Earlier this week I posted a link to some excellent videos by Joe Sansalone in Part 1 of this series.  In Part 2 I posted a discussion I had with Joe himself covering some of the questions that these videos generated for me.  Today, my hope is to wrap up with some final thoughts on these videos and how I intend to incorporate this information into my programming.

Perhaps the single most important thing I took from this series is that in our attempt to be efficient and do the whole YTWL series, we glaze over the fact that most people hardly have the required motor control to do any one of these movements on their own.  In doing so, we reinforce the poor motor patterns and let the scapula migrate upwards during the prone Y due to the upper traps instead of maintaining focus on scapular depression and upward rotation created by the lower traps.  Moreover, we make things worse by sometimes adding weight as soon as a person can hit the prescribed number of reps.

Being a former biomechanics geek, I feel that Joe has a good point about electrode placement affecting readings of muscle activation in any movement.  I also agree that individual performance of any movement will result in some variation in which muscles are recruited.  However, EMG can be a tremendously useful tool and discounting muscle activation studies based on these points might be a little neglectful.  After all, EMG is a huge component of Dr. Stuart McGill’s spine model and few people question this element of his research.  Granted, if you’ve seen the mathematics involved in this model you’d have to be Einstein just to make your argument anyway.

I still believe that performing a wall slide facing the wall will activate the serratus to a greater degree than the lower traps based on EMG and the angle of pull of both muscles.  Raising the arms from the wall at the top of the movement would certainly involve the lower traps if performed properly.   To me, this might be a great way to integrate both movements together to promote upward scapular rotation (a function of both muscles) in a very early progression.

So how would I program the prone Y?  Given what I’ve learned from Joe, I think that I’ll likely begin by performing the Y alone or in a pairing it with another movement promoting upward scapular rotation such as scap push ups or something similar.  Otherwise, I might perform it as part of an overall warm up circuit involving various other activation exercises and mobility drills.  Although I do tend to do some mobility/activation work between sets of exercises, I think that I’ll refrain from this with the prone Y (at least in the beginning) since people might be inclined to just hammer their way through it when their adrenaline is pumping instead of paying attention to the motor control element.  I’d also keep the reps low (between 5-8) to emphasize the importance of taking it slow and getting the movement right.

As far as progressions go, I figure that I’ll probably work from the Y through the rest of the YTWL (actually Nick Tuminello’s LYTP) series one by one before adding range of motion.  Only once this has been achieved would I consider stringing the movements together for strength endurance and finally adding weights.  Of course, some will move quickly through this progression and others will be slower, but that’s what I’m thinking right now.

What are your thoughts?  Would you do it differently?  If so, why?

When I was back in university I have to admit that I was a little bit of a biomechanics geek.  These days I like to think I’m just more of your normal all around kind of geek, but I digress.

Having spent a lot of time in a biomechanics lab I had the opportunity to do my fair share of EMG analysis.  And although it has been quite some time since I’ve done any hands on EMG work, I’d still consider myself versed enough to share a little bit about it.

But why are you going to bore us to tears with all this EMG crap?

Because I think that recently there has been a lot of talk about EMG for the determination of optimal training movements and I don’t think the general training public completely understands this enough to make a decision about whether this is valuable or not.  Instead, I think most people just believe that you slap on the electrodes, contract, and look at which muscles were most active.

Here’s a typical rundown on how EMG testing is really done in a biomechanics lab.

- EMG electrodes are placed on or in the muscle belly of the muscle group to be recorded.  Note that you can use either surface electrodes that just stick on the skin or fine wire electrodes that actually stick right into the muscle.

- The segment to be examined (leg, arm, etc)  is strapped into a jig of some kind to allow the researcher to measure the torque created when the muscle contracts.

- The muscle is contracted through a certain range of motion or for a certain amount of time and the EMG and torque are recorded.

- Since the EMG signal is recorded in millivolts it needs to be amplified before it is recorded.

- At this point, the raw EMG signal is atually quite messy looking and if you were to try and make any conclusions from this you’d be completely out to lunch.

- To make the EMG more usable it needs to be full wave rectified which basically means taking the absolute value of the signal.  This will put the whole signal on the positive side of the line.

 -  Then the signal needs to be filtered to take out all noise that might be impeding you from seeing the actual signal you’re trying to get at.  A filter is often used to filter out electrical noise (introduced by the electrical equipment), electromagnetic radiation, and motion artifact(often introduced by swaying wires as the limb moves during the trial).

- Once you’ve gotten this far you SHOULD have a relatively clean EMG signal, but you’re still not done.  In many cases people will average or integrate the EMG to get a clearer picture of muscle effort.

In the picture below you can see the progression from the raw EMG, to full wave rectified, to filtered, and then integrated.

 This signal can then be compared to the torque measured from the limb to establish a relationship between the amount of muscle activity and the amount of force that can be generated. 

So more EMG activity is equal to more muscle force?

Not exactly.  And this is where the problem lies.  Unfortunately the relationship between muscle force and EMG is not linear.  In other words, when EMG goes up, muscle force does not necessarily increase at the same rate.  One does not clearly relate to the other.

Why does this happen?

There are a few things about EMG that make it tricky (even for the best researchers). 

1.  Crosstalk between muscles often occurs when an electrode covers an area where several muscles are located.  For example, electrodes placed over the bicep will record bicep activity, but they can also pick up the signal of the brachialis which is deep to the bicep.  Fine wire electrodes that go directly into the muscle can decrease this, but many studies opt not to use these.

2. EMG best predicts muscle forces during isometric contractions.  Of course, when it comes to exercise, we want to look at movement so this creates problems.  When the arm moves, the muscle can move beneath the electrodes which means that different parts of the muscle are being picked up for the EMG signal.  This can make the prediction of muscle forces difficult.

3.  The real truth is that when you’re measuring the force of the bicep curl, you’re not really measuring the force of the bicep muscle itself.  You’re actually measuring something called torque which is a product of the force of the bicep muscle and its moment arm with the force at the hand and its moment arm.

Say what?

In non-geek terms, it means you have to calculate the muscle force of the bicep because you can’t really measure it directly unless you were to attach some sort of transducer to the muscle itself.  (I should note that a researcher named Paavo Komi used to do this buy surgically implanting a buckle transducer on his achilles tendon).   Sadly, since other muscles cross the elbow and contribute to flexion, the calculation of muscle force is damn near impossible without an elaborate computer program.  If you’ve ever seen the math behind Dr. Stuart McGill’s muscle force predicting model you know how crazy this can get.

But what if we had such a program, could we compare the EMG to other exercises?

You sure could, but again, you need to remember that when you’re comparing different dynamic exercises that are near maximally loaded with surface electrodes you’ve already introduced all sorts of potential error.  If you did want to do it though, you’d need some way to standardize the results. 

What researchers typically do is have the person do a maximal voluntary contraction and compare all EMG results to this.  For example, if you were doing two different bicep exercises you might say that the dumbbell bicep curl yielded 90% MVC whereas a pronated curl produced 65% MVC or something of the sort.

Not doing this can actually lead to erroneous results.  One such case was when it was reported that certain exercises hit the upper abdminals and others hit the lower abdominals to a greater degree.  However, after the signals were presented as a percent of MVC the differences disappeared.  In other words, this demonstrated that you cannot differentially train the upper and lower abs.

I guess my main point here is that while EMG is an incredibly useful tool in the hands of some researchers, it can lead to terrible confusion and inappropriate recommendations in the hands of others.

Always be open minded, but maintain some degree of healthy skepticism of whatever you read…except my blog…which is perfect.  

.

Want to understand how research is done?  Check out my product How to Read Fitness Research.