6

It seams that there is no actual scientific research in the physiology of muscle growth. All the studies done are far from scientific since all of those are statistical studies, the results of which are questionable.

The methods of such studies have been widely criticised by many experts and also a lot of those contradict each other, for example there are studies that claim that 8-12 repetitions are optimal for muscle growth while others claim that repetition number doesn't matter. In addition to that, there are many variables that are not considered for the experiments.

Lastly, another problem of such studies, as with any type of statistical study, is the failure to explain why the results are what they are. Why does that certain number of repetitions is better, what happens inside the muscle? What are the biological processes that occur during a specific type of training.

Statistical studies are not meant to prove something but rather to either confirm a theoretically proven theory, which in those cases does not exist, or to gain some insight about which direction to head in order to prove the theory and again none of those studies does that. So my question is, are all those experiments done actually scientific or pseudoscience?

10
  • 5
    Could you be more specific about which studies or which individual topics of sports science you're interested in? I can't tell if you're calling the entire field a sham, or if you are critiquing our current level of understanding of hypertrophy, or the relationship between training variables and hypertrophy? And when you refer to experts, do you mean sports science experts, in which case do they not present their own evidence and theory? Aug 25, 2020 at 10:23
  • 3
    OK. Could you be more specific about which studies or which individual topics of sports science you're interested in? You're kind of hand-waving an entire field here. Aug 25, 2020 at 10:50
  • 1
    How are these studies contradictory? Is it possible you're just seeing the healthy internal back-and-forth debate that should exist in a field, especially a field where getting gold-standard data is difficult? Aug 25, 2020 at 12:21
  • 2
    @Dave: I agree that getting good data is the problem here. However I disagree with the notion that back-and-forth is healthy. In 1686 Newton proposed his second law of motion: F=ma. Then the race was on to try to disprove his hypothesis. However nobody could disprove this and the law became accepted as valid. That was until 1905 when Einstein generalized this law. However F=ma still holds true as an approximation for objects that move with v small compared to lightspeed. Within the scientific method knowledge becomes more precise it does not move back-and-forth.
    – Andy
    Aug 25, 2020 at 19:30
  • 8
    @user33976 your physics background may be misleading you as to how reductivist other scientific fields are. In many fields (medicine being the most obvious one that jumps out at me), it's not uncommon for the underlying causes of phenomena to be poorly understood, but for statistical analysis of the data (which does indeed sometimes yield contradictory results, often due to difficult-to-control variables) to be used to glean actionable insights.
    – James_pic
    Aug 26, 2020 at 9:02

5 Answers 5

24

Muscle growth physiology is not a pseudoscience as researchers are using scientific methods to collect and analyze data and make logical deductions.

It is true that our knowledge of muscle growth is limited and the statistical studies do not explain and prove the results, but this does not make the field non-scientific.

It would be called pseudoscience, if researchers were using methods that are either logically or scientifically wrong, which is not the case.

As far as contradicting studies are concerned, I have seen quite a few myself, this is mostly due to the fact that, as you mentioned, we can't know all the variables, and there is of course some error. This however doesn't imply that non-scientific methods were used.

I agree that statistical analysis is a poor method to actually prove a physical phenomenon, but sometimes it's the best we have, biological phenomena are extremely complex and hard to model.

2
  • 3
    You make good points, but when a (scientifically backed) vague correlation is used to make strong statements that are not supported by said correlation, then I think it qualifies as pseudoscience. Not saying that all people involved do, but there is a clear trend of people overselling their methods.
    – Ant
    Aug 26, 2020 at 6:43
  • 7
    One can make a much stronger statement here: statistical analysis is the standard way to test hypotheses (like whether some drug cures some disease) in medicine and other fields. It doesn't prove anything however, it typically just gives you some correlation with some confidence (meaning the two things may not be related in the way you think and there's a risk of it giving the wrong result).
    – NotThatGuy
    Aug 26, 2020 at 11:51
19

I do not agree with your insistence on a mechanism. In 1964 the U.S. Surgeon General published a report where he concluded that smoking causes cancer. At that time the mechanisms behind this was not understood. However there was statistical evidence in the form of a study with over 1 million subjects that showed that smoking and cancer were highly correlated. That meant that either

  • A. smoking causes cancer
  • B. cancer causes smoking
  • C. another third factor causes both smoking and cancer

B. was easily be ruled out by only including those who reported to have smoked for some time prior to getting cancer. From what I understand C. was ruled out by adjusting for plausible third factors.

