Hypertrophy and Strength, correlation does not imply causation?

Other Discussions Training Discussion
1

Hey guys,

Recently came across some ideas (mainly outlined in this paper: Correlations Do Not Show Cause and Effect: Not Even for Changes in Muscle Size and Strength - PubMed ) from Dr. Jeremy Loenneke that give a hypothesis claiming that the current literature is not really designed to answer the question that increases in hypertrophy (beyond baseline adult levels) result in increases in strength. Although this goes against what many seem to believe, I find Dr. Loenneke’s case fairly compelling. He repeatedly states that’s there is an obvious correlation but the actual increases that result from training, do not seem to be connected in his opinion. He claims that it is very difficult to parse out the skill practice that inevitably comes with the training volume that is associated with increases in hypertrophy. In simpler terms, he expects people who are practicing lifting weights will get better at lifting weights to some extent, i.e. strength gain. Although I remain skeptical of the idea, I still would enjoy hearing your thoughts on it. Thanks again for all the content and the ability to consider new ideas.

Other related papers:

https://onlinelibrary.wiley.com/doi/abs/10.1002/mus.25488

Zac

2

This is an interesting topic worthy of discussion. I’m approving this topic to let people dive into this and digest it first.

3

Great, would love to hear some discussion on this. For those without access to the papers, this podcast gives the key points:

4

Great, would love to hear some discussion on this topic. For those without access to the papers ive found this podcast that I think gives the main points pretty well:

5

I’m just going to comment on the first paper, which a friend sent me a few months ago. I haven’t read the others.

There are two related methodological claims, one bad and one good. The first is that because we haven’t run a very particular kind of experiment to get certain causal data, we can’t be sure that muscle growth from training causes (some of) the strength increase from training. Now, causality is a tricky topic, and if you really want we could do a little philosophy of science dance and debate what “sure” really means here. But I’m pretty comfortable claiming muscle growth causes strength gain. Given the correlational data, it’s difficult to look at the relevant physiology and come away with the conclusion that the relation isn’t at least partially causal. Further, a lack of experimental data hasn’t prevented us from making causal knowledge claims in other domains, so why should it here? For instance, scientists didn’t conduct randomized controlled trials on people to establish that cigarette smoking causes cancer, yet we’re all pretty sure about that due to several other lines of argument. (See https://tobaccocontrol.bmj.com/content/21/2/87.)

The second claim is that we can’t precisely estimate the causal treatment effect of getting jacked on strength from purely correlational/observational data. This is correct, because there are a lot of confounding variables. For example, more educated people on average make more money than less educated people. And while some of that is due to their education, some of the correlation occurs because – on average – people who pursue higher education are smarter and more conscientious than those who do not, and those traits are rewarded by employers. So simply comparing the average salary of a high school graduate and the average salary of a college graduate will overestimate the casual effect of attending college on earnings. (This is known in the economic literature as “ability bias.”)

There are also a few statistical claims at the end that are bizarre. One is that when computing certain statistics, you should throw out all data points that lie within a certain margin of error (I guess measurement error here?) of zero. This seems obviously insane to me. I looked at their references and as far as I could tell none of the ones they cite in that paragraph actually suggest doing this, but if I missed something and they did, I’m pretty sure they’re wrong. Non-responders are important data, too, and while they’re correct to be worried that measurement error could be biasing their estimates, the “responder” data points they use are not magically immune from measurement error, so I’m not sure what the point is.

6

Patrick,

I would totally agree that part of the entire argument is a game of semantics and that the claim, “Hypertrophy causes Strength gain” is an asymptotic statement that can never really be proven without a doubt, due to the nature of science. However, I still think there could be something to learn here.

When looking at the third paper I linked (which I know you mentioned you didn’t get the chance to read quite yet) The group that was designed to elicit hypertrophy (by performing back-off sets) in addition to practicing the 1RM test daily and the group only practicing the 1RM test daily both had similar strength gains. The group performing the extra volume was the only group to see muscle growth, which makes sense given that it seems evident volume drives hypertrophy. Given the proposed relationship that increases in cross-sectional area automatically increase total force production of the muscle and the same amount of test-specific practice between groups, you would expect to see more strength gain in the group with the increased muscle size. The fact that this is not what is observed brings to question if the increase in CSA is actually increasing the amount of force production for the 1RM. One could then say that we know from training ourselves that initial gains in strength are caused by almost purely neurological adaptations and that if the study was long enough, hypertrophy would then be the main contributor to gains in strength. However, the authors claim that if you take this stance, the studies used to assert the causal relationship between size and strength gain are of similar length and that the only valid conclusion is that we don’t really know. I would agree with you that if I had to bet my life savings today I think the correlation is too strong to think otherwise but I think this will be an interesting discussion to keep an eye on going forward. The best way I can come to peace with it in my head is that increases in muscle size raise the ceiling for test-specific neurological adaptation that would result in strength gain, which again is speculation but it’s how I think about it given what I’ve read so far. I think it eventually comes to a point where you have to get bigger to continue to get stronger but as always its subject to change given new evidence.

appreciate your insight,

Zac

7

I hate to be so anecdotal and bro-sciencey but looking at the big picture here, if hypertrophy had no causation on strength gains then why would we even have weight classes? I admire anyone for digging deeper and questioning common “knowledge” but the level of scepticism seems absurd to me.

That out of the way, of course they are right that hypertrophy is not always the main driver of strength increases (and at certain parts of a trainee’s career like the start, not even close) I could probably get my squat from 100 to 300 without gaining a pound. That’s a nice 200lb increase but what about 300 to 500? Forget about it. I think the way my thought experiment fits with observed reality says more about the issue than a 21 day trial can totally disprove.

Now hold on, causality is a tricky topic as Patrick says and perhaps assuming that my thought experiment is 100% true that still doesn’t prove that hypertrophy causes the strength gains but it would prove that the hypertrophy was necessary for the level of strength gains. If I get a bigger wallet it doesn’t make me richer, but it allows me to carry more money in my wallet. So perhaps growing new muscles doesn’t make you stronger directly but it provides necessary material to be later trained for optimal strength. If we completely exclude this from our definition of causation without without making special note of that (I’m not able to read the whole studies so I may be wrong) then we may be misled to an absurd conclusion like “hypertrophy has no effect on strength gains at all” (I made that one up).

But the abstract of the first study is pretty imprecise in its language. It says " To answer the question as to whether training-induced increases in muscle size lead to training-induced increases in strength requires a study designed to produce differential effects on muscle size based on group membership". Notice the “lead to”. I could argue that having a bigger wallet “leads to” carrying more money around even if it doesn’t cause you to have more money.

That said, I would still find it hard to believe that hypertrophy has no immediate and direct effect on strength at all. I mean it is contractile tissue and it contracts. For it’s effect to be purely facilitating in nature the newly created tissue would have to be doing no work at all, just along for the ride until by further training it “learns” what its job is supposed to be. And then if not by doing the job it is supposed to do, how would it learn? By proxy of the tissue around it contracting? Maybe, but I find it more likely that since the tissue was grown in response to a certain loaded movement it would be set up to work through that movement right from the start.

2 Likes
8

A few questions about that paper: What exactly do they mean by trained (as in trained individuals)? What is the exercise and how much could the trainees lift before?

Now some possible explanations of the results: The subjects did actually gain more strength on the hypertrophied arm (2.2 kg vs. 1.9 kg which is a 16% difference). That’s not a huge difference but even if there was no difference at all if the subjects weren’t highly trained in that particular exercise then the novelty of doing it so frequently could have a quick effect on strength gains that is much more significant than the strength gains they could get from hypertrophy in just 21 days.

On the other hand, even if they were highly trained in the exercise they likely never practised a 1RM attempt for 21 days straight and the gains in top end strength from doing this could overwhelm the gains from 21 days of hypertrophy.

9

Thanks for your reply. I looked at just the abstract of the third paper because I’m lazy. If there’s a detail I’m missing below that’s contained only in the text, please let me know and I’ll read it.

Regarding this specific study, I think both you and neandrewthal are right. The people doing more volume and growing muscle did in fact get stronger. Further, the sample size was tiny (5 people!) and the study was short, so it’s difficult to draw conclusions about how hypertrophy potentiates long term strength increases. For instance, how much hypertrophy can we reasonably expect from a month of elbow flexion? I’m not sure, but I’d guess it’s pretty small, on the order of millimeters. That alone probably isn’t enough to provide much of a strength boost, so a 16% advantage to the group with more volume seems reasonable. Also, while these subjects may have been “trained,” I doubt they been doing the movement with a frequency corresponding to 21 consecutive days of training, so I find it plausible there were large neural strength gains just from the massive amount of practice. I’d be really interested to see what happens in the next 21 or 42 days, after those easy gains are exhausted. (And in general I don’t think studies with subjects numbering in the single digits should be given too much weight.)

The broader methodological point, that the arguments I just made can be turned around to show that other studies that purport to show a connection between muscle growth and strength gain are too short or underpowered to confidently assert a causal relationship, I believe is correct. But this is just a corollary of the general fact that most exercise science studies are too short and don’t use enough subjects to reliably detect effects of interest. This doesn’t really bother me, because as I said above, I don’t assert a casual relationship between muscle growth and strength solely on the basis of a few 4 week studies with a dozen people.

1 Like
10

Neandrewthal, some great thoughts,

Trained was classified 1 year, at least 3 sessions a week of resistance training experience and including the tested exercise (dumbbell curl adjustable up to 0.5 lb increments) at least twice a week. Obviously, a curl for a 1RM is not ideal but some considerations had to be made in a research setting. Now I agree If you look at the overall change in strength, the hypertrophied arm comes out slightly on top, although not statistically significant. This would stand to the current paradigm that hypertrophy maybe matters a little bit in this short of a time-frame but mostly neurological skill adaptations are at work. However, looking at the graph that plots the 1RM’s over the course of the study, the “testing arm” or non-hypertrophied arm actually exhibited the highest 1RM in the entire study around session 13. It looks like both groups had fairly linear progress, although the testing group drops off considerably in session 14 and never reaches back to true peak condition. This could be how Jordan talks about the effects of “Bulgarian” type programming, in which previously built strength is just more effectively able to be demonstrated. However, I would be interested to see what effects would come if progressive overload was applied to this type of protocol over the long term. Just as we know increasing training volume over time is what drives long-term hypertrophy gains, I’ve heard the author mention in a few podcasts that overload still applies but would just come in the form of more practice, or exposures. It would be interesting to see if the progress would begin to trend upwards again and not be limited to short-term peaking effects.

Overall, I think the short time frame makes this difficult to truly gather from and I’d like to see it extended to the length of a typical training study.

Also, just to be clear, I’m not totally sipping the “hypertrophy doesn’t matter at all” Kool-aid, I just enjoy playing devil’s advocate every once in a while and looking at things from the side opposite my bias.

Fahves tommorrow

11

Patrick,

Very much agree with you analysis. 5 people is garbage and makes the results of the study near impossible to validate whole-heartedly. Although I feel for these researchers, it would be pretty tough to get a significant amount of people to sign up for 21 consecutive days of training, even if it is curls lol.

Overall, I just thought the concept itself was interesting and would love to see a better protocol surface over the next few years, just to be a little more certain on what I think we all assume to be true based on the current evidence.

Great discussion

1 Like
12

@Austin_Baraki Would love to hear your thoughts whenever you get a chance, thanks again

13

Hey Zac,

I moved this topic to the unmoderated forum to allow the discussion to proceed without me needing to approve every post, as I’m currently slammed with work on multiple fronts and preparing/updating my lectures for this weekend’s seminar. If I have time to get around to this soon I certainly will, otherwise since I plan to write more about this topic in the future, I’ll be sure to include a discussion of it there too :slight_smile:

1 Like
14

Subscribed for nuanced. Kudos for sharing.

I think that that it’s a pretty interesting concept that like stated in the abstract needs more data on the subject but if it’s right it would explain my gain(zzz)openia. :(.

15

I say we take a look at this from the most fundamental viewpoint I can think of - Newton’s Second Law of Motion: FORCE (F) = MASS (M) * ACCELERATION (A). If our goal is to increase F, then one or both of our inputs to this equation, A and M, must be increased.

When we train, how and to what extent can we modify these two variables in order to get stronger? For the sake of simplicity, let’s assume that the lifts in question are perfectly performed for every rep irregardless of which two variables we change. Let’s also assume effects from fatigue are held constant between all tests.

In this case, acceleration would be how fast the muscles in our system contract against their respective attachment points from their pre-loaded, at rest (zero-velocity) state to their maximum lifting velocity. Likewise, the mass in this equation is the total amount of muscle contributing to force production. Total lean muscle mass would be directly proportional to the lifter’s total muscle CSA, since AREA * LENGTH = VOLUME, and VOLUME * DENSITY = MASS. We assume muscle length (the range in which it can contract and lengthen) is held constant between all CSA states. This assumption prevents muscle length from contributing to volume change.

For me to envision a scenario in which an increase in muscle CSA does not CAUSE an increase in force production, I would have to assume that there is an accompanying decrease in the ability of the muscle to accelerate under load or a decrease in active muscle density. Active muscle density in this case is the amount of muscle mass contributing to the production of force. In short, the muscle is bigger, weighs more, but it generates the same amount of force and acceleration because only a portion of its mass generates force. Would the explanation for this be that there is a high amount of sarcoplasmic hypertrophy compared to myofibrillar hypertrophy, and we simply just made the gas tank bigger? Would another explanation be that we cannot efficiently recruit the added mass from myofibrillar hypertrophy - like there’s an extra cylinder in an engine, but we can’t spark it yet?

In short, I believe this thought experiment suggests that training induced hypertrophy can be present alongside training induced strength gains and not add, in any way, to the strength gains. The necessary conditions for this would have to be one of the following:

  1. ALL of the hypertrophy that an individual experiences is sarcoplasmic.
  2. NONE of the net muscle mass from myofibrillar hypertrophy can be recruited yet.

How likely are these conditions to occur, you ask? I’m not qualified to answer that.

What are your takes on these nuances? Does my rambling make sense?

16

Swollverine,

I think you bring up a lot of valid points, the one thing that could add some nuance to the discussion that the authors suggest is that there may be a series muscular adaptations occurring locally (without increasing CSA) that we can’t detect or really understand quite yet.

This makes some sense to me in the fact that it comes to a certain point where our brains inhibit the amount of muscle we can voluntarily contract. I believe Austin has mentioned in one of the podcasts where it comes to a point where endurance athletes (which might not be the exact same) pretty much just have to psychologically endure the task rather than their bodies being unable to physically complete the task, and that higher level athletes with either a genetic predisposition, increases from training, or likely both are able to do this more successfully than the average person. I think this could shed some light on the possibility that there are local adaptations that can occur to maximize the force output of the given CSA.

However, logically, this would still suggest to me that eventually, CSA would need to increase. If you took the argument all the way out, you would get somebody so “efficient” at the skill that they would be able to cause damage from overriding the natural inhibition from the CNS, i.e. breaking your femur in a squat, which I’m not convinced is even possible. In addition, the authors contend that an adult gains most of the muscle mass they will gain with training within the first year, so from then on it’s the skill adaptation for the current level of CSA that matters for strength development, which has a lot of untapped potential in their opinion.

Overall, I still think the correlation is too strong to ignore but I’m still open to new evidence in the future that would maybe sway programming, and practical application slightly. If anything, I think this further stresses the utilization of singles @ 8 in training for powerlifters/strength lifters.

17

It’s almost like