Hypertrophy And Effective Reps - Incomplete Model?

[mikeynov]

Super smart dude Greg Nuckols made a post over at exodus barbell on the topic of hypertrophy models and effective reps that I thought was worth considering:

https://www.exodus-strength.com/forum/viewtopic.php?f=3&t=2928&start=40

I think this whole thing is a case where models can be useful without being strictly true. I think most people would agree that you need to a) recruit all or most MUs and b) induce some meaningful amount of local fatigue (i.e. not just doing 1RMs) if you want to maximize hypertrophy. Beyond that, you're just trying to come up with a model that more-or-less matches the experimental findings we have. I think the idea of "effective reps" puts you in the right direction, but it doesn't fit all the evidence we have (in Helms' dissertation study, growth was pretty similar between two groups stopping each set different distances from failure; in a recent study by Carroll et al out of ETSU, a group stopping further from failure grew substantially more than a group reaching failure each session. Then there's that old Sampson paper where growth was similar between failure and ~2 reps shy of failure). And then there's also the issue of model fitting - are we sure the hypertrophy we're seeing in the research is the hypertrophy we "should" be seeing? Do we see more growth with failure in a lot of studies simply because the subjects have more local edema? Is there possibly more sarcoplasmic hypertrophy going on with failure training (that's never been studied)?

I personally don't think we're yet at a point where we know enough to have too much confidence in any specific predictive model being "true," but effective reps isn't necessarily a bad lense to look through.

In my previous hypertrophy thread, we discussed the concept of "effective reps" that I first saw introduced by Borge/Blade. Basically the idea that the reps that really "count" in a set seem to be the ones at the end where you are in a state of maximal recruitment, particularly in sets of ~5+. So whether it's a set of 12 or 30, you only get something like 5 "effective" reps per set. Chris Beardlsey talks about the idea at length here:

https://medium.com/@SandCResearch/h...iding-failure-and-using-advanced-90e26d57bca9

What's interesting about Greg's note above is that "effective reps" doesn't seem to be consistent with the fact that training to failure does NOT actually generate more growth per set. One of the consistent findings is that sets that are even near-ish failure, like 2+ reps away, seem to generate just as much hypertrophy, usually. If there are only 5 effective reps per set as per the Beardsley thinking of effective reps, then missing something like half of those should very much NOT result in identical hypertrophy. But it seems to all the same.

Ron and I were previously speculating that sufficient work at high levels of recruitment and rate coding are necessary to grow optimally, but I'm not sure how to reconcile the rate coding model with this data either. Meaning that not all the motor units' rate coding are maxed out in submaximal sets that are in at least the ballpark of failure (probably still full recruitment), but still they seem to generate as much hypertrophy.

I was reminded of the old idea of growth being turned "on" ala HIT thinking, but even that's not quite correct. Because it's not that growth is being turned "on" so much as growth is being maxed out under certain conditions, and if those conditions are met, then you've done all you can in that set to grow.

So, the hypertrophic potential of a set is only maxed out when two things happen:

1) Sufficient total time/reps/fatigue
2) Sufficient levels of recruitment occur

For #1, we appear to need to do > ~5-6 reps per set. For #2, we need to reach a condition of full recruitment, either by being heavy enough or close enough to failure.

So testing this model, doing singles or triples satisfies #2, but not #1. So we'd predict that their growth potential would not be maxed out, and as per the research, this appears to be the case, though you can make up for it by doing a shitload of sets. Similarly, if we do really light weights but aren't very close to failure, we've satisfied #1, but not #2, and again, we'd predict suboptimal growth. Interestingly, like the singles example above, we can probably make up for it as long as we do enough volume, and there is some research where they still get pretty good growth doing a bunch of sets of 10 at ~60% 1 RM, for example, which is probably something like a ~15-20 RM, never really hitting full recruitment.

Anyhoo, I just thought these thoughts were worth sharing, as Greg's comments made me reconsider the arguments we had back and forth in that other thread I made on the necessity of load progression in hypertrophy. The problem I see vs. the other thread is that we were sort of concluding that growth was something tantamount to the amount of total work being done near failure, but the problem with this would be that submaximal sets seem to consistently generate as good of growth as sets to failure as long as the two criteria above are filled.

