Baseball Research Review #1 – Is Shoulder ROM a Risk Factor for Pitching Injury?
Authors: Wilk, K. E., Macrina, L. C., Fleisig, G. S., Aune, K. T., Porterfield, R. A., Harker, P., Evans, T.J., & Andrews, J. R.
The baseball player’s shoulder bears one of the greatest burdens in sport, undergoing numerous rounds of movement with significant speed, range and force. Glenohumeral joint internal rotation has been shown to be one of the fastest motions in the entire sporting world, reaching speeds of up to 7,250 degrees per second (Fleiseg et al., 2011). Couple that with a tremendous workload over a long and drawn-out competitive season, and it should come as no surprise that the shoulder has been the center of research and investigation for decades.
In general, a pitcher’s throwing shoulder exhibits greater external rotation and and less internal rotation than their non-throwing side. These side-to-side range of motion differences are likely due to combination of bony (e.g., humeral retroversion) and soft tissue (e.g., capsular thickening) adaptations, which occur in response to the substantial loads incurred during the throwing motion. Several theories have been proposed suggesting that shoulder range of motion deficits may increase the risk of pitching injuries, and previous studies have implicated deficits in glenohumeral internal rotation (Shanley et al., 2011), total rotation (Garrison et al., 2012), and flexion range of motion (Wilk et al., 2014) in shoulder and elbow injuries in pitchers, but the evidence thus far is far from conclusive.
What Was the Purpose of This Study?
The purpose of this study was to determine whether glenohumeral internal rotation deficit (GIRD), bilateral difference in flexion, bilateral difference in total range of motion (TRM), or bilateral difference in external rotation of the glenohumeral joint was associated with increased incidence of shoulder injury in professional baseball pitchers.
This was a prospective study. Prospective studies usually involve taking a large group of participants with shared characteristics (called a “cohort”) and watching them over a long period of time to see if they experience a certain outcome (say, the development of disease or injury) during the study period and relates this to other factors such as suspected risk or protection factors. In this study, the outcome measure was the development of pain or injury that was enough to land a pitcher on the disabled list and the suspected risk/protection factors were various measures of glenohumeral passive range of motion (defined below).
From 2005-2012, pitchers from one Major League Baseball organization were assessed during spring training for glenohumeral internal rotation (IR), external rotation (ER), and flexion passive range of motion (PROM)*. 296 pitchers were measured in total, of which “46 pitchers had their PROM measured in 3 or more consecutive seasons, 80 pitchers had PROM measured in 2 consecutive seasons, and 170 pitchers had PROM measured only once during the 8 years.” These pitchers were then followed until the end of the 2012 Major League Baseball season, to see if they spent any time on the disabled list over that time period. Average follow-up time was 48.4 months, and ranged between 5.5 and 89.5 months.
Definition of Suspected Risk Factors:
- Glenohumeral internal rotation deficit (GIRD). GIRD was defined as a difference of 20° internal rotation between the throwing and non-throwing shoulder, as originally described by Burkhart et al.
- Total rotation deficit. Total rotation deficit was defined as a difference in the total rotational motion of the shoulders where the total rotation of the non-throwing shoulder was at least 5° greater than that of the throwing shoulder.
- External rotation insufficiency. External rotation insufficiency was noted when a pitcher was found to have external rotation in 90° of abduction in the scapular plane of the throwing shoulder that was not at least 5° greater than that of the non-throwing shoulder.
- Flexion deficit. Flexion deficit was defined as a difference in the forward shoulder flexion between both shoulders where the flexion of the non-throwing shoulder was at least 5 greater than that of the throwing shoulder.
*Note: All PROM measurements were taken with a standardized protocol by the same two examiners over the entire course of the 8-year study. The technique used in this study to measure IR ROM has been shown to have good intra- and inter-tester reliability (Wilk et al., 2009). This is absolutely critical, as one of the main purposes of scientific literature with clinical findings is for independent reproduction of results. Full disclosure of valid and reliable methods is mandatory, in so much that studies without them are essentially useless for anyone looking for practical takeaways.
The main outcome measure was the incidence of shoulder injury. Shoulder injury incidence was defined as: “a player’s placement on the disabled list (DL) for any throwing shoulder injury.” Information regarding the time spent on the disabled list for each injury was gathered from trainer’s records and online databases (such as www.mlb.com and www.baseballprospectus.com) at the end of the 2012 Major League baseball season. As such, injuries that were not reported would not have been included in the author’s analysis.
Potential Confounding Variables:
All pitchers were pain-free and asymptomatic at the beginning of the study. This is important because this was a prospective study. In prospective studies it is important that none of the participants have developed any of the outcomes of interest before their baseline data has been collected. In this study, pain and/or injury was the main outcome measure, so it was important that all participants were experiencing neither pain nor disability at the beginning of the study.
