If training for a 70.3, I look at decoupling in each 30-40 minute repeat at race pace on the bike to see if my zones are set right. The 5% benchmark isn’t for all-out efforts, as far as I’ve been told, but sub max work (race pace intervals whose duration will generally depend on intensity and how much time you have until the race you’re training for). If it’s over 5%, you’re going too hard. Too low may mean that your strength endurance might need work.

Friel writes that decoupling should not exceed 5% for long, low intensity efforts, as long as you have the aerobic fitness to do the effort of this duration. If you’re over 5%, my understanding in that case is that you have either gone too hard (bad thing) or gone over your “time limit” (potentially good).

---

"FTP is a bit 2015, don't you think?" - Gustav Iden

Thanks.

I think in that case it wasn't what I thought it was. I was at 1.2% for 90min worth of 25mi TT pacing work. No need to pace for a run, empty the tank.

I think I'll try it out though, I have time to mess with it a bit.

Or....... foror long low intensity efforts if your over 5% you aerobic engine needs some work because that is a sign of endurance problems for that length of effort. You HR is going up to maintain an aerobic power level at the target length of time meaning you need to work on aerobic endurance.

This is interesting and not something I've given much thought before. I'd casually glance at decoupling to see if it exceeded 5% but never considered the implications of being well below 5%.

I did more riding than usual last year because I had no races or pool access. A lot of the riding was aerobic volume and I did some bigger challenge rides to see where I was at. For example:

• first/failed Everesting attempt (mid August): 5:20 of riding with an IF=0.79, Pw:Hr = 7.24%
I ultimately bailed on this attempt because I was overheating and had a mechanical issue, but maybe part of the overheating was the fact that I was going too hard for too long.

• second/successful Everesting (early Sept): 14:30 of riding with an IF=0.67, Pw:Hr = -0.93%
Did this one on a more consistent climb with a shallower gradient, hence the lower intensity factor and long ride time. Felt solid throughout the ride and basically zero decoupling. Seems like aerobic fitness was fine, could suggest strength was a limiter(?)

• this past Feb I did the PRL Full on Zwift: 5:30 of riding, IF=0.73, Pw:Hr = 1.42%, all seems good for this duration/intensity.

• a week later I challenged myself to 200 Zwift miles: 9:30 of riding, IF=0.66, Pw:Hr = -2.65%, if I'm understanding your post correctly this negative decoupling again suggests I was fine aerobically but might benefit from more strength work?

Don't know if I'm thinking about this/interpreting it correctly, but it's definitely got me thinking and curious.

It's just a way of establishing aerobic threshold. Not dissimilar to MAF, zone 2 HR etc. The idea was to warm up to target HR or pace and then continue for 1 hour. If the HR had to drift to maintain pace, or the pace decreases at the same HR by more than 5% the conclusion is you were past your aerobic threshold. The idea is you repeat the test untill you find a pace:hr decoupling of 5% or just under over an hour and use this as your aerobic threshold. It's useful for those following polarised or pyramidal training approach where 80%+ of training is supposed to be below aerobic threshold as they can use this pace/HR number as their upper limit for training.

Also once you have your aerobic threshold HR/pace you can test your lactate threshold HR/pace. If there is greater than 10% difference between the two you are said to have aerobic deficiency syndrome (ADS) meaning your aerobic system is lagging behind anaerobic system.

The guys that wrote training for the uphill athlete are big on it. If you Google "uphill athlete aet test" and "uphill athlete aerobic deficiency syndrome" you can get plenty of info. Remember they are focused on high altitude mountaineers/hikers/ski tourers/ultra runners who's demands are different to the average triathletes though. You may well be better served just using power meter and basing zones of FTP, and just using decoupling as a tool for corroborating your aerobic power zone is set correctly.

IIRC it measures the difference between average pace or power and average heart rate for halves of an activity or selection.

for a 60 min run it would look at average pace for 1st 30 mins and average heart rate for those same 30
mins and compare to the second half. if average pace is the same but average heart rate is higher then some decoupling occurred.

i try to clarify the ratio since even though it takes average heart rate which would be a “steadier” metric it still can be affected by temperature, humidity, hydration, elevation gain or even how well an athlete slept. so it’s not always black and white regarding if a pa:hr or pwr:hr is over 5% then one has bad aerobic conditioning.

also average heart rate can remain the same in second half but average pace or power be lower.

friel does suggest using it for steady long runs or rides. however there is some benefit to looking at runs that could be higher “intensity”
as half marathons or 10ks. same goes for runs where you would do lets say 60 min easy and 60 min moderate, even though pace changes if the ration of change is the same then decoupling is low.

what i do not use much is when the resulting number is negative. most times it is more of inaccurate hrm readings or not very clear relationship between pace and heart rate ie elevation gain or loss.

The way I understand Pw:Hr, it's not that meaningful over stochastic efforts like your Everesting attempt. Your pace/power should be more or less even over the analyzed period, e.g. 1.5 hours of easy running, or 40' at 70.3 race pace, or 15' at Olympic race pace. If I look at aerobic decoupling over an entire workout full of intervals, the numbers can be pretty random. They start making sense when I look into each repeat separately (but only a relatively long one, i.e. Pw:Hr in a 2-minute VO2max effort will always be crazy high).

---

"FTP is a bit 2015, don't you think?" - Gustav Iden

Update:

I did my "burnout" workout yesterday at lunch. What you've got for 20min, what you've got for 10min, what you've got for 5min. It really sucks, but accumulates a lot of time with a lot of lactate, high HR, and mental fortitude to suffer it out for time trial.

I did manage to get the HR to decouple the effort in the 20min and the 5min. I did not rest enough after the 20min effort and the 10min effort was sub-optimal and didn't decouple.

Poor route planning as my turnaround loop didn't take long enough after the 20min hard effort. 3min simply isn't enough.

But.......it did give me some pacing info to work with. Now I have a data point of "more than that one day" but "less than that 20min effort". As in, I think I can pick a 25mi power target now within 5 to 7w. I'll negative split it on the low end then up the ante in the end if possible.

So:
-80min at 260......too low (-1%)
-20min at 275......too high (4.5%)
-guess at 60min at 265......reassess at 35min into 25mi

Also, on reassessment, that 80min had a meaningful chunk of out of saddle hill riding during the effort. At which I can make like 20w more power than TT. That probably tossed that figure some.

So, despite it not being used as intended it seems I can glean some meaningful info.

Yep, looks like 265W will be close to your aerobic threshold where you can exercise for an hour with ≤5% decoupling.

As you say, you need a flat course as any hills will skew the data.

I guess the question is how do you plan to use this? You have the disadvantages associated with HR and seemingly little advantage of just using regular power zones. As a test every month or so to see if Aerobic threshold is increasing it has some merit. For actual training I'm not sure how useful it is other than capping training intensity.

Only for one and done race effort really this time around. Nothing useful otherwise. Otherwise I train by power. I do a Zwift climb or a real life TT to get a baseline for power based training.

I don't have the budget to spend too much exertion on getting pacing perfect by doing a ton of race day level efforts given the recovery necessary there. An hour is a lot to ask of my body all-out more than once in a while. 20min is more reasonable, or a 10mi TT is just over that if fast course.

So if I have the decouple for 20min and the value for some long/hard sweetspot at 80 to 90min...........I can get pretty close is my calculus here.

The 90min sweetspot doesn't knacker me nearly as bad as the true all out hour. Which is miserable for even battle hardened pros, much less a hobbyist like me.

I don't think there is much to be gained from looking at decoupling at high intensity or shorter duration? Does it really tell you anything? If you are at race pace you would expect way more than 5% decoupling as you will be past aerobic threshold.

Check out this site https://www.flammerouge.je/.../faqs_decoupling.htm. You might like some of his fact sheets on TTing as well. guy is really knowledgeable.

Best of luck

Don

There are a couple mentions about how elevation gain/loss may negatively affect the usefulness of aerobic decoupling figures - and at least from what I can tell of TrainingPeaks's documentation (see Efficiency Factor and Decoupling and Normalized Graded Pace) they seem to take into account elevation changes when calculating EF (which is used for decoupling calcs) by calculating NGP.

Perhaps there is still a kind of second-order/secondary effect, if the elevation changes are highly variable rather than relatively constant, but at least to some degree elevation changes (and other factors as noted in the linked documentation) seem to be normalized by NGP. Effect like this is probably diminished by having the same/similar changes in the first and second half of a session, such as an out-and-back run - but I'm guessing a bit here.

Please others chime in here if you're reading this differently!

FWIW I also enjoyed reading Triathlon 2.0 from Jim Vance when it comes to how to use these measures (and others) as information to adapt training.

The biggest problem with decoupling is that its tied to heart rate...so, everything that's "bad" about heart rate based training is bad about decoupling as a metric. Just like with basic heart-rate variation from day to day....coupling is even worse because your heart-rate response over time is a function of many of those same factors. So, what you might see as "decoupling" might simply be an accumulation of heat-stress, or dehydration, or you didn't get enough sleep last night, or you had a hard workout two days ago, or, or, or....These issues make it difficult to make practical use of, unless you control for them.