Any thoughts would be greatly appreciated. Ron, you out there bro?

 

[Old and Grey]

I think a lot of it has to do with where you are in your training career. Beginners, intermediates and even early advanced lifters can grow with sub maximal training. Hard core long-time trainees, I believe, need to be close to failure a good part of the time to induce hypertrophy and Myo reps, as set forth by Borge, is a way to achieve that within the shortest amount of training time. It can also be achieved using regular training methods such as a 3 minute rest period between sets but you are going to have to perform many more reps to achieve the same level of recruitment and the total training time may be creating a negative impact on your central nervous system that may actually inhibit or even reverse growth. My belief is based on anecdotal experience only. After 60 years of lifting, I thought I was pretty much maxed out without abusing steroids. However, 9 months after a continuous program of Myo rep workouts, the difference, at least to me, is astounding. Again, I do believe that this type of program is limited in its usefulness to those that are solidly advanced, i.e., 5-10 years of serious lifting.

 

[NWlifter]

Interesting post Mike..
First, zip over to your other thread, I posted some interesting info. on that. (rate coding, fiber tension, etc.)

Second,.. effective reps.. I think that term is misleading, it makes one think some aren't and some are. But really , it's more effectiveNESS of a rep. Every rep is effective. The more you do in a row, the more motor units that get to have effective reps. Rep 5 out of 10 is still very effective for many fibers, just not all fibers. The last rep of a set might not even be effective for the last recruited MU's. It might just not be enough time at a decent activation level.

What we need to REALLY remember, is fibers are individual entities. A muscle doesn't grow, fibers grow in the muscle. So we can stimulate some and not others. We can have 500 fibers grow and 400 atrophy, and only 'net' 100 fiber's worth of growth in the whole muscle.

So... do 10 sets of 4-5 reps with a 10-12RM, we might get hypertrophy in 1/2 the fibers, they are getting 50 'good reps' where the rest of the fibers are getting no good reps and getting smaller.
In that light, the last reps are 'better', not because the first reps did nothing for any fiber, but because it's the only place where the last or way later recruited motor units are 'on' enough to get fatigued and be stimulated.

Another interesting thing...
(can't remember if I posted this in that other thread)

Going to failure can be really different for various people.

https://pdfs.semanticscholar.org/7af3/b2ac5f8b1f8778ebab2bdd85a53f4b1f507e.pdf
Some people do get central fatigue and lose force with high effort ‘from’ that, some people can keep their neural output high. So high effort IS different between people. 6 out of 9 had central fatigue in 60 seconds. 3 did not.

This means people who 'can' keep the 'heat on' for the whole set, probably can really get a lot of stimulation in those muscles in that one set (failure IS from fiber force loss), those that feel like they are trying hard, but actually are losing neural force, appear the same, but actually are not pushing their muscles at the end of the set.(force drop is from neural decreases as they fatigue)

 

[NWlifter]

Any thoughts would be greatly appreciated. Ron, you out there bro?

I'm here, where'd you go?

 

[mikeynov]

Good thoughts as per usual Ron.

I agree in principle about some fibers growing and not others, but I'm not sure how practically true this is, and I still think back to some of Bryan's comments on the subject. I.e. the idea of the tension of one fiber affecting neighbors etc. Under normal lifting conditions with these high levels of force it seems hard to imagine some fibers growing and others literally atrophying. Not that that wouldn't happen, just that it's hard to envision.

In terms of central limiters to failure for some people, that is pretty interesting. Is there other/newer research on this subject? For me the exercise that comes to mind here is actually squats. For the entirety of my lifting career, squats were always one of the weirdest lifts in terms of what I could perform. Like if my lower back was fatigued/I was drained, I could cut my reps in half at a given weight vs. a good day. I remember at one point even testing my leg strength after a terrible squat session a decade+ ago, when I just couldn't push shit in the gym, and I was borderline PR-ing leg press after. So the problem clearly wasn't my legs per se.

