One of the common questions that comes up here is something like, "Please direct me to the studies that prove that _____ is the best or most effective exercise," or something to that effect. It's as if they think, "there must be studies about this, and the people on EXRX must know where they are, even though my own research can't find any." A recent example started out like this:
Louie Simmons mentioned somewhere that he often does light weight for a ton of reps (sometimes hundreds) in order to increase blood circulation in the selected area and improve healing of muscles and tendons (for example, he would do long sessions with indian clubs to rehab/prehab his shoulders). I still haven't found any study backing that.
Does somebody has resources on this potentially very important subject?
After a couple of people said that no, they didn't know of any such studies (I personally have MedLine memorized, so I can say for sure that there aren't any!) the OP said:
[quot]Don't you think it would be interesting to find out the optimal rep range/time length/intensity for recovery? I'm all for following what seems to give good results among top athletes, but studies can give additional insight.[/quote]
Yes, studies CAN give additional insight, but often they don't. A badly-done (i.e., the majority) study is worse than no study. Many studies in exercise science are very small, and provide very little statistical power. The truth is that studies often are not practical or even possible in the real world.
Many studies, like the one the OP proposed, are very impractical. Think about the question implied in his suggestion: Does light exercise involving and injured part, done shortly after the injury, speed healing compared to rest? I think that's what he was asking, at any rate.
Let's think about what it would take to do the study. For us to keep it simple here, let's just study light exercise (3 sets of 30 reps at 30% 1RM). If you want to study a range of each variables, the complexity mounts fast. If you want to study 3 rep ranges (say 10, 30 and 50 reps), 3 time periods (2 days, 5 days, 10 days), and 3 intensities (20%, 35% and 50% of 1RM), you'd need you'd need 28 study groups (3 cubed plus a control group).
First, you would have to decide who is the study population? Young, trained athletes? (This is common, since many studies are done at universities, and the sports teams are a popular source for subjects.) Young untrained exercisers? Sedentary middle aged new trainees? The elderly? It's an important question, because it would affect how the study is conducted, and it could affect the outcome. Just because something works in young athletes doesn't prove that it works for everyone. It would also determine where and how you would recruit the subjects.
Let's do a hypothetical study. I'd think (without remembering my undergraduate statistics) that you would need to study several hundred injuries in order to have a valid, powerful study. For our study, let's shoot for 100 injuries treated with light exercise, and 100 treated with rest, the control group. We'll need 200 injuries, and randomly assign them to either of the two protocols. That's just for our simple study. If you want to do all the variables, you'd need a good number in each of your different study groups, so that would multiply everything I'm going to suggest! If you still want to have 100 per group, that's 2800 injuries! If you only had 20 per group, you'd still need 560 injuries. Many studies that we see use even smaller groups than this, so even if you had 10 per group, you'd need 280. I doubt that would give much statistical power.
Second, what kind of injuries are we going to study? Obviously, we aren't proposing that we treat compound fractures with light exercise, are we? So, what then? Muscle strains? Sprains (ligament injuries)? Would a rotator cuff tear be included? Would a hamstring strain? Tendon avulsions? We'll have to define things in advance.
Third, how many subjects would you need? You'd have to think about how frequently injuries occur in your study population. If we're studying young trained athletes, and we find out from other studies that there are about 2 of whatever type of injury we've decided to study per 100 athletes per year, then we'll have to recruit 10,000 young trained athletes for a year's time, 20,000 if we expect to complete the study in 6 months. Whoh! That means that we'll have to involve colleagues in many universities to collaborate in the study. People from all those universities will have to travel to a central meeting point to plan, and later to evaluate the study. Everyone involved in doing the study would have to be trained, so that injuries could be evaluated and defined consistently. We'll have to have ways to make sure that all the athletes in the population report all of their injuries. And remember, that is just for the exercise-no exercise version of the study.
Maybe we should study old, sedate people, in whom the injury of interest occurs 10 times per 100 trainees per year. Then we'd only need 2,000 subjects to follow for a year, 4,000 for 6 months. We could contact dozens of community centers where they have senior exercise programs. Our staff members would have to visit each site to make sure that the staff there understands the study, and how to document the injuries of the subjects. They'd probably have to go back from time to time to keep the local staff up-to-date.
Fourth, how are we going to fund the study? There will be travel costs, staff time entering data, and compiling reports. If some of us are professional researchers, there will have to be funds to contribute to our salaries. We'd have to have insurance, including liability insurance in case a subject felt that they suffered permanent harm because of how they were treated following their injuries ("Your honor, it is common knowledge that light high-repetition exercise is vastly superior to inactivity after this kind of injury, but yet these researchers callously convinced my poor client to just sit around for a week after the injury, and now he can't move at all.") So, who is going to pay for the study? Either a government or other non-profit agency, or else someone who stands to make money as a result of a particular outcome. Grants from the non-profit sources tend to be small, and are usually only given for studies that are judged to be of great value to the public. I can't think of anyone from the second category who would be interested in this study. So we probably aren't going to be funded, and there will be no study, and we'll have to go on relying on Louie Simmons, Mark Rippetoe and KennyC.