013

Looking under the hood

Today I'm going to take a look under the hood of Functional Threshold Power and explore the relationship between four key underpinning physiological parameters that determine FTP:

  • VO2max
  • Energy yield from aerobic metabolism
  • Efficiency
  • Fractional utilisation of VO2max at threshold
I've prepared a chart (sample below), which I will come back to later to explore this relationship a little more. Those with an existing understanding of the relationship will likely need not look further than the various charts posted, and as normal you can click on them to reveal a larger version.

Scroll down and you will see several versions, showing the relationship between FTP, VO2max and GE at fractional utilisation of 75%, 80%, 85% and 90% of VO2max.

Our maximal sustainable aerobic power is primarily a function of our VO2max,
our gross efficiency, and our fractional utilsation of VO2max at threshold.

For everyone else, I'll introduce and explain the significance of each of these factors and then give an example of how changes affect the power we can sustain.

VO2max testing by the AIS
as reported by Katya Crema

VO2max is a measure of the maximal rate at which we can utilise oxygen. Normally it's also defined by how it is measured, e.g. during an incremental exercise test where the power demand is increased at a specified rate, and how long VO2max is sustained for, so that we don't rely on instantaneous peak values. Measurement of oxygen utilisation requires laboratory testing equipment that records the flow and composition of the body's respiratory gases while performing exercise.

VO2max will typically occur eventually when attempting to sustain a power output above functional threshold, and once reached is typically not sustainable for more than a handful of minutes. How quickly we attain a state of VO2max, and how long we can sustain it are determined by how far above functional threshold power we are attempting to ride, our fitness, power profile and some other individual characteristics.

VO2max is expressed in units of oxygen consumption per unit time, either absolute, i.e. litres of oxygen per minute, or relative to body mass, i.e. millilitres of oxygen per kilogram per minute.

e.g. if a 70kg rider has a VO2max of 60ml/kg/min, it means that for every kilogram of body mass, they can maximally utilise 60 millilitres of oxygen per minute, or 70kg  x  60ml/kg/min  /  1000 ml/litre = 4.2 litres of oxygen per minute.

VO2max sets the ceiling on our aerobic performance capability and as such is a reasonable determinant of our endurance performance potential, however it's not a particularly good predictor of performance. All one can really say is that to be an elite and/or professional cyclist, you will need a relatively high VO2max, typically in excess of 70ml/kg/min, however higher doesn't necessarily mean you will perform better. It just gets you a ticket to the game, but won't necessarily mean you'll be good enough to play.

That's because power output matters far more than how much oxygen we happen to use to generate it, and VO2max is not the sole factor in how much power we are capable of sustaining. And of course there are other factors beyond physiological that determine performance, but all things considered, in endurance cycling power output is a major factor.

VO2max is trainable, although it is also significantly genetically determined (perhaps half), so in a sense, you need to have chosen your parents wisely. You may not see much improvement in absolute VO2max from training, or quite a lot, or something in between. Trainability, which differs by individual and also has a sizeable genetic component, and starting fitness level are big factors. Improvements in VO2max of around 10-25%, can occur in a matter of months. Of course one can attain improvement in VO2max when expressed per unit of body mass simply through weight loss.

There have been some phenomenally high VO2max values occasionally recorded, well into the 90+ ml/kg/min range, with young Norwegian cyclist Oskar Svendsen reported to have the highest recorded VO2 max of 97.5ml/kg/min. Greg Lemond, the American professional cyclist of the 1980s and early 1990s and winner of three Tours de France, was reported to have had a VO2max of 92.5ml/kg/min, and he responded in an interview once that it was in the 92-94ml/kg/min  range. I don't vouch for the validity of these numbers, merely pointing out some of what's been reported.

Energy yield from aerobic metabolism
Citric acid cycle as per wikipedia

Without oxygen we'll die (well duh), and it's critical for sustaining our body's energy production needs, and just like many means of releasing energy through chemical reactions (e.g. rockets, campfires, internal combustion engines and many other chemical reactions), our bodies also use oxygen to help release useful energy from fuel.