However I agree with your sentiment that much of the research within this field is low quality (underfunded but performed according to scientific methodology). Looking at the metastudy linked by JustSnilloc I see some problems with many of the studies included:

1. The number of subjects is too low

3 of the studies have n = 7, 9 and 11. I doubt much trust can be placed in the results of a study with so few subjects. There is appearantly a statistical method called MBI that lets you conclude with high certanity from extremely small sample groups. This method has however been disproved and the only field where it is being used is sports science.

2. The timespan of the study is too short

For untrained subjects in particular the "newbie effect" causes hypertrophy and strength increases almost regardless of protocol to begin with.

3. It test the wrong hypothesis

It does not test the hypothesis coming from experience in the bodybuilding and strength communities: 3-5 repetitions are ideal for strength whereas 8-12 is ideal for hypertrophy. Instead it typically compares 8 repetitions with 20 repetitions. There seems to be a general problem with papers being produced by people with little or no practical weight training experience: Rippetoe: The Problem with “Exercise Science”

4. The methodology in the included studies are too different.

I also question the value of pooling many studies with slightly varying experiments in a metastudy. As an example say that study A says that 15 reps are better than 10 reps for hypertrophy in the leg press for young men. However study B says that 5 reps are better than 12 reps for hypertrophy in the bench press for women over 40. What are we to conclude from this? The statistical method used to combine all the studies seems highly complex. It also seems to involve a lot of weights and assumptions. I remember the quote "With four parameters I can fit an elephant". Meta-analysis seems to have been used to generate some strange conclusions. This paper questions the use of meta-analyses to evaluate resistance training: "In conclusion, considering the large number of variables involved in resistance training and the methodological inconsistencies in the current literature, it seems impossible to make comparisons of different studies or include different studies in the same analysis".

To me it seems that within this field a lot of reasearch is produced, but much of it is low quality or not really relevant. There may be a problem with funding. There seems to be enough funding for running journals and doing peer reviews etc. but maybe not enough funding to actually do long running experiments with a large number of subjects. Instead many resort to trying to extract more info from existing studies by doing meta-studies of questionable quality.

It may also be that such experiments have been conducted by large olympic teams in the past (USSR in particular) but never published. The result may now be common knowledge among top level coaches and they see no need to rerun the same costly experiments.

13
  • 9
    @Malady FWIW we know a lot more about the mechanisms of carcinogens these days. Back in the day though, the mechanism wasn't known so we only had statistics. It used to be thought that the main issue for smoking-related issues was tar content, and that's a mechanism which could easily be seen. But statistics said that low tar cigarettes didn't change the number of cancers much. So Andy isn't really right for today, but is right for when the link between smoking and cancer was made.
    – Graham
    Aug 25, 2020 at 23:43
  • 2
    I chuckled when I read B. So terrible is cancer it reaches into the past compelling one to smoke. Causality be damned.
    – DKNguyen
    Aug 26, 2020 at 1:03
  • 3
    Obligatory XKCD: Cancer causes cell phones Aug 26, 2020 at 3:30
  • 1
    @DKNguyen Don't confuse "cancer" with "diagnosed cancer". There are plenty of people with stage 4 cancer who don't realize it because they don't yet have any obvious symptoms (and no obvious "lumps" either). Suppose "lung cancer" typically started in infancy but was undiagnosible at that stage (e.g. because the invasive procedures required to diagnose it would kill more people than the disease itself).
    – alephzero
    Aug 26, 2020 at 15:17
  • 2
    RE: low number of subjects, I participated in a study like this, where muscle biopsy was used, I can tell you the procedure is invasive (and painful) and full recovery takes a few weeks, meaning a lay-man's before and after comparison is very hard to achieve because you have to either take samples very far appart in time (to allow for recovery) or physical location on the test subject (before in left bicep, after from right bicep)... you can imagine, even then there are a lot of variables including diet and habits. Establishing a reliable control is the same challenge! Aug 27, 2020 at 7:44
4

This is not much of a fitness question, but a question about what constitutes science and whether a specific body of research is scientific.

