Does anyone know why 42 days is the time frame used for CTL, why not 30 days, or for that matter 52 days? Is there any research out there that supports the 42 days?
1 Like
It takes about 6 weeks (42 days) to see changes in training.
3 Likes
Thanks for the response Gerald. Thatās my understanding as well, but as we all know, athlete response to training is very individualized. What Iām really interested is finding any research that explains how the 42 days was arrived at. My guess is that if there was any research done, the results probably showed individual variation but 42 days was the āaverageā. I would hope there is research and 42 isnāt some number somebody just pulled out of thin air. Iāve engaged Professor Google to see if there is any research out there, but not finding anything, hoping someone on the forum may have some links they can share.
Have you read this?
It gives a lot of info on how the time constants are derived and will give you enough info to keep searching.
42 is regurgitated everywhere with zero context as to where it came from and to whom it applies. 42 is for what phenotype of athlete with how much training age and etc., etc?
Clarke and Skiba have a method of determining personalized time constants and they give you a way to estimate how much taper you need before a race. Youād need to do regular time trials to develop the impulse response model.
2 Likes
CTL is the trademark of TrainingPeaks, and is set at 42days as default. I donāt have anything in research to show why itās that number. But itās a marker (42 days), so when you look at the data you will know that CTL will also be the last 42 days.
However, CTL also tells a story, over time, eg. the line increasing or decreasing will tell you if your training is building or not. Another is the slope; a gradual slope would indicate a steady increase in āfitnessā and a steep slope would indicate a possible overload and potential risk.
1 Like
I get CTL is a trademarked name. However, PMC is based on Bannisterās Impulse-Response model. In the link I attached thereās at least 6 papers mentioned that investigated the time constants.
All my point is, is that CTL, or long term fitness is not 42 for everyone and using 42 and 7 just blindly for the PMC is not good practice. However, developing personalized time constants is beyond most peopleās abilities.
3 Likes
A lot of the text in the link you posted is also showing in his book āTraining and Racing with a power meterā, so itās familiar to me.
1 Like
Thatās cool. I posted that link for OP. They seemed to want some additionaI info into where 42 comes from.
1 Like
Thank you, this is great info, Iāll need to read through it a few times to digest it.
Iāve just finished reading Manuel Sola Arjonaās book āThe Nature of Trainingā x.com, it reminded me of my post questioning why 42 days for CTL and 7 days for ATL. Manuel makes some very valid points which reinforce how much we donāt know and donāt understand and the dangers of taking a reductionist approach. I think CTL=42 days, ATL = 7 days is a perfect example of a reductionist approach and fails to consider the complexity inherent in human physiology, individual differences impacted by age, genetics, etc. Fitting all athletes in this model at best may mean they donāt achieve their full potential, at worst it increases the risk of injury and damages overall health.
5 Likes
So one of my best friends is Andrea Morelli and he has one of the best cycling/sports labs in the world (Mapei, multi-billion $ company) . He coached world tour cyclists for 20+ years and one of the top cycling scientistsā¦ Anyway he and I have discussed this 42/7 as itās more critical in the pro world then for most amateurs. So the problem is different workouts cause different levels of fatigue and individuals go through ebbs and flows so 42/7 is usually something that is changing all the time. People have done extensive research on how to optimize this number, their is even a name for it āPerformance Modeling.ā I did my own research to see if I could solve the puzzle but eventually gave up. I was trying to see if SMO2 would help give some answers. I know for me personally TSB related to a running focus and TSB related to a cycling focus are very different. I fall apart much faster running. These days (agingā¦) for me I have my own built in PMC which is sleep disruption. If I start doing too much too soon, I have sleep problems as a first sign. But yes if trying to plan volume for an event you need something more sophisticated.
4 Likes
Thanks for your reply Alex, Iām glad to know that I am not the only one questioning this. Over several months, I have read all the literature I could get my hands on related to this topic, and I have to say, I am not convinced that the 42/7 should held as the standard. I would love to see sport science researchers pick this up and investigate other options. Intuitively to me, it would seem that 42/7 may apply to certain athletes, but that other combinations apply to other athletes. I think all too often, sport science falls into the trap of āone-size fits allā and sadly, I think to the detriment of athlete potential and performance.
I fully agree on the fact that 42/7 is not applicable to everyone. But it has been accepted as a ālarge populationā parameter. And the subject is so difficult that you canāt expect a newbie to know what he should set it at.
If you have an advanced understanding of training methods or if you are a coach, you can adapt the standard settings. But I would only advice to do so after sufficient data is available and there are clear signs that this default isnāt working for the athlete.
Just like the 3/1 or 2/1 for weeks ON/OFF as a training regime, you must evaluate what works best for yourself or the athlete your coaching.
I do know that some senior athletes set the fatigue days to 8 or even 9. I, myself, have no knowledge of people changing the Fitness days.
Newbies and Proās are probably the biggest groups where this default is āwrongā given that they are on the outer side of the bell curve. For non-proās training 4 - 6 times a week with volumes up to around 15 hours, most will feel ok with these settings. But as you say, everyone is different.
2 Likes
I, 100% agree, itās a bit like 220-ageā¦ I also feel like using 42/7, HRV, monitoring your sleep and how you feel, you start to know what is going on and it becomes just one of the variables and is very broad but you know when to start paying attention. If you are -30 TSB for a week and your ankle starts hurting, instead of thinking it is pedalling technique related maybe realize you are over doing it.
I just asked Andrea if he even changes 42/7, based on previous conversations I think he gave up too. It requires too much monitoring and analysis even as a coach to do it is my belief/understanding.
But I am very interested in this and happy to discuss so if you decide you want someone to bounce ideas off, I am happy to discuss. The biggest struggle is you need direct feedback of what is going on. For example how fast do you recover? How can we adjust this 7 days figureā¦ And because these values seem fluid it will be hard to capture everything that could affect it. Itās possible that HRV in some individuals could help paint the right picture, but it would also be nice to understand performance but who wants to do an FTP test every week or more often? That is why I started using SMO2 because if you could compare power to muscle oxygen for example on a daily basis it might be easy. Heart rate has so many factors that trying to have clean data can be tough. SMO2 seems a bit hit or miss. @MedTechCD was planning on getting a lactate monitor. This is probably the ultimate solution. Compare lactate to power on a daily bases with some strictness to your protocol and you might be able to get useful data.
1 Like
It seems to me there may be more value in closely monitoring day-to-day status versus relying on the 42/7 model. Something that Iāve started doing is using DFA1 as a tool intra-workout, over several weeks Iāve been able to āget to knowā what my normal power/pace and HR is for a given DFA1 value, during a workout if I noticed that the DFA1 is lower for the same power/pace & HR than previous workouts, I back off a bit and maybe donāt go as long, if the DFA1 is staying higher, then I go a bit harder and/or longer. If DFA1 is hovering in my ānormalā range for power/pace & HR, I hold the planned intensity and duration for the planned workout. This seems to also mirror fairly closely with morning subjective scores. Iāve also been intrigued by the work done by Andreas Venhorst Dr. Andreas Venhorst, MD, PhD | sportmedizin, this is also an interesting [related] article: Fatigue and Human Performance: An Updated Framework | Sports Medicine
I have lived by your, āā¦it becomes just one of the variablesā¦ā since I started any kind of structured training, not from any wisdom just skeptical.
I find this kind of discussion between well informed and educated individuals extremely helpful. I think too many take as gospel truth what they are given because they donāt read the fine print, myself included. Seems obvious that most everything would vary by individual, but thatās not obvious at all. There is a tendency among many to use unmodified training plans for example, although they are certainly better than nothing.
Of course if these things were so simple, all you coaches wouldnāt be needed! 
I had the 2nd HRV app in the Apple App store and was a big proponent of HRV long before I started TrainerDay or all the modern tools and techniques existed. I have not tried DFA1 but assume itās another good source. I feel these days I have learned to feel it and my sleep patterns are so easily disrupted when I push ever so slightly too hard that itās almost enough in itself but I am also not very serious. Hopefully when (when not if 
) I get back to previous levels of fitness my body needs something like TSB or HRV to back up how I feel and not let me lie to myself. But yes I agree completely with what you are saying which is we can each find something that works for us that keeps us from lying to ourselves. The main good part about the TSS model is once setup you donāt need to do anything, and once you learn how it works for you it seems a pretty good indicator for me and many others even if itās not so accurate.
As some people here know, one of my best friends is a world tour coach and very scientific and while he likes coaching and coaches in general, he says coaches fall into the same patterns as cyclists self coaching which is they just have their quick little go to patterns and can totally miss the bigger picture. Obviously probably less true for a world tour coach and athlete but he still feels it happens even at that level.
Not saying coaches are bad, they can add tremendous value but someday the combination of human augmented with AI should theoretically do a lot better. Letting the human coach kind of forget about some of the details and focus on the big picture or explore nuanced ideas. All this thinking aligns with your thinking, meaning itās a never ending process of learning about yourself and finding what works for YOU.
2 Likes
Alan Couzens, in chapter 9 of his book, dives deep into how to personalize your training parameters using Python.
1 Like
I stumbled across this the other day about AI fitness coaching.
Ostensibly they reviewed ChatGPT in this study because it is one of the most robust AI systems out there and could be considered state of the art. I donāt know a lot about cycling AI systems. I presume they donāt write the entire AI system themselves, but use an existing Large Language Model and piggyback off that.
As I expected as I read through the study, AI programs are not ready to replace a coach. They can certainly write training plans, but the ability to tailor, evaluate, and update a training plan to an individual lacks precision. I have no doubt many individuals find their AI cycling programs very helpful. For them to believe that replaces a live coach would be a mistake.
The studies conclusion matches my own experience with AI in general. I find them to be very nearsighted and lacking context and background in their various evaluations and conclusions. In other words, they miss the big picture.