Assuming you can control for those variables...decoupling is specifically for steady state, sub-threshold, efforts. The effort level itself irrelevant. A decoupling value over 5-ish% or so, is simply saying that effort was a harder effort relative to your current fitness. By harder, I mean higher-training-load. Maybe that's a 3 hour z2 ride, or a 90min z3+, or a 45min z4. The idea of decoupling is that you are exhausting your slowtwitch fibers and starting to recruit more fast-twitch fibers which requires more blood flow to more places.

For the most part, decoupling should just be expressing your training plan. The whole point of a training plan is to push the envelope somehow. So, if your training plan is appropriate for your current fitness you should see higher decoupling figures a couple times a week. Decoupling is just another way to look at duration for a specific intensity. So, if you are going out for an endurance ride, then based on the decoupling theory you should ride at z2 until you reach ~5% decoupling. Ie, its a measure of "training effect". Bumping the intensity is the wrong solution for driving higher decoupling...doing that changes the basic character of the workout/stimulus. If I bump a z2 ride up to z3 to get the 5% magic number then I'm doing a different ride with different benefits.

Also, >5% decoupling is a pretty taxing workout---especially, if you are already pretty well trained. In general, that's going to be a workout that's beyond your current max. So, if you can do ~3hr z2 @1% decoupling, it might take 4+ hours to get to the 5% magic number. I'm typically "spent" after a 5+% effort, and often need a day off.

All that is to say, I don't find it very useful because its confounded by environmental factors, and there are other more effective means for planning workout durations.

ETA: I decided to look back at my own training log. I took Jan - Mar completely off....COMPLETELY. I restarted running again in late March. All my long runs since:

10 Apr 4.5mi 1.47%
17 Apr 5.6mi 2.58%
24 Apr 4.6mi 4.89%
01 May 6.8mi 4.89%
08 May 7.5mi 8.20% <--- HOT/Humid AF
15 May 8.1mi 6.71% <--- Still Warm/Humid, but less
22 May 9.3mi 2.62% <--- acclimating to the conditions

Thanks for the response Tom! Would you mind picking up at the end of your post, and sharing a couple of the other top alternatives? Always looking to up my training planning game!

Travis

Sorry, I didn't respond quicker....its been a busy week.

To be honest: What's wrong with plain old duration? And making adjustments based on how you feel/respond to the training? I mean, that is what stimulates adaptation. I will also use TSS, since I'm a training peaks kinda guy.

My own process? I make a general plan using TSS to outline the plan, and divide based on sport priorities (if I'm multi-sporting it). Then I will select workouts of different intensities based on my goals for a block of training. The intensity and TSS will derive a duration. Once I've got a base week established using this approach, I tend to copy the week and increase the durations each week using my best judgement. I know from past experience that I can add ~5-10 minutes to a long run, or ~15-20 minutes to a long bike.... each week.

I assess the upcoming week based on how I feel on day1. I follow a 6 day schedule with on day completely off (Sunday). Saturday is my long day. Monday is my "day 1"...and is usually planned as a harder day of some kind...might be 2x20 on the bike, or a hard fartlek run, or some such. If my legs are still tired from Saturday, such that I can't hit the goals of the hard run/bike...then I know I need to maintain for the week (not increase durations / repeats / etc...or even back down a smidge).

At the start of a new plan (after a period of time off), it takes me about 4 weeks for all the "fatigue" to sink in, and for training to stabilize. So, I will tend to plan the first 4 weeks on paper per the above concept. Then, I begin adjusting from there.

As indicated above, I do keep my eye on Pa:Hr or Pwr:Hr. But, I don't usually do much with it. You really have to bounce the number off the situation to decide if it should be "trusted". I really find it doesn't tell me anything I didn't already know. If the long ride seemed easy, I check Pwr:Hr and its 1.05%...yep it was too short. Again, I already knew that. If the ride seemed easy, but my Pwr:Hr was 8%....why? Oh, it was hot out and I got dehydrated. Fine, drink more next week and ignore the decoupling (its bogus). Or....I'm a "good tired", the day was nice...I check Pwr:Hr and its 4.89%. Just right. Increase by ~15 minutes next week.

I'm not doing anything with the data. I'm just confirming it matches (or doesn't) with other indicators. But, its those OTHER indicators that I use to adjust my training.

I've NEVER had decoupling tell me I went too hard (or easy), when my body told me differently....and the decoupling turned out to be right two days later.

I don't know if that makes sense. Let me know....