Functional Threshold Power
I first heard of the acronym FTP maybe 10 years ago. At the time I was deep into coaching CrossFit, not so much into trying to completely understand what was going on physiologically with the athletes. I was mainly concerned with the performance aspect. At the time, I thought I had real knowledge. Looking back it was only cursory. When I had heard FTP mentioned or read about it online somewhere, I just skipped over and continued on.
My next contact with FTP was in listening to a video from the great Mark Burnley. This was a few years ago, maybe 2020? I can’t say for sure. I bet he would hate me referring to him as great, but I really think he is a great researcher, one of the best in his field for sure.
In his video, he made a quip to the effect of (paraphrasing) “I was asked to mention FTP, so now I’ve done it.” The reason I was watching anything from Mark Burnley or reading his research was to understand the power-duration relationship. So, I thought it was an interesting comment and again, I just moved on.
I was next reminded about FTP in a book by Philip Skiba. If you are a fitness coach you need this book! There is a short discussion about FTP in the early chapters, where he basically says it has its uses and is practical. Again, nothing definitive. But, both Burnley and Skiba were less than enthusiastic to recommend its use. So, naturally, I was interested.
There’s been a lot written about FTP. Recently, I have been seeing it more and more on my social media feed. Not a lot, but just more than in the past. For me, when I find something I am interested in I need to understand where it’s coming from, what is its origin.
In the research literature, the continued reference for the origin of FTP is from a book by Allen and Coggan about using power meters in cycling (1). Well, I believe they were the first to put a name to this idea. Although, others have discussed a similar idea (using one hour average power as a correlate to performance), but just never named it (2).
What is FTP?
Let’s look to Training Peaks to give us a definition. From author Joe Friel, “Functional Threshold Power is defined as the greatest mean maximal power you can currently produce for one hour” (3). The one thing not mentioned, which I guess is implied but readers of this post may not know is that this test is performed on a cycle ergometer of some sort.
How do you test for FTP? First, find a bike with a power meter. “The most common field test for finding FTP is a 20-minute, steady-state ride done at maximal effort. It should be treated as a race, so be rested for it… Once you’ve finished the 20-minute field test and downloaded the data look to see what your average power was. Subtract 5% from it and you have an estimate of what your FTP currently is.” (3).
The 20 minute test (using 95%) is the recommended approach due to ease of testing, versus doing the full 1 hour time trial which was the original idea. Joe Friel makes the case that the 1 hour test requires too much motivation for a training scenario (3). An interesting observation.
OK, let’s just say an individual holds 200 Watts as an average for 20 minutes on this test. 95% of this would be 190 Watts. So the FTP estimate in this scenario would be 190 Watts.
Again, I must emphasize, this is for cycling. I have not yet read any literature mentioning the value of this test using other modalities. It may exist, but I haven’t read it. We’ll cover that issue later.
What do I like about FTP?
As Skiba mentioned, it is indeed very practical, which I love. As a fitness coach, most of what you are doing has to be extremely practical because coaching and training design are time sensitive. FTP is relatively easy to collect and the results are readily implementable.
On the contrary, taking blood lactate samples and or measuring VO2 Max on a particular modality is not practical in any way for fitness coaches. Therefore, these are rarely done in fitness gyms around the world. And by rarely, I mean never. So by comparison, FTP is extremely practical..
What I love most about people using FTP is that they are showing curiosity in their own physiology and their performance. It means they are trying to figure something out. They are trying to best navigate their own scenario. Trying to optimize.Trying to create knowledge and then make use of that knowledge. This is what the basis of improvements in training programs is about. Admirable.
What Is The Problem With FTP?
1) Is It A Threshold?
The word “Threshold” appears in the name FTP. So lets start here. Is FTP an actual threshold? A dictionary definition of Threshold yields the following:
“the magnitude or intensity that must be exceeded for a certain reaction, phenomenon, result, or condition to occur or be manifested”.
In this case, the magnitude or intensity would be the power output while cycling. Therefore, FTP is claiming that something readily observable physiologically will occur above this power output versus what will occur below this power output. In the hypothetical scenario, I presented earlier, if the FTP is 190 Watts, then what happens at 200 Watts should be different from what happens at 180 Watts.
Sounds clear enough. But, is that actually true?
A recent study (4) looked at just this question. They investigated to see if FTP was indeed, the maximal metabolic steady state. If FTP is a good estimate of this, then it will represent the highest power output one could sustain and still achieve steady state behaviour in numerous variables (ie – VO2).
This study recruited 13 cyclists, ranging in ability from untrained to well-trained. FTP was assessed using the standard 20 minute maximal effort self-paced time trial. On a separate visit (crossover design) they had participants perform a time to exhaustion at the estimated FTP (95% of 20 Min Average Watts). An additional visit to the lab was to have the participants perform another time to exhaustion trial at FTP +15 Watts. The group average for FTP and FTP +15 were 222 Watts and 237 Watts respectively.
What were the relevant findings?
From the paper (4), “the key findings were that VO2 peak was not reached at both intensities”, and that “VO2 stabilized at both intensities”. So what does this mean? The authors then state “the present study also demonstrated the FTP and FTP +15 are within the heavy intensity domain and, therefore, should not be used to represent the physiological threshold between the heavy and severe intensity domains”. Long story short, FTP in this and many other studies has not been shown to be a good estimate of the maximal metabolic steady state, and therefore does not really fit the description of the threshold it seeks to be.
(screenshot taken from Wong et al.)
If you look at that image and are not sure what to look for, if FTP were indeed an significant threshold, the line with the white dots (FTP +15W Group) would continually go up as the duration of the time trial goes on. Meaning, the trend would be towards VO2 Max (FTP +15 reached approximately 90% VO2 Max on average). As you can see, it does not do that, it basically follows the exact same pattern as the black dot line (FTP Group). To quote the authors once again, “FTP and FTP +15W are the same intensity”. Meaning, they effectively will have the same physiological response. They each reside within the heavy intensity domain. Obviously, exercising 15 Watts harder for a time to exhaustion test will result in earlier failure, but that doesn’t equate to evidence for it being a threshold.
Then then plainly state, “the intensity corresponding to the FTP is not appropriate to use as a reference for designing training programs as it has a high chance of overestimating the VO2 response” (4). Ouch!
2) Does It Reasonably Predict 60 Min Average Power?
Now, let’s look to see if FTP does indeed represent what it portends to… to represent a 60 minute average power, as that was the original intention of FTP.
The test has now been shortened to 20 minutes using the 95% adjustment, but the intention is that it does represent the 60 minute average power.
Looking at the same study (4) “only seven out of the 13 participants were able to sustain exercise at FTP for 40 minutes”. The mean time to task failure at FTP was 33 +/- 7.6 mins. In this case, the 20 minute FTP test using the 95% adjustment was not a good estimate of a one hour performance.
A different study from Sitko et al. (5) aimed to resolve the same issue. Is FTP, established via the 20 min test, a good estimate of one hour performance? This study was much larger in terms of participants, 87 to be exact. The participants were all trained cyclists, of varying degrees. I do believe all 87 were Men. Ranging from recreational level (VO2 Max Group Average of 46.9 ml/min/kg) up to Professional (VO2 Max Group Average of 74.3 ml/min/kg).
One of the main methodological differences from the previous study (4) was that this one had a more robust warm-up section prior to the 20 minute test. In theory, this would maybe lower the FTP power estimate as compared to the previous study in which the warm-up was simply 5 minutes of easy cycling (100 Watts). Regardless, once FTP was established for the 87 participants, they came back on a different day, and rode at the estimated FTP until task failure.
The various groups had the following FTP estimates and accompanying time to exhaustion:
Recreational – 235 Watts/35 Min TTE
Trained – 266 Watts/42 Min TTE
Well Trained – 279 Watts/ 47 Min TTE
Professional – 308 Watts/51 Min TTE
(taken from Sitko et al.)
As you can see from the image, there are indeed many individuals that did make it to the 60 minute mark or longer, and on a percentage basis, this goes up as the training level of the group goes up. But if you look to the left, the recreationally trained, nobody made it to 60 minutes. And with regards to the people reading this article, you yourself are likely at around this level or below. And for coaches, most people you coach will be at around this level or below.
From the study, “it is recommended that FTP and time to exhaustion are assessed, and reported independently for each subject. Lastly, cyclist performance level and experience should be factored in when attempting to study time to exhaustion, as higher performing and more experienced cyclist consistently show longer time to exhaustion at FTP” (5). Meaning, the FTP estimated from the 20 minute test should not be used as a proxy for 60 minute average power. Additionally, one must also consider the level of experience of the rider.
Anecdotally, in a recent social media conversation I had with a Exercise Physiology Researcher (PhD project is related to cyclists) he pointed me to a new paper in which 17 of 19 riders made it to the 60 min mark. These were indeed “competitive” cyclists, so high to very high level ability. One thing to note from our exchange, is that the data he was reporting was blood lactate levels over the 60 minute duration amongst the riders. And to quote him, he said “consistent with heavy domain”. Which is consistent with what I have cited above.
Moving on…
3) Applicability
Here is what I see as the MAIN limitation for using FTP or even thinking in those terms as a fitness coach. It is not an applicable model in ANY OTHER SCENARIO outside of cycling. Its range of use is so extremely narrow as to render itself nearly useless to you. You need a better model, see below.
Here’s one example. The study from Wong (4) was very elegantly designed. Their intention was to show that even working above FTP did not elicit severe intensity domain behaviour. If you don’t know what severe intensity domain behaviour is, that’s the main problem in and of itself with FTP. Because intensity domains are based on real exercise physiology. FTP isn’t.
FTP is to exercise physiology what “blocks” are to nutrition. A great way of getting started. However, when other known and well described options are available (i.e. macronutrients), continued use of such imprecise metrics should indeed be abandoned by the serious practitioner.
You yourself may respond, but Michael, my FTP works perfectly for me. If I happen to go a little bit faster than FTP then I will fail quickly. If I go a little bit slower than FTP I can go for a very long time. That is fantastic and I’m happy for you. But that’s happening not because it’s FTP, it’s happening because you are lucky enough that your FTP score is the exact same as your critical power (which is the real thing).
FTP is a descent can opener. Critical Power/W’ is a swiss army knife.
A Crash Course In Critical Power
What is Critical Power?
The concept was developed in the 60’s (6). However, it has made great strides since then. “Critical Power is now considered to represent the greatest metabolic rate that results in ‘wholly-oxidative’ energy provision, where wholly-oxidative considers the active organism in toto and means that energy supply through substrate-level phosphorylation reaches a steady-state, and that there is no progressive accumulation of blood lactate or breakdown of intramuscular phosphocreatine (PCr) i.e., the rate of lactate production in active muscle is matched by its rate of clearance in muscle and other tissues” (7). Critical power or critical speed or critical torque best approximate the limits of sustainability in human activity.
How do you establish it?
Almost like you would for FTP. There are really 2 main options available to you, the first option (if you’re a sick puppy) is to do the 3 minute ALL OUT test (8). The next option is to do to time trials between 3 and 15-20 minutes and then do some math.
Is Critical Power mainly for biking, like FTP?
No, it is observable in movements involving small muscle groups and movements involving large muscle groups (9). It isbservable with isometric testing, normal concentric/eccentric motions like bicep curls and leg extensions. Running, rowing, skiing, biking, etc. Pretty much do any movement you can think of. And, if you’re a decent detective, you’ll see it everywhere. Oh, and it is likely ubiquitous across the animal kingdom (10) in case you want to test your dog on the treadmill.
Is it precise?
“When the athletes ran at a speed that was ~6 s per km slower than Critical Speed, VO2 was in steady state and the prescribed 30 min of exercise was completed, whereas when the athletes ran at a speed that was ~5 s per km faster than CS, a VO2 steady state could not be achieved and exercise tolerance was limited to ~17 min, indicative of exercise within the severe-intensity domain. These data indicate that Critical Speed provides a rather precise demarcation of the highest running speed at which VO2 can be stabilised. This observation is consistent with several previous studies which indicate that Critical Power or Critical Speed is the metabolic threshold which partitions severe-intensity exercise, which, by definition, is characterised by an inexorable increase in VO2 to its peak value at the limit of tolerance, from heavy-intensity exercise, during which a VO2 steady state can still be achieved“ (11).
Another extremely important aspect of Critical Power research is that papers readily state the standard error associated with the measure (their confidence). So, if the critical power estimate is 195 Watts, it will generally be reported as critical power can expect to fall between 190 and 200 Watts or similar. It is extremely important to acknowledge when things are not exactly precise. Or, if they can’t even be exactly precise. But what the research will also show you is if you are working 10% above the estimated critical power then you will achieve a VO2 max and task failure. The same cannot be said for working 10% above FTP, as evidenced.
Lastly, it must be mentioned that critical power (critical speed, critical torque, etc.) is not a specific point on the power spectrum, say 195 Watts. It is more likely that it represents a narrow range of power output in which a phase transition from sustainable metabolic rates to unsustainable metabolic rates will occur (12).
Summary:
1 – FTP is not really a threshold of any real sort.
2 – the recommended FTP test does not fulfil the original obligation of estimating 1 hour power.
3 – the utility is so narrow as to render it useless for fitness coaches.
Tis better to err in the pursuit of truth than to err in the pursuit of fiction.
—Michael
Reference:
3 – Analyzing Functional Threshold Power in TrainingPeaks
6 – Monod and Scherrer (1965). The Work Capacity of a Synergistic Muscular Group.
7 – Poole et al. (2016). Critical Power: An Important Fatigue Threshold in Exercise Physiology.
8 – Vanhatalo et al. (2006). Determination of Critical Power Using a 3-min All-out Cycling Test
10 – Burnley (2023). Invited review: The speed-duration relationship across the animal kingdom.
great article Mike!