This then reminds me of Jester's comments on why we fail in that other thread, and you start to see the appeal of Arthur Jones like thinking of directing as much stress to muscles as possible in comparatively low skill environment (e.g. machines). Though, as I said to him in that thread, I've never seen research showing a superior effect of something like a leg press vs. a squat where the latter clearly has a higher skill component that might be more sensitive to central limitations along with the obvious weak link in the kinetic chain of the lower back. So, I dunno.

Anyways, just some rambling thoughts.

 

[NWlifter]

Good thoughts as per usual Ron.

I agree in principle about some fibers growing and not others, but I'm not sure how practically true this is, and I still think back to some of Bryan's comments on the subject. I.e. the idea of the tension of one fiber affecting neighbors etc. Under normal lifting conditions with these high levels of force it seems hard to imagine some fibers growing and others literally atrophying. Not that that wouldn't happen, just that it's hard to envision.

they do see that in some cases, like with endurance training, ST fibers will grow, they are the only ones activated enough to experience tension. It's why we have to activate and fully recruit, otherwise, any not recruited or 'barely recruited' wouldn't experience enough work and tension (and higher intracellular calcium) to start the signal cascade.

In terms of central limiters to failure for some people, that is pretty interesting. Is there other/newer research on this subject? For me the exercise that comes to mind here is actually squats. For the entirety of my lifting career, squats were always one of the weirdest lifts in terms of what I could perform. Like if my lower back was fatigued/I was drained, I could cut my reps in half at a given weight vs. a good day. I remember at one point even testing my leg strength after a terrible squat session a decade+ ago, when I just couldn't push shit in the gym, and I was borderline PR-ing leg press after. So the problem clearly wasn't my legs per se.

I think that study had other references at the bottom, I know Enoka spoke of it in Neuromechanics of Human Movement. I feel that with all exercises and thought I was weird, now I see it's not uncommon, where if I go to failure, as I approach it, I have to REALLY mentally try hard and can get a 'burst' of force, but it almost kills me, even with small exercises.

This then reminds me of Jester's comments on why we fail in that other thread, and you start to see the appeal of Arthur Jones like thinking of directing as much stress to muscles as possible in comparatively low skill environment (e.g. machines). Though, as I said to him in that thread, I've never seen research showing a superior effect of something like a leg press vs. a squat where the latter clearly has a higher skill component that might be more sensitive to central limitations along with the obvious weak link in the kinetic chain of the lower back. So, I dunno.

Anyways, just some rambling thoughts.

Sure, good thoughts!
I think for people like me (the 6 out of 9 types), going to failure just pushes the neural stuff harder with none to barely any muscle benefit. But some seem to really get some good muscle stimulation from it. It kinda fits what we see too, seems like there are fewer people who do really grow with that 1 set to failure stuff. Most seem to get more strength, less size and burn out in a fe weeks. I bet that is why.

 

[mikeynov]

On the one set to failure thing, I've had an idea in mind I'd like to see tested in research at some point. In fact, if you know of any examples of something like this being tested, I'd love to see it as I'm not sure I've come across it.

I'm not sure how many bodybuilders you've known, but one of the weird trends I've noticed in the actual bodybuilding crowd is that a lot of bodybuilders train with a ramping style that reminds me a bit of how Dorian Yates trained. Which is to say when they do an exercise for "3-4 sets of 8-12," what they're actually doing is a series of sets in which the weights ramp up set to set, terminating in one, all out effort.

So a classic HIT set of bench might be 1 sets of 185 for 8-12 reps or something after warmup. But done in the above manner, it might be 145 x 8, 165 x 8, 185 x 8-12. What's interesting is that the RPE is low enough on the "working warmup" sets that, past getting used to the conditioning of more total work, they're not REALLY eating into the top set too badly performance-wise. But they also feel like they're doing something, i.e. the cumulative effect of those ramping sets + the all out set, at least based on my own experimentation, sure feels like more cumulative stimulus than just skipping to the all out set. But you have more practice doing the actual lifts, as much practice as more traditional sets across (e.g. 3 sets of 185 for 8-12 reps per set).

So my thought experiment would be comparing that ramping style vs. sets across. Sets across should in principle involve more effective sets, only 1 in the case of ramping vs. 3 in the case of sets across. But I can't help but notice a lot of very strong people train in that ramping fashion, and having done it myself, it seems to work pretty well to the point where I've questioned if it might be working as well as multiple sets at the top weight.

