Is Load Progression Necessary For Hypertrophy?
- Thread starter mikeynov
- Start date
Well, per Beardsley logic it certainly would be full recruitment, as in theory the last ~5 reps of any set to failure at > 5 reps should be at maximum recruitment. For clarification, bear in mind the formula we're working off of here is Reps in Reserve = 10 - RPE. So an RPE of ~6 means only having 4 reps in reserve. That is definitionally what RPE is, as per Tuscherer, it's a metric of performance, not "how hard does this feel?"
For example, a set of 10 @ RPE 6 would mean that's 10 reps with a 14 rep max. That's pretty heavy/reasonably hard, particularly in something like a squat or deadlift. Some of the trouble with terminology is that Tuscherer's original descriptions pertained to low reps. So a single @ 6 is a 5 RM, and I think he originally described a 6 as "speed work," but that had a very particular meaning in a very particular context, in this case what research actually supports is a good range for speed work when dealing with very low reps. But in a set of appreciable reps (like our usual ~8-12), an RPE 6 is no joke, it's definitely a "working" set.
On the research end of this I'm pretty sure RPE ~7-8 (2-3 reps from failing) is often what produces similar results to failure, so to be honest ~6 is probably right around the threshold, I was just erring on the side of being conservative if it's a legit 6. Mike Israetel starts all his hypertrophy programming at an RPE ~7 (was just looking at his RP templates) based on the research to date, and if we're suggesting this still works as well as failure, then we have to ask why, and as per Beardsley logic, it's probably because we're already at full recruitment and are producing at least some fatigue. Is 6 enough? As above, it's probably right around the threshold. 7 I guess is safer, if we want to be as safe as possible. I could actually see even RPE ~5 being adequate at least for the full recruitment part, though it's then a question of how much fatigue we need.
edit: expanded discussion of the proximity to failure and full recruitment.
So, if ~80% of 1 RM is the usual tossed around number we've used for full recruitment from rep 1 in this thread, that would represent a ~7-8 RM for most lifts for most people, though there is some variation here. 85% of 1 RM is more like a 5-6 RM if we want to be extra super careful about full recruitment, and is what I think Beardsley is using for the basis of his logic.
So let's take a set of 12 reps to failure ~70% 1 RM. After we perform 7 reps, we're at an RPE of ~5, i.e. we have 5 reps we could perform before we'd fail on the 6th (13th) rep. When looking at the remaining reps, we have 5 potential reps to go, so in terms of recruitment, the argument is that we are using what is, in effect, our temporary 5 RM, and every single rep should be at full recruitment thereafter.
Note that, as per my observation above, this is actually being extremely conservative. More realistically, we might only need to be ~7-8 reps away from failure to technically have full recruitment if we were to use this logic, as that more meaningfully corresponds to the lower end of the range that might be at full recruitment from the first rep on (80% of a 1 RM).
But based on the above logic and then a bunch of other research, I tend to use ~5 reps from failure as it seems more right-ish. RPE ~5 should land us at full recruitment, and any reps past this should be reps done in the context of already full recruitment, hence my recommendation of RPE 6 (RIR 4) at the bottom end, as we'd get at least some fatigue in the context of full recruitment.
Last edited:
As per my reply to Ron, this is coming from a combination of why these submaximal sets are probably working in the first place (full recruitment + at least some fatigue) and then actually looking at the research, too. One of the studies I had in mind was fixed 10's at 60% 1 RM, which would actually be an RPE of like ~4-5 tops (5-6 reps in reserve per set), and they actually produced really good hypertrophy.
Similarly, the 5 x 5 type programs that appear to produce the best results (without grinding people into dirt) tend to hang out in the ~75-80% 1 RM range, where it'd be something like an RPE of ~5-7.
So yah, just combining these factors and guessing that ~6 is probably right around the threshold. RPE ~5+ should see full recruitment, and RPE 6 is at least one rep that's performed in the context of that full recruitment. Still, just a guess on my part.
A little more on this topic if people want to do some reading.
https://www.strongerbyscience.com/training-to-failure-or-just-training-to-fail/
This article is pretty good, and makes the case of the emerging research on not to failure vs. failure, though does note that there is some conflicting research, here. Though as per the other studies I'll cite below, the trend is definitely towards not to failure probably being fine if the RPE is high enough. Unfortunately this is now a little old and I hope Greg updates it at some point.