Injury history. Of the 296 pitchers in the study, 18 had been placed on the DL for a shoulder injury and 14 had been placed on the DL for an elbow injury before their first PROM examination. Given that previous injury is a major risk factor for new injury (Fulton et al., 2014), it is possible that pitchers with an injury history could have confounded the results of this study by inflating the number of injuries reported. To account for this, the authors looked for a relationship between injuries that occurred prior to the study period and injuries that occurred during the study period in pitchers with an injury history, and found: “no significant associations between pitchers with a history of shoulder (P = .53) or elbow injury (P = .72) and the development of shoulder injury or shoulder surgery (P = .61 and >.99, respectively).”
- Over the course of the study, 51 pitchers (17%) spent time on the disabled list for shoulder injuries, of which 20 required surgery, mostly for various labral and rotator cuff injuries.
- GIRD, total rotation deficit, and flexion deficit were not significant predictors of either shoulder injury or surgery.
- On the other hand, there was a significant relationship between external rotation insufficiency and both shoulder injury and surgery: “Altogether, pitchers with insufficient external rotation were 2.2 times more likely be placed on the DL for a shoulder injury (95% CI, 1.2-4.1) and 4.0 times more likely to undergo shoulder surgery (95% CI, 1.5-12.6).”
This was an interesting study that adds to a growing body of literature which suggests that shoulder ROM may be an important risk factor for pitching injuries. In this particular study, pitchers who did not have at least 5° extra ER ROM in their throwing shoulder compared to their non-throwing shoulder were at a significantly increased risk of developing a shoulder injury during the 8-year study period. In addition, a 20° deficit in IR ROM on the throwing side, as well as bilateral differences in total rotational or flexion ROM were not associated with an increased risk of shoulder injury. These findings are intriguing, but they need to be interpreted with caution. To demonstrate that point, we want to outline some limitations when it comes to applying these findings in a training setting:
First of all, let us be abundantly clear: Our comments below are not directed towards the authors of this study. Our intention is not to denigrate their efforts, nor are we in any way suggesting this is a poor study. The reality is: Research is hard. And it is iterative. This study is simply one more piece to the puzzle; another cog in the wheel. With that said, there are a few limitations with this study that are worth pointing out. The authors have done a good job acknowledging many of these limitations in the Discussion section of the paper — we’ve just elaborated on them here:
- No data was provided on the time between the initial passive ROM assessment and when the injury actually occurred. A pitcher’s shoulder ROM is dynamic, which means it could easily have changed between the time ROM measurements were taken and when the injury occurred. In fact, 170 of the 296 pitchers — that’s 57%! — were measured only once, and there was no data provided on the time between the PROM test and when the injury occurred. Thus, we cannot assume that measurements taken at the beginning of the season would be static throughout the course of a month, let alone a few seasons. We know that long-term adaptations to throwing include a reduction in internal rotation (which includes but is not limited to humeral retrotorsion) as well as an increase in external rotation. External rotation doesn’t seem to be affected as much by an acute bout of throwing as internal rotation, which shows a significant reduction after throwing (Kibler, Sciascia, & Moore, 2012; Reinold et al., 2013). However, there is evidence that external rotation increases significantly over the course of the baseball season (Dwelley et al., 2009). Considering that shoulder ROM was measured on the first day of spring training, it’s possible that pitchers without internal rotation or total ROM deficits in spring training developed deficits over the course of the season. Likewise, it’s very possible that pitchers with external rotation insufficiency in spring training could have developed “sufficient” external rotation prior to injury. However, one interesting theory is that the less external rotation ROM a pitcher has going into spring training, the more he must (theoretically) acquire during the season. Considering recent evidence that large acute increases in training load increase injury risk (Hulin et al., 2014), and given that injury rates are highest in the first month of the season and drop substantially after that point (Posner et al., 2011), external rotation insufficiency may indicate an athlete who is physically under-prepared for the substantial increase in throwing load that occurs in conjunction with the start of the baseball season. It is a bit of a stretch, but this study may indirectly lend support for earlier and more gradual return-to-throwing programs.
- What about the findings from previous ROM studies? In 2011, Wilk et al.concluded that total range of motion deficits of the throwing arm were contributing to shoulder injuries in pitchers, but now that has changed to ER deficiency? Have pitchers evolved? Has the mechanism somehow changed in just a few years? Is the previous study all of a sudden meaningless, and if so, what does that say about this current study 5 years from now? In our opinion, this just lends more evidence to the idea that injuries are very multifactorial and that losses in ROM lead to a decrease in afference, regardless of where they come from AND lead to diminished system rigidity. (We go into much more detail on this in our webinar.)