What I like about this comparison is that it gets to the heart of effective reps/sets. In both cases, you'd have people practicing the lifts as much such that total reps/TUL is actually similar between the groups. What's different is the number of effective sets/reps.

Obviously my gut instinct would be the sets across should work better - more total effective sets/reps. But I'm also not totally sure it would, and in cases where 1 vs. multiple sets are performed in research typically showing an advantage for the latter, it's almost never done like it is above, which seems to more reflect how actual bodybuilders do all out top sets.

 

[NWlifter]

I've wondered that too, same with the Doggcrap trainees. a few warm ups, then one killer set.
I would bet that the one set is pretty darned stimulating. If a person can do that often enough, and for long enough, they'd accumulate some good muscle size. Well, they must as many actually have. Now I'd go back to that study I posted above, those 3 of 9 would do well on that, the 6 of 9 might not...I've done that, I get stronger, but don't grow, I've LOST size while getting stronger when switching to that from more volume. I would bet, that if a person can really nuke the muscle it works great, those that 'nuke the nerves', just get stronger without size.
This puts a monkey wrench in the the 'effective reps' idea too if you think about it....

 

[mikeynov]

I've wondered that too, same with the Doggcrap trainees. a few warm ups, then one killer set.
I would bet that the one set is pretty darned stimulating. If a person can do that often enough, and for long enough, they'd accumulate some good muscle size. Well, they must as many actually have. Now I'd go back to that study I posted above, those 3 of 9 would do well on that, the 6 of 9 might not...I've done that, I get stronger, but don't grow, I've LOST size while getting stronger when switching to that from more volume. I would bet, that if a person can really nuke the muscle it works great, those that 'nuke the nerves', just get stronger without size.
This puts a monkey wrench in the the 'effective reps' idea too if you think about it....

For a video on this topic from Nippard:



He gives a pretty good overview of the different studies to date on benefits of failure vs. not to failure. Two interesting points that caught my eye:

* Failure probably works better in the context of overall lower volume, which has a certain "no shit" factor.
* Failure might work better in novices than in the comparatively trained. This struck me as having a certain logic in that novices may have comparative trouble truly tapping into their muscular potential (volitionally perhaps?) relative to more advanced trainees. Maybe it's as simple as the more advanced have an easier time reaching that full recruitment/higher rate coding zone without giving up.

I also found Israetel's S-curve description of effective reps very much in line with what you've been saying, so give yourself a nice pat on the back.

Thought this was a good video though, and as per usual, I'd love your thoughts.

 

[NWlifter]

Very cool video
* IMO, the main reason untrained do better going to failure, is due to how untrained usually 'guess' where failure is really poorly. A study I've seen found that untrained people usually 'thought' they were close to failure when they actually were many many reps short of failure, where trained people could indeed estimate when they were a rep or two short. So I would bet that if indeed the untrained actually went 2 RIR compared to trained going actual 2RIR, then the results would be the same.

* S curve, yes that makes total sense, it fits physiology too.

* One thing I have yet to see anyone put out there is the actual main reason there are good reps. All the text and talk about it seems to consider it like a whole muscle phenomena, that 'a muscle' is more stimulated by getting closer to failure. In reality, what it is, is due to orderly activation and recruitment levels, in light of how muscle fibers are usually 'faster orderly' also. What I mean is, is the less a fiber is used in daily life, the faster it usually is and faster fibers have more growth potential. Later recruited fibers are used way less in daily life.

OK so if you do 3 sets with 4 RIR, think what really is happening...By stopping with 4RIR, some of the last recruited motor units , even though recruited, didn't get to be 'on' very much, they never experienced very much 'work'. They flashed on and off a few times at the hardest part of the ROM for a few reps, but not nearly enough to tax and stimulate them. So the earlier recruited fibers did get enough work to be stimulated and many of those were slow fibers so wouldn't grow much, this means maybe 1/3 of the FT fibers were stimulated to grow.
Now you do 3 sets with 2RIR, those larger last recruited fibers did get to be 'on' a lot more, they probably had 2-3x as much 'on time' during those sets than they did with 4-5 RIR, so now are stimulated and grow. You'd have almost twice as many fibers that experienced enough tension and work to grow so of course, the whole muscle shows more growth.