Some select quotes:
Then take a gander at this post by Greg Nuckols:
http://thinkmuscle.com/community/threads/hypertrophy-and-effective-reps-incomplete-model.43530/
The usual study cited in failure vs. not failure is the Sampson et al paper, which I think is where the original RPE ~8 figure is coming from, as they compared sets of 4 vs. 6 between groups:
https://www.ncbi.nlm.nih.gov/pubmed/25809472
I think the Helms dissertation research was with Zourdos, and refers to this:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4961270/
That's a really, really good read on Reps in Reserve in general. The discussion around the prescription of sets of 8 for hypertrophy at 70% 1 RM is pretty illuminating in terms of variations and the sorts of RPE/RIR they're aiming for. In this case, trying to get as many people as possible in the RPE ~6-8 range (RIR 2-4).
The Carroll et. al. study Greg references is this:
https://www.mdpi.com/2075-4663/7/7/169/htm
This one is a little more confusing in terms of just what the percentages correspond to as per their own charts, but the idea here was training substantially shy of failure here actually produced better results than literal failure.
But yah, the above I think are the best resources I've come across on this topic literature-wise. There are obviously hurdles in general when trying to use RPE/RIR when estimating useful hypertrophy sets, but I do think we can make some statements if we just make the assumption the assessment itself is accurate:
At a guess, conflicting data is probably related to how well people are estimating RPE in the first place, and as per the Helms and Zourdos paper, this is quite possibly a function of who the population studied is. Trained people are seemingly much better at estimating RPE/RIR, whereas novices are maybe not so good, so it'd be much harder to prescribe stuff for them based on these concepts.
https://www.strongerbyscience.com/training-to-failure-or-just-training-to-fail/
This article is pretty good, and makes the case of the emerging research on not to failure vs. failure, though does note that there is some conflicting research, here. Though as per the other studies I'll cite below, the trend is definitely towards not to failure probably being fine if the RPE is high enough. Unfortunately this is now a little old and I hope Greg updates it at some point.
Some select quotes:
Then take a gander at this post by Greg Nuckols:
http://thinkmuscle.com/community/threads/hypertrophy-and-effective-reps-incomplete-model.43530/
The usual study cited in failure vs. not failure is the Sampson et al paper, which I think is where the original RPE ~8 figure is coming from, as they compared sets of 4 vs. 6 between groups:
https://www.ncbi.nlm.nih.gov/pubmed/25809472
I think the Helms dissertation research was with Zourdos, and refers to this:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4961270/
That's a really, really good read on Reps in Reserve in general. The discussion around the prescription of sets of 8 for hypertrophy at 70% 1 RM is pretty illuminating in terms of variations and the sorts of RPE/RIR they're aiming for. In this case, trying to get as many people as possible in the RPE ~6-8 range (RIR 2-4).
The Carroll et. al. study Greg references is this:
https://www.mdpi.com/2075-4663/7/7/169/htm
This one is a little more confusing in terms of just what the percentages correspond to as per their own charts, but the idea here was training substantially shy of failure here actually produced better results than literal failure.
But yah, the above I think are the best resources I've come across on this topic literature-wise. There are obviously hurdles in general when trying to use RPE/RIR when estimating useful hypertrophy sets, but I do think we can make some statements if we just make the assumption the assessment itself is accurate:
- RPE > ~5+ should be full recruitment.
- The most cited study comparing failure vs. not to failure found comparable results at RPE 8.
- Research to me looks like it's trending towards even lower RPE maybe being comparable, too, though where the exact line is, I don't know.
- There is some indirect research showing similar underlying muscle activity once you hit RPE ~5-7.
- Similarly, if you go looking for empirical examples of tried and true "hypertrophy work" that's not to failure, you'll probably arrive somewhere in the RPE ~5-6+ range when dealing with sets >= 5.
At a guess, conflicting data is probably related to how well people are estimating RPE in the first place, and as per the Helms and Zourdos paper, this is quite possibly a function of who the population studied is. Trained people are seemingly much better at estimating RPE/RIR, whereas novices are maybe not so good, so it'd be much harder to prescribe stuff for them based on these concepts.
