The past 5 months have been some of the busiest in my life, and that’s saying something. To put it into perspective, since the beginning of September,
That last part is what I wanted to talk about today.
(NOTE: if you want to attend the seminar, I’m presenting it in Vancouver May 2nd, just click HERE for more info and to register)
In total I’ve looked through about 400 scientific publications on mobility, tissue adaptive qualities to stretch, neurophysiology, and all sorts of stuff related to how to become more bendy and supple in life.
Most of the research done on tissue adaptive qualities comes from non-human subjects, primarily due to the pesky need to dissect the tissues to see what’s happened after the interventions. It’s tough enough to get studies through ethics without the unfortunate challenge of killing the subjects, so researchers will use mammals that have somewhat similar tissue qualities as humans, maybe a faster relative lifespan to accelerate the results they get, and then publish those studies.
The obvious challenge is that as these studies aren’t done on humans, the end results have to be inferred as being similar as with humans. We could pick apart why this is a step away from true human observations, but for the time being, let’s just say that indirect assumptions are likely less than totally accurate when it comes to comparing the hind leg hamstring mobility of a rat with that of your grandma trying to get better at gardening with a few stretches.
Because they may be similar doesn’t mean they’re the same.
For the studies that directly look at stretching interventions, there’s usually a few very BIG methodological issues that come out of them.
Pre-Post Test Specificity
It’s common to see testing in mobility studies look at using something like a sit and reach test, a test where you overlap your hands, keep your knees straight, and slide a dial forward as far as possible to indicate posterior chain flexibility.
A challenge with this type of test is the number of confounding variables that could influence the total body mobility of the individual, not specific to the intervention being used. For one, there’s no definition of what tissues may have been adapted during the intervention and how they would influence this kind of test. If I stretch my hamstrings all day long, but the limit to me rolling forward is spinal chord tension or a hard end point to bone on bone hip flexion contact, my score will be limited. Contrast that with someone who gets a parasympathetic response to the stretch intervention and feels more at ease with the tester on the second round, and without any structural changes to their hamstrings at all, their scores will be better.
Essentially, while a test like this is fine to use in large scale fitness assessments like the military or school settings, using it to determine tissue related changes to interventions may make it too non-specific. It’s a shotgun approach when we want to use a sniper rifle.
Now I completely get that there’s a significant financial challenge to acquiring imaging devices or more technical stuff on incredibly tight budgets where funding models are trying to go towards stuff like cancer or diabetes or figuring out how to make the military more deadly, so finding cash to pick up a flouroscopy machine for a stretching study is likely going to be a tough sell, but this is just me thinking in ideal situations, not necessarily the reality of a grad students thesis on a time crunch.
Inadequate Definition of Interventions
Most of the studies I’ve seen on stretching interventions are incredibly open-ended and don’t look at any level of standardization within participants. They’ll outline a stretch routine you would see posted at your local commercial gym for members to give a shot, and then reassessed after X number of sessions, weeks, whatever. Some may get deeper and say they have guided mobility from a certified trainer or therapist.
Sounds good, right?
The challenge is that in a scientific study, details matter a lot. Were they going to any specific joint angle, perceived discomfort, pre-determined sarcomeric length, or some other factor that would be much more specific, or was it just “hey do this stretch, hold it for 30 seconds, and think about all of the wonderful things you want to buy on Amazon after your workout”? I only saw a small handful of studies where an intervention looked at involving a specific joint angle or relative position, and those were for the ankle and elbow and in injury recovery situations.
If I’m just left to stretch to what feels comfortable for me, it’s going to be entirely different than someone who doesn’t have the same history of mobility training or pain tolerance. This isn’t saying I’m amazing with my awesome training and superlative pain threshold. There are a lot of people out there who have WAAAAY more mobility training history than me and a significantly higher pain tolerance than I do, and a lot of people with way less of both, but comparing our approaches will be very inconsistent in terms of scientific rigour.
For a study to compare apples with apples, you’d have to design interventions to be specific in terms of relative joint angles to pre-testing scores, specific levels of discomfort the individuals have to get into without exceeding (which can be very hit or miss when it comes to perceived discomfort), positional specificity to avoid cheating the movement or pushing harder than outlined, and even using imaging to make sure the interventions are consistent with one another.
In scientific studies, blinding refers to whether the testers or participants know what’s going on. If I have a trial for a new medication, I could blind the participants as to whether they’re receiving the medication or a placebo to rule out potential placebo effects.
the same thing can happen with the testers. They may know you’re in the intervention group, and coach you to get a bit further into the post test group, or assess you in a slightly different way to coax a different outcome. Most studies are done on shoe-string budgets so getting funding for a blinded tester is kinda hard, and most are run by exercise physiology grad students toiling in their supervisors publication mines, so it’s understandable that they don’t do much in terms of specific blinding, but it does have an impact on the results.
Ideally, the testers would have no idea whether the individuals in the study were in the intervention or control group. They’d just line up, assess the person, record the results, and move on, not knowing whether they were getting stretched out or just hanging around doing nothing. Often though, the same person doing the testing is also doing the interventions, so they would hopefully have some idea of what’s happening.
This isn’t to say all studies are falling into this concept, but there’s certainly a lot of them from what I’ve seen, or what hasn’t been stated in the methods of the papers I’ve looked through.
Because of a lot of these limitations to the body of research available, you see a ton of studies with huge variations in outcomes. When the standard deviation is larger than the effect, you know there’s some seriously messed up business going on. It’s hard to believe that participants in dedicated stretching studies would spend weeks on end doing their intervention and wind up with LESS range of motion, but here we are.
So most research in stretching and mobility has to be taken with a really big crunchy grain of salt, as the methodologies are often kind of all over the place and the specific outcomes can be really varied from one study to another depending on how rigorous they are with their intervention specificity and testing, but hopefully the field will just continue to refine and improve over time and we can get more info on how and why mobility actually improves, plus specific recommendations as to what to do for the best results. In the meantime, keep moving and have fun with it.