The Sport Journal
February 7, 2020
Concussion in the Collegiate Equestrian Athlete
Equestrian sports, in general, pose a significant risk of concussion. Minimizing the risk of concussion has been a focal point in recent years. The purpose of this paper is to describe concussion and explore potential association(s) between groups of musculoskeletal injuries and Body Mass Index (BMI) on the risk and odds of concussion in the collegiate equestrian athlete. Forty-three schools, ranging from DI to DIII, from the Eastern United States were selected from the NCAA and Intercollegiate Horse Show Association’s websites. Self-reported injury and demographic data was collected through an online survey created in Mach Forms. Seventy-three participants completed the online survey (women n=71, men=2). Aggregate descriptive data is reported on all subjects. After removing data on 2 men, and a single female with incomplete data, the data from 70 females with complete data was analyzed using chi-squared and Fisher’s exact tests and ordinal logistic regression. Pearson’s chi-squared as well as Fisher’s exact test (p-value =.0288 and.0297 respectively) indicates the risk of having concussion with 0 UE injury is not the same as with 1 or 2+ injuries. The average number of injuries per athlete increased from 0 to 2(+) concussions. Concussion is a commonly reported injury. Upper extremity injury is identified as having the strongest association with concussion risk in the equestrian athlete. Knowing UE injury status could be useful in gaging the risk and odds of concussion in equestrian athletes.
Female participants dominate the approximately 10,000 athletes competing in U.S. collegiate equestrian sports (6,23). While not as well researched as other female collegiate sports (e.g. soccer), equestrian sports, in general, are otherwise widely recognized by researchers as dangerous (16). In a 2002 data analysis, it was noted “horse-riders can expect a serious accident once in every 350 hours of participation, which is twenty times more dangerous than motor cycling” (Silver 264). The head was the most commonly injured body part amongst riders 19-49, with upper extremity fractures just ahead of concussion (16.6 vs. 15.2%) in riders age 0-18 (2). Data compiled from the National Electronic Injury Surveillance System (NEISS) for the years 1997-2015 showed an average of 24.19% of all equestrian injuries came from the sport setting, with injuries to the head making up an average of 20.07% of all injuries; Concussion made up an average 5.01% of all injuries(19). Winkler (2016) noted 45.2% of the over all population and 21% of those aged 18–29 years old sports-related traumatic brain injury (TBI) were from equestrian and related sports. In an Australian population, equestrian activities accounted for the second highest level of mean participation-adjusted rates of hospitalization for concussion over a 9-year period (130.3/100 000)(8). Falls have been shown to be responsible for 51%- 82% of the injuries that occur in equestrian sports (2,4,5,14,20). Indicative fall times,(i.e. time it takes to contact the ground) based on rider center of mass height, from four to ten feet have been calculated and range from 505-782ms (20).
Non-collegiate sources were utilized to estimate the rate of falling from the horse as data regarding the rate of falling in collegiate competition is not available. The rate of falling, as reported in the 2010 United States Eventing Association Cross Country Competition Safety Report, for each starter is 1 in 42 and the probability of a fall for each jump is 1 in 879 (24). Despite the limited data, the mechanism creates a fall is consistent across disciplines and skill levels. There is a loss of Center of Mass (CoM) control and position as it relates to movement of the horse. This can be intentional (e.g. bucking, rapid stop) or unintentional (e.g. rider error) and most often associated with an obstacle (e.g. fence) (10, 24).
The high incidence of injury has led to increased attention on reducing the risks associated with falling as a way to decrease the number of injuries, especially concussions. Continuing efforts include improving personal protective equipment for the athlete, designing safer competition obstacles and teaching an emergency dismount to help athletes escape from the horse under certain situations. Prior to 1970, gymnastic style tuck and roll training was also taught as part of the Pony Club curriculum and as such, widely accepted as a way to minimize the risk of injury should the rider fall from the horse (16). This type of training has been revived on an extremely limited basis and is presently not part of the equestrian’s standard training or curriculum.
Research indicates that concussion, musculoskeletal injury and body mass index (BMI) can influence injury and re-injury risk in subjects, secondary to increased demands on postural control (3,7,9,11-13,15,17). While equestrians accept the high risk of falling as well as the association between falling from the horse and concussion, the risk and odds of concussion that can be associated with musculoskeletal injury and BMI is unknown in female collegiate equestrians.
The purpose of this paper is to present a description of concussion in collegiate equestrian athletes and to explore further potential association(s) between lower extremity (LE), upper extremity (UE) and spinal (SP) injuries, and body mass index (BMI), on the risk and odds of concussion in the female collegiate equestrian athlete.