Last edited:
So, tangent on the failure point;
Many compound exercises will reach failure due to ‘weakest link’ reaching failure. I.e lower back in squat or deadlift, anterior delts in the bench press etc. When I can’t break another deadlift rep, 99% of the time it’s nothing to do with my glutes.
How/should we be accounting for this when we consider failure? Compared to say failure a cable row or skullcrusher etc.
This is a good point. The answer is I'm not sure, and my thinking on this has even changed a bit.
As per that Menno video in my first post, in the context of something like Kaatsu, occluding the bloodflow of your quads in something like a squat ramps up metabolic activity and makes the muscle hit full recruitment comparatively earlier. But what's a bit like black magic to me, it's not just your quads that benefit, but your glutes, too. Which aren't occluded. Which basically means that we've made the quads a super weak link in the kinetic chain, right, and they fail way sooner. But as a consequence, another prime mover, the glutes, appear to experience the exact same proportionate increase in stimulation/recruitment, presumably because they're picking up the slack to a greater extent.
When watching that video it immediately occurred to me that a lot of our "weak link" arguments regarding failure might be wrong, at least from a recruitment/fatigue standpoint. To take your example, the anterior delts or triceps might fatigue first in a bench press. But what's the effect of this on chest recruitment? If it's parallel to what happens to Kaatsu, it would actually mean the chest would have to pick up the slack of the weaker links, and might receive the benefit despite them being a weak link.
To put this in a pseudo-equation, if (net bench press force) = chest (C) + shoulders (S) + triceps (T), where each letter contributes a certain amount of force, if you reduce the force potential of S or T (i.e. they're the weak link and fatigue first), and C is in a position to contribute more force via increased recruitment/rate coding, it almost by definition would have to as volitional effort increased to keep performing reps, right? So while the shoulders/triceps might be a weak link in how much you can actually lift, they wouldn't actually impede the recruitment or fatigue the chest experiences.
That said, I'm not sure it works like that. But it did occur to me that it plausibly could.
In testing this against the research, the Sampson paper is on elbow flexor exercises, so basically isolation exercises. In that context, we can say super confidently that failure would mean high fatigue of the elbow flexors. But most people apply it to compounds where, if the above logic DOESN'T hold, you're right, it throws a wrench in things. However, other emerging research (stuff that Helms cites, that Carroll paper) also shows comparable or even superior effect of not to failure training explicitly for compound movements, so my intuition is that it probably doesn't matter, and the reason it might not could be something similar to the Kaatsu logic above.
edit: added another little paragraph to better illustrate the concept.
Last edited:
Interesting thought-point re: recruitment of non-failing muscles in a compound lift.
I suppose another way to frame the query is; do we, and if so HOW do we distinguish between failure of a movement/repetition and muscle failure?
Good question. I'm hoping Ron has some input here as surely someone has like, tried to measure different EMG patterns in the prime movers of an exercise as one approaches failure. That's not a perfect proxy, but it would give us some insight, I think.
On a literal level, I feel like the pseudo-equation I offered above should work in principle, right? In something like a bench press, we tend to get stuck in the midrange somewhere, several inches off the chest, when truly failing. If you took a snapshot of what's happening in the chest, (anterior) delts and triceps, what's actually happening in each? In theory, all three should be able to contribute force that causes the bar to move upwards. If the bar is now stuck, what does that mean if our volitional effort is as high as possible, i.e. we've hit legit failure?
I'm not sure it makes sense that the chest wouldn't be at full recruitment. If you're trying as volitionally hard as possible, the chest shouldn't be able to "hold back" barring something like pre-existing muscle damage or central inhibition. If the S and T components as per my equation above drop in force enough to stop the bar, and C is still capable of causing bar displacement via contraction, then we should be at full recruitment of the chest by definition.
The wonkier thing, I guess, would be the degree of fatigue. If the triceps or shoulders fatigue first, then the high threshold motor units in them would have experienced longer duration of fatigue, at least, compared to the chest, even if it does pick up the slack at the end. But the Kaatsu research makes me think this may still not matter from a stimulation perspective, otherwise Kaatsu would be better for the quads in occluded squats than the glutes, which doesn't appear to be the case.