One of the most common distinctions between science and pseudoscience comes from Karl Popper, who is perhaps familiar to you from your physics background. For Popper, what makes science different from pseudoscience is that tests or experiments have clear point of demarcation that distinguish in a black and white way true from false. All statistical hypothesis tests are therefore scientific (because they have a clear demarcation in the critical values that distinguishes true from false unambiguously), but things like astrology and tarot reading are not.

Is research on muscle growth scientific? According to this simple presentation of Popper's view, yes - absolutely. In fact, any body of research based on null-hypothesis significance testing is, so long as the critical values are clearly specified.

There are of course other ideas, but I would recommend you to the SEP article for a better review.

Perhaps at the heart of this question is a concern about the quality of the research being conducted. Being scientific doesn't imply that the research is high quality or substantively correct. That being said, the methods being used are similar to the clinical methods used in my current field (education) and other fields in which I am familiar. I would suggest that carrying over your own field's lens of what constitutes proper research methods may not be appropriate to other fields.

3

A study like those you describe (i.e., gathering statistics about some small population and doing statistical calculations on it) is by definition scientific because it is falsifiable. A reader can either figure out that the conclusions have been wrong; or a future study with identical constraints can get different results.

To be scientific literally means nothing except being falsifiable. A pseudo-science is defined by the fact that its postulations are not falsifiable.

Whether the fields you mention produce "good" results is a completely different question, and there may be many reasons why one would be disappointed about that. Some areas of research might just have a hard time scrounging together the money to perform "good" (=expensive) studies. Some areas might just not have enough scientists being interested in them. And some might just be exceedingly hard fields of study.

Also, you seem to imply that there are "only" statistical studies, and no work approaching these topics from "first principles" (maybe dissecting muscles, looking at them through microscopes, building computer programs which physically simulate the workings of muscles in a non-statistical way and so on and forth).

All studies involving humans are notoriously hard to perform in a good manner. Human (and other) bodies are extremely complex systems of interlocking parts, and very rarely is there some aspect that has a simple cause-and-effect chain. Relatively few aspects are up for inspection; much is hidden below our skin, or within our cells - which it is hard to inspect for practical reasons (if you take them out, they are not part of the processes anymore; and you can hardly inspect the exact same cell multiple times to see how it specifically changes...).

So, TLDR: the scientific community is doing its best to tackle these problems with the scientific method; and statistics are the most approachable method we as humanity know today. Statistical results do need a lot of interpretation and context, which does not make it any easier to publish these results to the general public, and many famous errors have been done when taking a result of a study and publishing it in non-scientific newspapers.

Grotesquely, there is pseudo-science surrounding physiology, which often actually works better than real science: having a grizzled old coach with 50 years of experience training 100s of athletes probably has better end results than reading 100 scientific papers. That coach might tell you plenty of advice you will find in no paper, and find no reasonable scientific justification for, but it might still simply work anyways. But unfortunately the knowledge of that coach is distributed somewhere in his or her brain matter, and very hard to extract in a concrete manner. :)

-1

I understand why you might think that. Sciences like physiology and psychology study very complex systems, that is the human body, and for this reason they have to rely on statistics to conduct research.

But statistics has two major flaws.

First it requires a large sample and one needs to consider as many variables as possible to be accurate, the studies conducted on muscle growth work with really small samples, like 100-200 people which is a joke. Hence the contradicting studies.

Second, statistics can't explain the "why" it does give you a relationship between variables, but one cannot know why this relationship is valid, for example if a study shows that sprints build more muscle that jogging, we know what happens but this doesn't explain the deep reason of why it's true, we won't know what happens to the muscles and what are the mechanisms that cause this.

But there is also another problem, physiologist don't understand statistics at all, they just throw the data in R and automatically get the results. One needs to really understand statistics to get the correct answers.

Moreover I would like to address the fact that many of those studies are biased in order to promote particular equipment or anything the "researcher" can use to make profit.

Now to answer your question, physiology is not a pseudoscience because they are using the scientific method, but it is not a good science and the results are often questionable.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.