Caught Looking: Jaw Clenching, Predicting Elbow Varus Torque, UCL Surgery and Command

Here are three studies the R&D department at Driveline read this week:

The Effects of Concurrent Activation Potentiation on Bat Swing Velocity of Division II College Softball Athletes

Potentiation is often discussed in sports literature, more commonly in terms of training using post-activation potentiation.

The definition of potentiation from Merriam Webster is, “to make effective or active or more effective or more active; to augment the activity of (something, such as a drug) synergistically.” Post-activation potentiation is the idea of leading one exercise with another that will theoretically improve the performance or outcome of the trailing exercise. 

Concurrent Activation Potentiation (CAP) is the idea of activating some contraction or muscle activation while the activity you want to optimize is being performed. One practical example of this used is jaw clenching or jaw opening while an activity is being performed, such as lifting weights. 

Mace et al. at Florida Southern College looked into the effects of CAP on bat speed of 13 NCAA softball players by having them test their bat speed off of a tee in two different conditions: while clenching their jaws and one while they relaxed their jaw muscles. 

The order was counterbalanced so that the effects that might be observed from the order of the two conditions were minimized. Using a simple paired t-test, the authors found that the group mean bat speed was significantly higher in the jaw-clenched condition than the jaw-relaxed condition (62.7, 65.8 mph). 10 of the 13 participants observed an increase in mean bat speed from the relaxed condition to the clenched condition. 

This study is awesome because the results are clear cut and the readers can easily apply the results. However, the idea of a clenched jaw improving athletic performance and strength/speed output is not something I imagine is surprising to many. 

Many performers will make a “game face” or grunt, clench their jaw, contract other face muscles, clench fists even in a lower-body activity, etc. Maybe this factors into why some people have seen performance increases from wearing a mouthpiece while performing—because it allows them to clench harder more comfortably. Regardless of the cause, this idea is very cool and deserves to be considered for those interested in optimizing athletic performance.

Effect of Ulnar Collateral Ligament Reconstruction on Pitch Accuracy, Velocity, and Movement in Major League Baseball Pitchers

Despite the increasing number of Ulnar Collateral Ligament (UCL) constructions in baseball, the performance effects of this reconstruction procedure, if any, are not fully understood. Some say that pitchers come back from rehabilitation throwing harder since they have a new ligament, or they come back with worse command. McKnight et al. in Los Angeles used MLB in-game data paired with injury data of MLB pitchers to investigate performance differences before and after UCL surgery. 

This study included 39 pitchers who underwent UCL reconstruction surgery in 2011 or 2012. Fastball velocity, baseball accuracy, and curveball movement were recorded and compared between the one year pre-operative and three years post-operative. Fastball velocity and curveball movement are pretty straightforward measurements, but fastball accuracy is a tough metric to measure with confidence. This study used COMMANDf/x, which incorporates pitch characteristics and the location of the catcher’s glove with the pitch is released. 

Out of the three measures included in this study, the only one that was significantly different between pre-op and post-op was fastball accuracy. There was just over a 9% accuracy degradation one year and three years after UCL reconstruction for fastball accuracy in this sample, both of which were statistically significant difference. However, fastball accuracy two years after UCL reconstruction was not significantly different—but was about 4% worse than pre-op on average. Fastball velocity and curveball movement did not exhibit significant differences before and after UCL reconstruction, despite the confounding factor of age. 

There could be something to the idea that command gets worse after UCL reconstruction. With different tissues in the arm, movement, stability, and strength may be affected—especially since accuracy was worse on average in all three years post-op.

However, the validity of the COMMANDf/x measurement needs to be considered since there are a lot of confounding factors. Measurement variability aside (error in measurement), we cannot be sure that using the position of the catchers’ gloves when the pitch is released as a measure of the pitch target is a valid measurement. There is a lot of movement in the position and orientation of catchers’ gloves before receiving a pitch. If the variability in the pre-receipt glove movement of the catcher is not random, it will show up in statistical tests, confounding the command measurement. 

It’s not to say that these results are invalid—the measurement quality of fastball accuracy just needs to be considered when interpreting these results. I think these results warrant a more controlled look at accuracy pre- and post-UCL reconstruction. A study like this would be tough to execute because it would require a prospective design that would involve measuring the accuracy in a controlled setting of many pitchers since the researchers would not know who will need UCL reconstruction in the future. This is the main reason injury research is so difficult. 

Aside from command, the null results (no difference) of fastball velocity and curveball movement are both interesting in their own right. One thing that should be noted is that comparing performance after surgery to the year before surgery could have limitations. Pitchers may not be achieving peak performance the year before surgery if they are technically injured.

A closer-to-perfect comparison, in my opinion, would be to compare each pitcher’s post-surgery performance to a measure of their peak performance in their career. However, I still think these results show how successful this operation has been for keeping athletes in the game longer since they can consistently return to their pre-UCL reconstruction performance. 

Stride-Phase Kinematic Parameters That Predict Peak Elbow Varus Torque

Tanaka et al. in Japan collected an impressively large dataset to test for biomechanical parameters that are associated with peak elbow load. 107 high school pitchers with an average age of 16.3 years were included. 

This group used multiple linear regression to test the association of 26 different parameters with peak elbow varus torque. The authors ended up with a model that included five different parameters that explained 38% of the elbow torque variance. Those five parameters are:

• Increased wrist extension at stride foot contact
• Less downward displacement of the pitcher’s center of mass during stride
• Front knee flexion at stride foot contact
• Knee extension of the drive leg at stride foot contact
• Elbow pronation at stride foot contact

Interestingly, the relationships between these measurements and velocity in the study, on average, were weak. The authors explicitly reported on this part of the analysis. This is interesting because, at first impression, the role of velocity and performance in the increase of elbow load was something I assumed would account for most of these relationships. In other words, often we find that indicators of performance are decent indicators of torque since it takes more torque to accelerate and move faster. However, the authors reported on these relationships, so based on their analysis, that doesn’t seem to be the driver of this load.

One of the problems with these types of analyses, which is something we have to be careful about at Driveline as well with our biomechanical data, is that they are descriptive and cross-sectional. In other words, they only describe the dataset at one point in time. This study’s results are very interesting, and I think studies like these warrant more controlled experiments in which the subjects try to manipulate their kinematics with experimenters observing the effects on load measures. Alternatively, a follow-up could include a longitudinal dataset as well: measuring pitcher’s kinematics at two or more points in time and observing the association between the changes in these measurements and elbow load. 

The last thing to note is that elbow load is not entirely understood, and we still cannot predict injury with kinetic measures like peak elbow varus torque. Right now, it is merely our best proxy for the workload an athlete is undergoing. Either way, the results of this study are thought-provoking and something to consider for biomechanists to investigate further. We are not in a position to start making sound applications of these results in the applied settings, however. More testing is required.