This made me laugh for some reason. I assumed we weren't talking about the same thing RPE-wise. To be honest RPE isn't even the greatest term, and I'm not entirely sure the original usage of it even by Tuscherer matches what most people are now treating it as. I should have clarified more what I meant.
The first RPE I'm aware of related to cardiovascular exercise and was the Borg scale, 6 to 20. Weird ass scale, and was literally perception of effort, with ~15+ being "hard." There are newer scales that are just 1-10 to simplify things, and I think Tuscherer borrowed this when coming up with his original concept and applying it to lifting. At some point, people used it entirely as a metric of performance, not "how hard this is," even though that's also a reasonable thing to ask as perception of effort does relate to gross training demands, injury prevention etc. But yah, reps in reserve and RPE became effectively interchangeable, where they're just inverted. E.g. RPE 10 is RIR 0, RPE 9 is RIR 1, and RPE 0 is RIR 10. Weirdly RPE and RIR 5 mean the same thing. Incidentally and confusingly, you can have 0 or even less than 0 on an RPE scale. Like if you did 5 reps with your 30 RM, that's an RPE of -15...lol.
It also has other weird implications. For example, if 85 lbs is 85% of your 1 RM of 100, then the minimum RPE even possible with that weight would be 6 (1 rep with 4 reps in reserve). Picking up the weight and doing no reps at all would technically be an RPE of 5.
So yah, it's a little muddled, and the more accurate and meaningful term is probably just reps in reserve. For some reason in the lifting world, however, probably due to Mike Tuscherer's influence, a good % of the people doing this still use RPE.
The reason I'm trying to introduce these concepts is because I think it's an extremely useful tool that we can use when endeavoring to come up with a modern HST. Since relative intensity of effort is probably the most important factor in a set, it also makes sense to actually prescribe our loads based on the RPE they represent, and adjust accordingly as necessary.
edit:
Here's a link over on exodus of me fucking around with RPE tables, including a link to a little Excel calculator if you want to play around with this stuff:
https://www.exodus-strength.com/forum/viewtopic.php?f=3&t=2967
Last edited:
NWlifter
Active Member
Yes, I've thought about, and even posted about that on forums before, mostly to HIT people who think a person HAS to hit failure to even have growth. I always would tell them, if failure is required, then doing a compound is only causing failure, maybe, in one of the prime mover groups, all of them are not going to hit failure at the same point. Some people probably have 'pec' failure with bench, some tricep, depending on their personal joint leverages, bone length, tendon attachments, etc. Since those guys were growing doing those basic routines, and not just in the 'failing' muscle group, that in and of itself proves failure isn't required.
Ha yes, I'm brain dead here lately, work plus about 4 hours a day working on our house, surprised I can even tie my shoe laces lol.
Great info. in all those posts Mike!
Yes RPE is and can be kinda vauge I guess. I think I like the RIR way a bit better, instead of 'feeling' how hard it is and rating perceived exertion , you find the failure point, then later your 'guess' is a bit more accurate with reps in reserve. Kinda the same thing but it's the real number of reps rather than a percentage of how hard it felt. I'm sure a decent load with 2-3 reps in reserve is plenty, it's how I've trained my whole life 90% of thte time, long before anyone even talked about all this stuff. I'd just do a set till I knew I only had a couple left in me.
[QUOTE="mikeynov, post: 262277, member: 28924"
On a literal level, I feel like the pseudo-equation I offered above should work in principle, right? In something like a bench press, we tend to get stuck in the midrange somewhere, several inches off the chest, when truly failing. If you took a snapshot of what's happening in the chest, (anterior) delts and triceps, what's actually happening in each? In theory, all three should be able to contribute force that causes the bar to move upwards. If the bar is now stuck, what does that mean if our volitional effort is as high as possible, i.e. we've hit legit failure?
[/QUOTE]
Thinking on this some more, I think it’s fair to blame leverages/biomechanics for the difference between movement failure and muscular failure.
