Audience: Health professionals I used to be a researcher (exercise biomechanics, physiotherapy, chiropractic) - one of my goals was to quantify how hard muscles worked during different exercises. This was important for determining which exercises may be best for targeting a certain muscle or determining how modifications to exercises (e.g. doing it barefoot or on a wobbly surface - for a simple paper look here) changed the targeted muscles response.
I used surface EMG which quantifies the electrical activity of that portion of a muscle that was under the electrodes. Surface EMG is messy and you are required to process the crap out of it to get something meaningful.
The problem is that you can't just say that because there is 56 volts (this is an arbitrary unit you might as well say 56 bananas) of activity in the vastus lateralis in Mary and 24 volts in Mary's vastus medialis that Mary is somehow deficient in her VM muscle. But people do this and they are wrong. Don't get me started on the chiropractors that use surface EMG scans of people's backs to tell them they are dysfunctional - that is a lot of poppycock.
What also happens is the huge amount of variability across people and within people in EMG readings that is not related to how active a muscle truly is. This means that just because a muscle looks like it works very hard (i.e. lots of EMG activity) in many people does not mean it works that hard in everyone. Further, there is a lot of variability across people in how they respond to different exercises - we have to be careful in how we generalize research to different patients. For example, tall guys with short torsos get very little abdominal challenges with modified curl ups (I had a subject with less than 15% of max whereas others would have 40-60% of max).
A specific example EMG Research that you may want to question
A big problem with applying and interpreting surface EMG is the lack of a reader's and researcher's healthy skepticism. I will now pick on a paper that has always bothered me. This study was published in the JOSPT by DiStefano (2009). The paper looks good, I love the concept and the execution, the methods look totally appropriate but it does not make any sense.
The authors aimed to determine what was the best exercise to activate the Gluteus Medius. This is an important muscle for spine and hip stability and is often implicated as problematic for Runners. Groups of physiotherapists have been touting its importance for years. But what the authors concluded does not make sense nor fit with any research before them.
Brief Paper Findings and my perceived flaw
The authors found that the best exercise to activate the hip abductors was the side lying leg raise (hip abduction). They found that the average muscle activity was 82 % (Standard Deviation = 42) of the subjects maximum. This is crazy. I have nothing against the sidelying leg raise but it does not recruit your hip abductors to more than 82% of your max. And the standard deviation was 42, suggesting that some people were over or near their maximal ability. Yet somehow they completed 8 repetitions. That in and of itself is a huge give away that your hip EMG stinks. Something is wrong if your EMG process tells you someone is working close to near maximum yet they can do 8 repetitions. That is a big red flag. Yet this paper got published, someone got a PhD and there was no mention of this simple possibility.
Further, previous papers by Ekstrom et al (2007) and Bolgla and Uhl (2005) investigated the exact same muscle during the same exercise. These two different research groups found muscle activity between 39 and 42% of maximal activity. No mention was made in the discussion of these two other papers explaining why the 20 or so subjects had muscle activity nearly twice what others had found.
Take home point
This looked like a very well designed study. It just made no sense from what others had previously found and what our senses tell us about how hard an exercise feels (do a leg raise on your side - its not that hard). This finding should certainly warn us about the limitations of research and that you certainly should not believe or accept what is published.