Of the biochemical reactions that release energy aerobically (i.e. with oxygen), we utilise two primary fuel sources, one being glycogen and the other free fatty acids (from our body fat stores). Most of the time we obtain energy from both, but when exercising at near threshold power and above, we are heavily, if not solely, reliant on glycogen to meet the energy demand.

Using glycogen as fuel, our body can release around 21.1 kilojoules (kJ) of energy per litre of oxygen. We get a little less from aerobic fat metabolism, around 19.8kJ per litre. So in general the energy released per litre of oxygen utilised is somewhere around 20-21 kJ depending on the mix of fuel substrate used.

We do have the means to also produce energy without oxygen (i.e. anaerobically), however such energy pathways are available to us only for brief periods and are not sustainable, but are good for rapid energy demand (e.g. sprinting) and to supplement the energy provided via aerobic means when the energy demand exceeds our ability to supply via aerobic metabolism alone. Due to the limited supply of such energy though, such efforts are of short duration (seconds to minutes).

Here's a summary of the main energy pathways used by our bodies. It's a fairly complex topic (e.g. just look up the Kreb's Cycle for starters), and is one for the physiologists to chat about over a beer, beer being another key fuel substrate and one of the major food groups, along with burritos, donuts, caffeine and chocolate.


The basic definition of gross efficiency (GE) is the ratio of work done during the specific activity to the total energy expended and expressed as a percentage.

In the case of cycling, GE is the ratio of the energy delivered to the cranks of the bicycle to the total energy metabolised by our body. Sometimes this is referred to as gross mechanical efficiency (GME), just to emphasise the relationship between the mechanical work done at the cranks, to the total energy metabolised by the body.

There are a number of definitions of efficiency in exercise physiology and if you'd like to read about them in a little more depth, then this paper: The reliability of cycling efficiency by Lukh Moseley and Asker Jeukendrup (MSSE 0195-9131/01/3304-0621/$3.00/0 ) is a reasonable place to start and I'm sure there are others. That's just the PubMed extract which doesn't say much about the definitions, but you can find full text version online if you search, and it's a little more instructive.

As highlighted in that paper, trained cyclists typically perform with a GME of around 19-24%, meaning that of the energy metabolised, only about one-fifth to one-quarter actually ends up propelling us forward on a bike. The balance is mostly given off as waste heat, with a little energy of course needed for life support functions!

Have a quick think about that: for every watt you generate at the cranks, you are geneating around another 4 watts of heat. A rider performing longer intervals at 300W is generating somewhere in the vicinity of 1200W of heat! This is precisely why cooling is so vital for performance, as we need to dissipate that excess heat in order to continue to perform at that level.

To measure efficiency we need to measure both our energy output to the bicycle (via a power meter) and our total energy metabolised, which is done via the same respiratory gas exchange analysis equipment used to test VO2max, indeed the two factors are usually measured from the same test. Perhaps one day there will be practical and portable means to measure energy metabolised in the field but for now, the only reliable means is in the lab.

Gross mechanical efficiency can be acutely affected by things such as fatigue, hydration status, glycogen levels, environmental conditions and so on. Chronically we have an efficiency level granted to us by genetic inheritance plus however much we can manage to improve over the course of our cycling careers.

Efficiency is trainable, in particular over the long term, perhaps not to the degree of VO2max or lactate threshold, however other than by performing large volumes of training over many years, it's not totally clear whether or what specific training one can perform to achieve short term improvements. There's lots of noise from many purveyors of a fast performance gains to do with changing pedalling "techniques" or equipment choices, and while some are based on sound science and worth paying attention to, some are far more speculative, while others fall into the snake oil category.

One thought on efficiency is it's related to mix of muscle fibre types, as slower twitch fibres tend to operate with greater efficiency than their faster twitch cousins (which are better at utilising rapid energy release but less efficient metabolism), and so a fast twitch dominant sprinter is more likely to have a lower overall gross effiiency than their diesel mate. The science demonstrating the scope for chronic improvement in efficiency is a bit more limited than for VO2max and lactate threshold, and longitudinal studies are not common.