And in turn, makes me withdraw some if my thoughts about studies using leg extensions instead of squats
On a literal level, I feel like the pseudo-equation I offered above should work in principle, right? In something like a bench press, we tend to get stuck in the midrange somewhere, several inches off the chest, when truly failing. If you took a snapshot of what's happening in the chest, (anterior) delts and triceps, what's actually happening in each? In theory, all three should be able to contribute force that causes the bar to move upwards. If the bar is now stuck, what does that mean if our volitional effort is as high as possible, i.e. we've hit legit failure?
[/QUOTE]
Thinking on this some more, I think it’s fair to blame leverages/biomechanics for the difference between movement failure and muscular failure.
And in turn, makes me withdraw some if my thoughts about studies using leg extensions instead of squats
Failure of the movement, i.e. a deadlift, compared to failure of the muscle/s involved in that movement - i.e. erector spinae.
If I take my deadlift to failure, ‘failure’ is occurring because of my core (which is always weaker than my leg muscles or glutes), compared to taking my quads to failure on a leg extension or the glutes on a hip thrust or kickback etc.
For the logic I outlined previously with the bench, the triceps (or shoulders or whatever) are actually a prime mover. But the deadlift is an interesting example, ditto squat, because, in theory, the spinal erectors aren't (or shouldn't be) causing any actual bar displacement. For the deadlift, the movement is coming primarily from the hip extensors with a little bit of quads, and the squat obviously a lot of quads but still hip extensors (though more glutes and adductors than hamstrings).
But for either exercise, the weak link is a stabilizer. In that case, I'm not sure how my previously outlined logic would work. Like if your back gives out in a deadlift, I'm not sure what's happening to the glutes/hams. I suppose in principle they should still be at max recruitment (maybe?), though it's a question of inroad/fatigue.
Like if you took this to an extreme example, imagine having the grip as the limiting factor by forcing double overhand or using one of those thick bars. What's actually happening to the prime movers when you fail?
Just an interesting thought experiment.
Well where this makes me wonder, critically speaking, is how reliable are some of the studies if they’re examining or comparing ‘failure’, but using a compound (especially barbell, e.g. squat vs leg press) movement to failure. How trustworthy are any outcomes re: failure? Should be re-considering some of those studies as though ‘failure’ is actually just ‘more fatigued’?
Just a tangent.
It is for sure a really good question. As I said above, that Sampson paper that's most commonly cited on failure vs. not to failure was actually with elbow flexor exercises, essentially isolation movements. The Carroll study did use compounds, however, but I need to look at that more to see exactly how far from failure they were.
One thing that occurs to me, however, is that if a leg press doesn't have the lower back as a limiting factor, which it doesn't, and if there is a substantial difference in stimulation as a result of the lower back being the weak link in the kinetic chain, you'd think research would actually show superior growth of the leg muscles in a leg press vs. a squat. To the best of my knowledge this hasn't been demonstrated, however, so it might be a question of just how weak the weak link is.
Like with a squat or deadlift, it's possible your lower (or even upper) back would give out, but it's probably not wildly out of sync with the prime movers. Whereas in my example of doing something like a 2" bar double overhand deadlift, your grip strength will be wildly out of sync with your hip extensors such that I'd imagine it would have to be worse for growth vs. a conventional deadlift in that it would be almost impossible to meaningfully fatigue any of the prime movers in any given work set.
NWlifter
Active Member
Oh ok like volitional failure
NWlifter
Active Member
Yes for sure, and that's all failure is anyway, just a bit more fatigue.
Hit failure with 200, when fatigue limits your strength to less than that at the sticking point, failure
Use 190 instead, it takes a bit more fatigue to limit below 190.
the only linear marker with any RM to failure is activation. If neural output is maxed at failure, then it's the point where as many fibers as possible have hit tetany.
And finally found a reference I read many many years ago...
http://www.indiana.edu/~nimsmsf/P215/p215notes/LabManual/Lab9.pdf
tension must be sustained beyond the fraction of a second generated by a twitch. Therefore, most skeletal muscle contractions in the body are tetanic contractions.
Now I'm thinking again, that 5RM really won't increase 'per fiber tension' over a lighter load for example. It will just have those fibers create that tension a lot more often.
Last edited: