First things first, please give this post a little time to get rolling. There are bits of it that are primarily for nerds like myself, but there are also directly actionable parts, so be patient while we get there.
You may have heard of EMG before. EMG stands for Electromyography – essentially measuring the electrical activity in your muscles.
Muscle contraction starts with a nerve impulse. If the nerve impulse is strong enough, it creates a small electrical current that runs down the muscle (wave of depolarization), setting off the chain of events that leads to muscle contraction (excitation contraction coupling). The harder your muscle needs to contract, the more muscle fibers will be activated in this manner, so the stronger the electrical current measured in your muscle will be.
That’s the basis of EMG. If you can measure the electrical activity in the muscle, you have a good idea of how hard that muscle is contracting and how many muscle fibers are being used. So, a higher EMG measurement means more muscle fibers used, which means a greater training effect, which means more strength gained, right?
Well… not so fast. That’s the story you’ll hear from people who fully endorse the use of EMG as a primary factor in assessing exercise effectiveness. It sounds nice in theory, and it “feels” good for an uninformed reader. 100 people may have 100 different opinions, but with EMG you can actually measure something, and get a nice objective number you can use to compare one exercise to another.
In that regard, overreliance on EMG can really give you a false sense of security.
There are a few issues with this, though.
First and foremost is that EMG readings are relative to load, not just exercise. Let’s say you bench press 30% of your max, and then bench press 80% of your max. The 80% load will give you higher EMG readings than the 30% load. So EMG isn’t necessarily telling you one exercise is better, but that a given load with a given exercise may be better (i.e. comparing an 80% bench press load with an 80% incline press load may be more appropriate). Of course, extending this idea further, you’ll also get higher EMG readings with 100% of your max than with 80%. So if you’re relying solely on EMG, the logical assumption is that you should do singles with your max and nothing else, which is obviously absurd because it doesn’t take training volume – the much larger driving force – into consideration
Furthermore, to normalize EMG data to compare results for different people, you have to base it on something – you typically don’t just use raw electrical data. That “something” is the Maximal Voluntary Isometric Contraction (MVIC). An MVIC is essentially how much muscle activity you can attain without actually moving a load. By definition, your MVIC is 100%MVIC. However, the basic recommendation that goes along with EMG data is that exercises and loads that produce average EMG readings above 60%MVIC are probably going to be good bets for increasing size and strength. As an illustration of this, average %MVIC for squats at 70% 1rm (an exercise and load that we can all agree will build quad size and strength) averages around 60-80%.
If EMG was all you used to compare exercises, you’d think that training your quads purely with isometrics was better than squatting with 70% of your max.
We know that volume, load, and range of motion are all major contributing factors for getting bigger and stronger, but EMG doesn’t account for these by itself.
With that brief primer on the function and limitations of EMG out of the way, we can dig into the study at hand: Bench press and push-up at comparable levels of muscle activity result in similar strength gains
The first thing that jumped out at me, right in the abstract, is this statement which is pretty damning for people who have put too many eggs in the EMG basket to this point:
“While researchers assume that biomechanically comparable resistance exercises with similar high EMG levels will produce similar strength gains over the long term, no studies have actually corroborated this hypothesis.”
However, from there, things really start looking up for the usefulness of EMG.
There were 30 subjects (22 men and 8 women), mostly in their early 20s, who weighed ~70kg. They were lean (average bodyfat about 14%), healthy, and had been training an average of 2 years (1 year minimum). They met the NSCA’s definition of “advanced trainees”: minimum of 1 year of resistance training experience, performing at least 3 strength training sessions per week at moderate to high intensity, and currently involved in strength training (more on that later).
After 2 familiarization sessions, the subjects worked up to a 1rm bench press on their third session, and during their 4th session, they worked up to a 6rm for both band-resisted pushups and bench press (half of them did pushups first, and the other half benched first to make sure exercise order didn’t affect the outcomes). During their 6rm test, EMG readings for their pec major and anterior deltoid were recorded.
After that, they split the subjects into 3 groups. One group trained bench press for 5 weeks, one group trained doing band-resisted pushups for 5 weeks, and the third was a control group.
The pushup and bench press group both trained with the 6rm load they established at the beginning of the study, doing 5 sets of 6 each training day, twice per week.
The researchers controlled for basically everything that could have affected the outcomes. Volume and intensity were identical, rest periods were timed (4 minutes between each set), hand placement was standardized (50% wider than shoulder width) and enforced every set, rep cadence was standardized (2 seconds up and 2 seconds down. Read how intended rep speed can affect results here) ROM was dictated (elbows to 90 degrees. So at the bottom of each rep, the bar was a couple inches off the chest/the chest was a couple inches off the floor for most people. However, if you’re a short-armed barrel-chested grizzly bear like myself, that’s your normal bench stroke. Yes, the bar rests on my chest when I do floor press.), and bench was done in the smith machine to make sure technique was similar between subjects (i.e. there wouldn’t be differences from some people flaring their elbows and others tucking their elbow).
I’m sure some people will read “didn’t touch their chest” and “bench in the smith machine” and throw the baby out with the bathwater. But that’s how good research is done – control as many variables as possible to make sure that what you’re studying is the thing making the difference, not some unrelated variable that wasn’t accounted for. However, I would be the first to admit that what we’re getting out of this study (as with most studies) is a principle, not direct application.
After 5 weeks, they hit new 1rms and 6rms on their bench press.
The EMG differences between bench press and pushups were negligible for both the pec major and anterior deltoid.
Additionally, there were no significant differences between 1rm and 6rm bench press for the two groups at the start.
The pushup group and the bench press group saw similar strength increases for both 6rm and 1rm. The control group did not get any stronger, which should be expected. However, this study did something interesting – the subjects were currently training when they started this study, and were told they could continue training as they currently were, so long as the bench press and pushup group didn’t do any pec-dominant exercises or pushing movements. We can assume that the control group, however, WAS still pressing in their training. So the fact that the control group didn’t get any stronger during the same time period that the two intervention groups made substantial improvements speaks to the inadequacy of their training program.
This study shouldn’t be taken as a ringing vindication of using EMG at all times for all purposes, but it did show that EMG can be used to draw a comparison between two biomechanically similar movements. However, more dissimilar movements (i.e. squat vs. deadlifts rather than bench press vs. pushups), or movements that are going to be performed with different volumes (i.e. trapezius activity between Olympic lifts that you may do for 10-20 total reps in a workout, and shrugs that you may do for many more reps), or through different ranges of motion (i.e. comparing hamstring activation in the RDL vs. the glute ham raise) should still be approached with caution when trying to compare using EMG.
Also, I just want to point something out if you missed it in the “procedures” section. They found these peoples’ 6rms, and made them do 5 freaking sets with the same load for the same 6 reps. This isn’t one of those exercise science studies where they put the subjects through a foo foo training program that doesn’t have any relevance to the real world. 5 sets of 6 with a true 6rm would be somewhere between very difficult and impossible for most people. Since they could actually perform all 5 sets, I think that speaks to their initial training status, though. With less neural efficiency, they couldn’t induce as much fatigue on each set, so they could keep the same intensity across 5 sets, whereas if a 400 pound bencher hits a 6rm of 330 or so, I can promise you they aren’t doing 4 more sets of 6 with 330 in the same workout.
Another interesting thing about this study (though I’m not entirely sure what you’d do with this information) – they didn’t apply any sort of progressive overload. They had data that the initial 6rm loads resulted in similar muscle activation, so those are the same loads they stuck with for the entire study. However, the average strength increase was ~20% in spite of doing the exact same workout twice per week for 5 weeks. If someone’s fairly untrained to begin with, a fancy lifting routine isn’t required to make them stronger. Heck, progressive overload isn’t even required to make them stronger. Just making them work hard (and 5×6 at a 6rm load certainly classifies as “working hard!”) is enough to make them stronger.
In terms of an interesting finding about EMG research itself, from the study:
“…using EMG levels below the threshold of 60%MVIC has been considered ineffective to produce strength adaptations, another relevant and novel finding in our study is that lower EMG values (i.e., 52%MVIC) induced a high intensity stimulus, which were adequate to produce muscle strength gains.”
As an aside, assuming a threshold of 60%MVIC was required for strength gains, that would mean the bench press with 6rm loads (~85% 1rm – 52.7%MVIC in this study) would have been ineffective in strengthening your pecs. Sorry Arnold, for science has spoken. Hopefully, this study will cause people to reevaluate applicability of EMG research, since there are so many other factors in play.
In terms of exercise selection – if you’re going on vacation or your gym is closed for a holiday that coincides with Monday (international bench press day), don’t lose any sleep over it. Just get a heavy elastic band to provide resistance and you can still get a decent faux-bench workout in. It would be a stretch (get it? Elastic band? Bad joke, and I’m not sorry) to assume it would be as effective for you if you’re more highly trained than the subjects in this study, but it’ll do the job well enough that you don’t have to worry about losing all your hard earned gainz.
Another point worth bringing up is the difference between muscular strength and efficiency of movement. Since these people clearly weren’t training optimally (as evidenced by the control group that continued to train how they were previously without getting any stronger), they probably didn’t have an incredibly efficient bench press stroke. So getting stronger at pushups, since they’re similar enough, transferred directly to a bigger bench at the end of the study. However, if you’re already very efficient, the transfer probably wouldn’t be 1:1 as it was in this study, even if you strengthened your muscles to the same extent doing a similar movement. We *may* be seeing that starting to occur with these subjects already via the % change in 6rm vs. 1rm. Both intervention groups increased their 6rm by essentially equivalent percentages (21% vs. 22%), but the bench press group increased their 1rm by a larger % than the pushup group (14% vs. 20% – though the difference didn’t quite reach statistical significance), perhaps indicating slightly larger improvements in efficiency.
Finally, and most importantly from an application standpoint, if you want to generally improve your upper body strength, or you’re training someone for a sport other than powerlifting, band-resisted pushups can get someone stronger effectively (more effectively than however these study participants had been training for the past year or more, at least!). Since they allow the scapulae to move through their normal range of motion (instead of being pinned to the bench) as well, I’d argue that heavy band-resisted pushups (along with other exercises that let your scapulae move freely like push press, dips, and landmine press) will probably improve performance more than bench press will. Just my 2 cents.
• • •