One thing efficiency is not: it isn't how you pedal, nor the way in which you apply forces to the cranks. It would really help if manufacturers of various cycle training aids would stop misusing the term - it confuses people no end.

Fractional utilisation of VO2max at threshold

This is the percentage of your VO2max you sustain when riding at your functional threshold power. It might range for example from 75% of VO2max to ~ 90% for very fit riders and is an aspect of our fitness and performance that is very trainable over the short to medium term, and can be the element of fitness we gain the greatest improvement from, but it is also something that can be developed over many years of training.

To briefly illustrate, if two riders have an FTP of say 300W, and one is doing so at 80% of their VO2max, while the other at 90% of their VO2max, the rider at 80% of VO2max has quite likely far more scope to further improve their threshold power output.

Ok, so how do all of these factors relate?

There have been suggestions VO2max and efficiency are inversely correlated, although I'm not sure how firm that relationship is, if indeed it exists across the board, or if there is a sound physiological reason why that might be the case.

The maths of the relationship is pretty straighforward though:

FTP = Energy per litre O2 (J)  x  VO2max (ml/kg/min)  x  Fractional VO2max at threshold (%)  x  GME (%)  /  60 (seconds/minute)  / 1000 (ml/litre)

Energy per litre of O2: 20,900 joules (say, refer above for details)
VO2max: 65ml/kg/min (say)
Fractional utilsation of VO2max at threshold: 80% (say)
Gross mechanical efficiency: 22% (say)

FTP = 20,900J  x  65ml/kg/min  x  0.80  x  0.22  /  60,000  =  3.98W/kg

So, now we can see that FTP is a function of those four variables, although we can reasonably assume the energy released per litre of oxygen at threshold is fixed, leaving us three variables to tinker with, and of course you can flip that equation around to ascertain any of the variables chosen given the other factors are known or assumed.

So back to the chart I posted earlier, see this example:
FTP W/kg for a fractional VO2max of 80%.
This rider has scope to improve threshold power by
increasing the fraction of VO2max they can sustain at FTP

On the vertical axis is gross efficiency, the horizontal axis VO2max, and plotted are curves representing various threshold power to body mass ratios, in steps of 0.5W/kg from 2.5W/kg through to 7.0W/kg, for a rider whose threshold power (FTP) occurs at 80% of their VO2max.

So for instance, if a rider has a VO2max of 65ml/kg/min and a GE of 22%, then we can see these intersect at around the 4.0W/kg curve. They could also have the same FTP with higher VO2max but lower efficiency (and vice versa).

If this rider managed to improve their fractional utilisation of VO2max at threshold from 80% up to 90%, then this is what happens with the very same GE and VO2max:
FTP W/kg for a fractional VO2max of 90%.
At same GE and VO2max, rider can sustain a far high power output

All of the W/kg curves have moved down and to the left. Now we can see that the same combination of GE and VO2max results in an FTP of ~4.5W/kg. 
Check the maths: 20,900 x 65 x 0.90 x 0.22 / 60,000 = 4.48W/kg.

To attain the same level of power improvement without increasing fractional VO2max utilisation, it would require an increase in GE from 22% to a little under 25% (not very likely in the short term), or alternatively an increase in their VO2max from 65ml/kg/min to 73.4ml/kg/min.

Of course, one can attain a power improvement via a combination of all three factors, although it's more likely that one will improve VO2max and/or their fractional utlisation of VO2max at threshold in the short to medium term, than attain any short to medium term improvments in gross efficiency.

So what's possible for the freaks exceptionally talented?

The magical troika of high VO2max, high fractional VO2max utilisation at threshold, and a high GE may well be exceptionally rare in the same individual, if it's possible at all, but given a GE of 25% is not exactly unheard of (higher GE values have been reported although some question the validity of those measurements) and we have seen VO2max values reported well beyond 90ml/kg/min, and very fit cyclists will have a fractional VO2max utilisation of ~90% at FTP (I'm not sure if or how much higher that may potentially go), then if we have another look at the above chart and see where a combination of 25% GE and a VO2max of 97.5ml/kg/min intersects, it's beyond the 7.0W/kg curve. It's actually at 7.6W/kg. Yikes.