This study was approved by the institution’s IRB. An online survey instrument was developed in Mach Forms to gather demographic and self-reported, whole body, injury data from the respondents. Concussion history was self-reported as a separate category. Consent was assumed upon voluntary completion and submission of the survey. Anonymity was assured to all participants. The form reduced mailing costs and encouraged participation in an uncomplicated manner. An outside expert in the field of equestrian reviewed the survey instrument for appropriateness of the content, content validity and also provided feedback relative to the format of the questionnaire. A link to the survey was sent to 43 equestrian teams, representing 10 percent of the total number of teams, randomly selected from the National Collegiate Athletic Association and Intercollegiate Horse Show Association’s websites. Seventy-three athletes (male=2, female=71) completed and submitted the survey. We report descriptive data on the entire group of 73 athletes. The data from the two male subjects was removed to avoid potential gender bias and one female with incomplete data was also removed, resulting in 70 total subjects for further analysis (mean age of age = 20.3 years +1.90yrs, mean weight=62.15kg +9.42kg, mean height+8.2cm).
For the purpose of analysis, concussion was organized into 0, 1 and 2 or more (2+), categories. Independent risk factors were total number of SP, LE and UE injuries. The data was analyzed using Chi-squared and Fishers Exact tests.
Aggregate total number of concussion are reported in Table 1. The 32 individuals reporting concussion represent 44.5% of the population. There were 61 total concussions (ave. appx 2 per individual reporting concussion). Concussion represents 11% of the 568 total number of injuries (which includes concussion) reported. The no concussion group accounted for 39% (221/568)of injuries; while the concussion group accounted for 61% (347/568) of injuries.
Table 1. Aggregate Number of Concussions
1 2 3 4 0
No. of Athletes 17(22.9%) 5(2.7%) 10(13.5%) 1(1.3%) 40(44.5)
We then homogenized the group into four concussion categories x body part (head, neck, R. shoulder, etc). The zero concussion group contained 39 athletes, totaled 155 affected body parts and appx. 4 inj/athlete); Single concussion (N=15, Total 62 body parts affected, appx 4inj/athlete); Two concussions (N=5, Total 27 body parts affected, appx. 5 inj/athlete) and more than two concussions (N=11, Total 96 body parts affected, appx. 9 inj/athlete).
For BMI, values ranged from16.31 to 30.18. Mean BMI was 22.57. According to standard BMI interpretation, 1subject was graded as obese, 11 overweight, 4 underweight and 55 normal weight. The discussion regarding BMI is limited as after further analysis demonstrating the association between BMI and concussion was not significant for our data.
The three tables below report the distribution of athletes categorized by concussions and number of injuries.
Table 2. Athlete Distribution by Concussions and UE Injuries Categorized into 0, 1 and 2+
No Conc 1 Conc 2+Conc
0 25 8 6
1 8 1 1
2+ 6 6 9
Table 3. Subject Distribution by Concussions and LE injuries
No Conc 1 Conc 2+Conc
0 11 6 2
1 4 0 0
2+ 24 9 14
Table 4. Subject Distribution by Concussions and Spinal Injuries
No Conc 1 Conc 2+Conc
0 19 8 6
1 5 0 1
2+ 15 7 9
The chi-squared test indicates the risk of having a concussion with 0 UE injuries is not the same as with 1 or 2+ injuries (p-value=.0288). Since, some of the cells are sparse (less than 5 subjects) we also used Fisher’s exact test to study association between concussions and UE, LE and Spinal injuries. Fisher’s exact test also showed a significant association (Fishers p=.0297) between Concussions and UE but not the other two (LE, Fisher’s exact p=0.205 and Spinal, Fisher’s exact p=0.568). Thus, an association between incidence of concussions and UE injuries was indicated by chi-squared test as well as the Fisher’s exact test. However, BMI, Spinal injury and LE injury did not show any association with concussion occurrence.
Our finding of 11% of all injuries being concussion is more than double than the cumulative NEISS average of 5% from 1997-2014 and within the range (9.7-15%) of all injuries reported by Zuckerman (2015). While no data to compare collegiate concussion rates is available, in a survey of concussion and concussion symptoms in equestrian athletes registered with the United States Equestrian Federation, Buetow (2016) found 71.6% of the population reported concussion symptoms, with 40.9% being officially diagnosed with a concussion. While we did not ask about concussion symptoms, the 45% of our population reporting at least one concussion is comparable. Only motor vehicle-based sports are consistently reported as having higher TBI rates (8, 21, 26).
Perhaps more importantly, in a study of Swedish equestrians, the head was reported as the most frequently injured region and the most commonly injured region associated with a fatality (18).
Helmet use is chronically emphasized as a way to reduce concussion in equestrians. While helmet use is mandatory during competition (25), and perhaps the most easily modifiable factor in terms of reducing head injury/concussion, it’s overall use is chronically low (27). So while important to continue to emphasize overall helmet use among equestrians and improve helmet function, it is perhaps necessary to expand industry focus to include complementary modes of protection.
The average difference in number of body parts injured between zero and one concussion is negligible (3.97 vs. 3.87) and in fact, one concussion having a lower average is interesting. While we cannot suggest a scientific reason for the lower value, it is possible that athletes are taking extra precautions to minimize the risk of a second concussion. The 20% increase from 0/1 to 2 concussion (4 vs 5) and then almost doubling (5 vs. 9) in injuries per body part when more than 2 concussions occurs is a revealing, but not surprising finding. The interaction of concussion and the variables that influence postural control is not fully understood; yet it is clear postural control (Howell 2018) and or neurocognitive performance (Herman 2015) is negatively affected, hence influencing injury risk. In riding athletes, this would be an especially critical problem, as they would be falling from height and often at significantly higher speeds than other athlete’s experience.
We would however suggest that, as it relates to equestrians, the pathway proposed by Herman (2015) would be circular vs. linear. The hypothesis being, as the effects of concussion compound, the athlete’s ability and capacity to minimize the risk of falling and hence the risk of injury decrease. Thus more injury leads to greater risk of falling and more injury.
While no association between LE and concussion was found in our data, a short discussion of the potential interplay of these variables as it relates to equestrians is warranted as this is a noted finding in other sports (Herman 2017). Equestrians accept falling off as an inevitable part of riding. However, they rely on the strength of “their seat”, which is commonly understood as the interaction between the LE, saddle, stirrup and movement of the horse, to minimize the risk of falling off. Equestrian athletes are not traditional foot on the ground athletes; unlike as an example, soccer players. The lack of stable ground under foot provides fewer options to the athlete to maintain body position and postural control compared to more traditional athletes. Hypothetically then, a LE that is compromised by (multiple) injuries could increase the risk of falling and hence head injury and or concussion.
The use of the UE to protect the head is instinctual (e.g. attempting to catch/deflect a projectile thrown at the head). However the concepts of when and how to use the UE to protect the head first, when a fall is inevitable, is not taught as a primary part of riding education. Traditional riding education teaches athletes to stay on the horse if at all possible. Riders are taught the UE’s primary use is to maintain a connection with the horses head through the reins. Athletes learn by feel and experience when their “seat” will no longer keep them in the saddle and a fall is inevitable. In this situation, they are instructed to grab the pommel of the saddle, neck or mane of the horse or tighten up on the reins to stay on the horse. By attempting to ride out the dangerous situation instead of dismounting when the situation is less dangerous, the athletes can place themselves in a position where an injury could be more likely to happen, as opposed to dismounting when their position could allow for less risk of injury.
For example, the fall time for an athlete falling from a horse is typically measured in milliseconds (20). In this extremely small amount of time, he or she has to decide they are in danger, plot an escape and generate a movement pattern. This is not likely enough time to place the UE in a position to protect effectively the head. The overall poorer body position and lack of time to properly position the UE, likely compromises the UE’s ability and capacity to spread out the increased ground contact forces, allowing for increased chance of head contact with the ground (think darting into the ground vs. tucking and rolling through the fall). When the above scenario is considered, the model results of UE injuries having the strongest association with concussion appears plausible. Anecdotally, a 1997-2013 trend in the NEISS data demonstrates an overall increase in concussion and flat trend of overall UE injuries, which could indicate anyone of the above scenarios (19).
The thought to ride out the dangerous situation may be due to how falling is treated in equestrian. While equestrian sports do air on the side of caution, generically, if an athlete is separated from the horse or touches the ground, and requires assistance to get back into the saddle, he or she is can be penalized either time or points. In the situation where a concussion could be suspected, the athlete is removed from competition until cleared by a physician (25).
We fitted an ordinal logistic regression model to explore the association of MS injury while adjusting for BMI. Although the odds of 2+ vs 0 or 1 concussions were almost three-times greater going from 0 to 5+ UE injuries (OR=1.250 and 3.051 respectively), while adjusting for BMI but despite large effect size the results were not statistically significant. While a larger sample size is needed for further exploration and to reach stronger conclusions, the results are worth a short discussion. BMI presents an interesting challenge for equestrian athletes. Horses are taught to move at a given pace during competition irrespective of the load it is carrying. Physics dictate athletes having larger BMI’s moving at like pace, compared to athlete’s with lower BMI’s, are subject to not only larger destabilizing forces that challenge body position, but larger ground contact forces. During the fall, the forces generated by the horse and imparted on the rider can often place the rider in a head down position or turn the rider into a head-first projectile. As such, the athlete must now manage his or her body weight with the UE, which does not have the same physical capacity as the LE to dissipate the ground contact forces. When the above scenario is considered, the model results, albeit inconclusive, appears reasonable.
Concussion is a common injury in equestrian sports. This investigation is the first to report potential associations between concussion and categories of musculoskeletal injury in collegiate equestrian athletes. Our findings of UE injuries having stronger associations, than LE and SP injury, as well as an increasing average number of injuries per athlete going from 0 to 2(+) concussions is unique. While focusing on improving obstacles and personal protective equipment to decrease injury risk should continue, our findings provide a starting point from which to consider athlete specific factors (i.e. musculoskeletal injury, BMI) that influence the risk and odds of concussion in the collegiate equestrian athlete.
Applications in Sport
Despite the large number of participants and teams in collegiate equestrian sports, the frequency of concussion and potential interaction(s) with musculoskeletal injury on the athletes who participate in them are not well understood by the people charged with their care. Concussion, and their co-factors, pose as significant a problem in this group of athletes, as in any other group of athletes. Our findings indicate concussion is common, average number of injuries per athlete increased with increased concussions and upper extremity injuries potentially increase risk and odds of concussion. These findings provide a rationale to explore potential injury prevention programs beyond the present focus of improving personal protective equipment and obstacle design.