If we look at these, you can see the big difference,.... further into the good rep zone, it's that MORE motor units were active enough 'to be ' stimulated. Thus, whole muscle size, protein synthesis, etc. would be higher. 5000 fibers growing with 2RIR vs 2500 fibers growing with 4-5RIR is a big difference in whole muscle growth. Of course, the earlier recruited are also getting more 'on' time with more reps too.

Let say for this hypothetical model, that the fibers in a motor unit need at least a 10 worth of activation time for minimal stimulation and it's that same laying on it's side S curve for that stimulation, ie a 15 is way more than 10, but 20 is a little more than 15, 25 is a LITTLE bit more than 20, etc.
So, after 3 sets with 4-5RIR
First MU's , in this arbitrary number setup I have, got 3x10=30, max on every rep
MU 6 got 3x6=18
MU 10 got 3x4=12
after 3 sets with 2RIR
First Mu's, got the same 3x10=30 (they got plenty of time with either, fully on every rep either case)
MU 6 got 3x9=27
MU 10 got 3x7=21
------------------------------------------------------
As far as recruitment and activation with lighter loads, this graph shows 'where' it happens with heavy vs lighter. this graph shows where 'full recruitment begins with 2 different RM levels.

 

[mikeynov]

A few quick thoughts for you...

• Agreed about failure probably working better in novices due to their poor ability to judge effort/RIR.

• I agree with your general logic, but the thing that gives me pause is papers like Carroll which aren't getting THAT close to true failure and actually showing better results. I think they used no higher than 8 RPE, and a lot of sets seem to be below that. I understand you're illustrating your point with 2 RIR vs. 4, but it really seems like, for more advanced trainees, that line might be even blurrier, or the curve flatter, even a little further out from true failure.

• The other thing I thought I'd point out is some of the conclusions here seem a little off reality, like twice as many fibers being stimulated to grow in 2 RIR vs 4 RIR, even ignoring the previous bullet point. E.g. even the 2005 study that had the 30 second break in the middle of the set showed decent growth in the rest group, and the continuous group was like ~13% better hypertrophy results. So in terms of this applying to the individual muscle fibers, it seems like the net effect of failure vs. not to failure when you're even vaguely in the ballpark of failure (like RPE ~5+) isn't very high in magnitude when you zoom back out to the whole muscle.

 

[NWlifter]

Oh yeah, that was 'exaggeration for illustrative purposes for sure.

it was mainly to show, that what is a good rep, is per motor unit. We can think of them for the whole muscle, which implies enough work for all fibers, but we just need to keep in mind that different motor units experience more or less 'work' than other motor units.

So.. 5x5 with a 10RM would have less work for the last recruited motor unit, than would 10,8,7,6 with that 10RM for example. How much difference, if any, that would make long term, I bet not a lot if any

 

[mikeynov]

https://www.strongerbyscience.com/effective-reps/

New article by Greg on the topic. I'm reading now, but as per usual will share thoughts as I have them

 

[NWlifter]

I read that this morning, was going link it here myself, you beat me to it~

 

[mikeynov]

Nice.

So my head is kind of still in the same place after reading Greg's article. The different contexts also lead us to some nuance in our recommendations.

So I think the "effective sets" model I introduced still explains what we should be doing about as well as anything. As per Greg's article, it would need some modifications, differentiating between compounds vs. isolation lifts. What's fascinating is that, for the prime movers in exercises like the squat or bench, it's possible even lower RPE efforts might still get the job done. But pretty clearly in isolation exercises, particularly the "peak contraction" sort, we'd probably want as high as RPE as is possible.

This line in particular by Greg stood out to me:

So, if you’re benching for triceps growth or squatting for glute growth, you may need to go pretty close to failure to recruit most motor units, but if you’re benching for pec growth or squatting for quad growth, you may be recruiting the vast majority of your motor units from rep 1, even with moderate (i.e. 70% 1RM, or 12RM) loads.