Of course no-one we know has been measured to have an FTP near that level, certainly no-one without blood/oxygen-vector doping assistance, but just what is actually physiologically possible or plausible? Who really knows?

And what's possible for us mere mortals?

Well have a look at the chart, find a threshold power line near where you are, or would like to be, or perhaps a VO2max value if you happen to know it, and see what various combinations of W/kg, VO2max and GE are required. See what happens at different fractional VO2max utilisation levels.

I dislike setting limits on what's possible, but it's clear that if your highest VO2max is say 60ml/kg/min and there's limited scope for pushing that up much further, then I hate to be the bearer of bad news but you will never see an FTP of 6.0W/kg, but 4.0-4.5W/kg is definitely within reach. 

What does it mean for training?

Well while it's fun to occasionally look under the hood to see what elements of physiology we need to work on to improve our threshold power, one doesn't really need to get too hung up on these individual factors, as they are all inter-related and power measurement not only conveniently condenses the outcome for us but is the primary physiological measure of performance that matters. So keep training hard and smart. There's still no short cuts to improved fitness. There's also no real need to rush out and get your VO2max tested, the power meter will tell you most of what's important.

When you are the limit of  your current improvement in power, then perhaps it might be time to consolidate those gains, and then consider whether a change of tack is necessary to make the next step up. Do you need to give your VO2max a bit more attention, or have you still room to move in lifting your fractional VO2max utilisation? How well do you personally respond to such training?

We can gain some insight into these relationships though inspecting our power profile, and relationships between shorter and longer range power outputs, or for example, how our Functional Threshold Power and our Maximal Aerobic Power relate.

Of course a focus on training to improve one element can and does impact the others, but not always. Perhaps some additional weight loss is required. What one chooses to focus on may be different for a seasoned pro than a local club amateur, but the principles are the same.

OK, enough of that for now - it's time to close the hood, get back into the saddle, and rev the engine!

OK, here are the same charts for each of the fractional VO2max levels I mentiioned earlier:

75% of VO2max:

80% of VO2max:

85% of VO2max:

90% of VO2max:


alex field said...


I went through most of the linked articles, but couldn't seem to find where you got the general FTP equation or the 20,900J/L O2 number from? You proposed 21.12 kJ/L O2 on a cyclingforums post from 2007, have you found this to be correct for VO2max figures?



Anonymous said...


Thank's for understandable article. I have done VO2max and FTP tests. Can you please explain, how am I able to calculate my 1) Fractional utilization of VO2max at threshold and 2) Efficiency (GEM %).



Alex Simmons said...

Hi Jari

Well to do that, you just rearrange the formula I posted in the blog item:

FTP (W/kg) = Energy per litre O2 (J) x VO2max (ml/kg/min) x Fractional VO2max at threshold (%) x GME (%) / 60 (seconds/minute) / 1000 (ml/litre)

If you did a VO2max test, you should already have your GE (since VO2 is measured, and your power output is measured).

That just leaves your fractional utilisation, so rearrange the formula accordingly, using same units to give:

% fractional utilisation VO2max @ FTP = (FTP W/kg x 60000) / (20900 x VO2max x GE)

e.g. (4.0W/kg x 60000) / (20900 x 70ml/kg/min x 22%) = 74.6%

Unknown said...

Alex, need advice! Have results from lab test:
FTP: 3.4
VO2 max: 70ml/kg/min
VO2 @ FTP: 55ml/kg/min

Is fractional utilization of VO2 max then 55/70 = 78% ? If that is so, then to derive GE:

3.4 = 20,900 x 70 x .78 x GE / 60,000
3.4 = 1141140GE / 60,000
204,000 = 1141140GE
GE ~ 18%

I'm trying to see where my weakness is to focus on for making improvement. It seems like #1 is low GE followed by #2 FU @ VO2 max ?