- Asymptomatic does not equal healthy. The literature is full of examples of pitchers with “junky” shoulders and no symptoms (e.g., Miniaci et al., 2002) and vice versa. Stating that a pitcher was “healthy” from a tissue pathology standpoint at the time of screening cannot be assumed without advanced imaging; ultrasound / MRI. This is an important distinction: When pitchers are put on the disabled list, we cannot infer that tissue damage occurred prospectively, especially without baseline imaging!
- What about humeral torsion? Emerging evidence suggests that humeral retrotorsion may be protective against throwing injuries (e.g., Polster et al., 2013) and Whiteley & Oceguara (2016) recently suggests that ROM without knowledge of humeral torsion is limited in its clinical interpretation. That being said, considering that humeral retrotorsion and external rotation ROM are related (Chant et al., 2007; Reagan et al., 2002), passive ER ROM may be a good proxy for humeral retrotorsion in situations where it cannot be directly determined via ultrasound, and thus may provide clinicians with important information about how they should manage individual athletes. (However, more research is needed to determine the validity of passive ER ROM as a measure of humeral retrotorsion.)
- Not all passive ROM is equal. Neurological tension and mechanical tension within passive ROM are different. This is a topic for another discussion, one that is critical to understand and interpret, and is in the works (in which a video demonstration will be included).
- Can we really reliably detect a 5° bilateral difference in ER ROM? Let’s assume that 5° was indeed the magic injury-identifying number. How many coaches and therapists can actually measure this reliably, even with a goniometer and an assistant? This is a difficult skill that requires technical proficiency and lots of experience. The examiners in this study had a combined 30 years of experience measuring professional baseball shoulders!
- No data was provided on injury severity. Because injury records were obtained from online databases, the only measure of injury severity was time spent on the disabled list, which, as you can imagine, is probably not the most accurate, as it can be influenced by other factors. That being said, the authors were looking at whether or not injury occurred (a binary, yes/no distinction) so this is not necessarily a huge limitation. The authors bring up another good point about injury data: “We are not sure whether all teams use the same definition of injury and whether all days on the DL are truly injury days.”
As you can see from our discussion above, there are a lot of things to take into consideration when interpreting these findings. (And this is by no mean an exhaustive list!) So what can we take-away from this study?
What this study actually tells us:
- This study suggests that a pitcher with external rotation insufficiency may have an increased risk of developing shoulder injury, while a pitcher with GIRD or total rotation deficits may not.
What this study does not tell us:
- This study does not tell us that external rotation insufficiency causes shoulder injuries. You may have heard that, “correlation does not imply causation”. It is applicable here, and it is likely something we will bring up again and again in our research reviews.
- This study also does not tell us that improving ER ROM in pitchers with “external rotation insufficiency” would reduce injury risk and, if it does, how best to acquire that new ROM. (A topic perhaps for another time!)
- Finally, this study does not tell us that we should ignore other measures of shoulder ROM such as internal rotation ROM, total ROM, or flexion ROM.
Glenohumeral rotational ROM is probably important for baseball pitchers, but it’s not clear why it is, which (if any) measure of shoulder ROM is most critical, or how best to acquire it.
What happened when this study first came out was that there was a (rather unfortunate) strong knee-jerk reaction towards dismissing GIRD, and it has even been suggested that training or attempting to acquire IR in the shoulder is either unnecessary (or even detrimental!). We believe this is misguided for many reasons.
From a neurobiological perspective, we believe that acquiring and training new range provides the CNS with more afference and improves resiliency of the entire system. (Check our webinar for more information on this topic.) This means that our goal in training should be to acquire and control as much range of motion as is possible within the adaptive and recovery capabilities of the individual athlete.
The reality is: We don’t have the data to definitively say that there is an “optimal” amount of shoulder rotation ROM, and we may never. But, with the extreme ranges of motion at which the shoulder must accept and disperse force during the pitching motion, why would we not want to improve control in such ranges? Neglecting the terminal end of our force-length curve isn’t in our best interest. The mere fact that we go into full internal rotation when throwing at game speed means that we must train our tissues and our nervous system for it!
In our opinion, it is illogical to have read this article and concluded that we only needed to correct external rotation deficiency and dismiss GIRD. Although this study is a good start towards investigating the correlation between PROM and future injury, there is simply no way of rationalizing a causation at this point in time.
Well, there you have it! If you like what you’ve read, and you’d like to receive updates about future instalments of the Baseball Performance Research Review, be sure to sign up for our newsletter using the form below.