So if you're reduced to using compounds only in general, as high of RPE as is reasonably possible starts to look more attractive again as we'd want to grow as many muscles as possible with the exercise. However, if we separate the two, it would seem we could definitely get away with lower RPE per set, and I think programming-wise, this makes a lot more sense.

So for compound lifts (squat variants, bench variant, deadlift variant etc.) I think the rule of thumb I came up with starting this thread still works well. ~5+ reps per set to be safe, and ~6+ RPE per set to at least be in the right ballpark for an effective set in terms of the prime movers of those lifts. However, we should also probably complement this with isolation exercises where possible at even higher RPE, maybe legit ~8+, to make sure we're not just growing the prime movers, but also the accessory muscles, as well.

 

[NWlifter]

that makes sense, I remember waaaaaaaaaaaaay back, maybe 2003, Dave Maurice (who was awesome with the physiology 'stuff') recommended I read a few books, one was Komi's Strength and Power in Sport, I think that book it was mentioned that when they tested powerlifters for quad activation a 3Rm was the same as an 8RM, only 'other muscles' increased with the higher RM. This also explains how bodybuilders use 'less weight' but what they call a working weight, and get great stimulation 'in the target' muscle. (like pecs). So benching a 15RM for 10 pec focused reps probably hits the pecs as well as an actual 8RM does.

 

[NWlifter]

IMO, ,the key to ALL this stuff is REALLY really being able to visualize recruitment and activation, most of the answers are right there.
Total number of 'ons' for each fiber and that in reference to their condition and specific MHC type is probably the pure answer to volume, reps, TUL, ... etc.
(OK some of the columns total wrong, dag nabbit lol, should have checked that before I put all the layers together and saved it, but you get the idea)

this might help visualize things.

 

[mikeynov]

I made this really terrible graph (I'm no Ron) to illustrate the "effective sets" idea, putting it together with the S-shaped curve Israetel spoke of and relating it to fatigue.

The red curve essentially represents the amount of stimulation we get from the accumulation of reps. So the first few reps do very little, then the level of stimulation starts rapidly rising, but begins tapering off again around wherever the "effective sets" threshold might be. I put RPE 6 here to illustrate the concept, and I think the ~90% stimulation effect is probably ballpark correct, but obviously this is open to interpretation and context.

But the above to me describes compound movements in particular with respect to their prime movers. To me, it makes the most sense to get as much stimulation as possible while trying to minimize the fatigue to the extent possible. In other words, stimulation is a sort of S-shaped curve where we get a disproportionate reward by the time we hit an RPE ~6+, but fatigue is just linear and cumulative. So the sweet spot is probably somewhere a few reps shy of failure or so, at least in the context of compound movements and trained lifters.

 

[Jester]

Are we certain fatigue is accurately described or represented as linear? This definitely doesn’t correlate with my training history.

 

[mikeynov]

I thought about adding a disclaimer. Fatigue is a combination of a few different things, depending on how we define it. There's the literal metabolic fatigue the muscles experience, obivously, but you can kind of throw in central fatigue and even muscle damage in the same bucket. Something like "cumulative stress" might be a better expression to represent the idea I'm after. For compound lifts, you could probably argue that fatigue gets a little exponential-y towards true failure with respect to the central component in particular, maybe?

But if anything, this would just add even more weight to the idea of intentionally limiting RPE to prevent undue fatigue (or more accurately cumulative stress), so the same idea still applies.

edit: I redrew the picture trying to capture this idea a little better, as per Jester's comment. I think the idea is that cumulative stress/fatigue also has an inflection point of sorts around RPE/rep 6 as well, but instead of turning pseudo-logarithmic, it turns a bit exponential. So while the degree of stimulation quickly levels out once we're in the effective sets/reps zone, the cumulative stress/fatigue (the thing determining recovery cost) starts rapidly rising.

So we're trying to weigh the degree of stimulation from a set vs. its recovery cost, basically. So somewhere around our theoretical inflection point into an "effective set" the recovery cost still isn't too high, but if we get too near failure in our quest to maximize stimulation, we might well incur too high of a recovery cost for our efforts.

This, again, applies more to compound type movements vs. isolation, however, as I think the latter's recovery cost really isn't a huge deal in general.