The National Interscholastic Cycling Association Mountain Biking Injury Surveillance System: 40,000 Student-Athlete-Years of Data

Abstract

A web-based injury surveillance system was implemented through a collaboration between University of Utah researchers and the National Interscholastic Cycling Association (NICA) to better understand injury characteristics in mountain biking. Data were collected from NICA leagues during the 2018 and 2019 seasons. Injuries were tracked in 41,327 student-athlete-years, identifying 1750 unique injuries during 1155 injury events. Rider-dependent and rider-independent variables were analyzed. The most commonly reported injuries were concussion (23.6%), injuries to the wrist/hand (22.3%), and shoulder (15.6%). Half of all injury events occurred on downhills. Men and women reported similar yet significantly different injury rates (2.69% and 3.21%, respectively; P = 0.009). Women sustained more lower-limb injuries (37.8% vs 28.3%; P = 0.003). Nearly 50% of crashes resulted in an emergency room visit. Youth mountain bike racing is a rapidly growing sport. Acute traumatic injuries are common. Injury surveillance system data are now being used to inform injury prevention strategies and direct future research.

Introduction

The sport of mountain biking has grown significantly since its modest beginnings in the hills of Northern California in the 1970s (¹). The popularity of mountain biking has facilitated the development of multiple disciplines within the sport (cross-country, enduro, and downhill) and has enabled wide-spread participation across the demographic spectrum. The growth of racing, since the first adult competition in the 1980s, has followed closely on the heels of the sport's burgeoning popularity among recreational enthusiasts. Youth competition, initially limited by accessibility and therefore participation, lagged behind that of adult competition. Now, over the past decade, youth cross-country mountain bike racing has become one of the fastest growing sports in interscholastic athletic competition.

The National Interscholastic Cycling Association (NICA) was formed in 2009 as the oversight body for middle and high school mountain bike racing in the United States (²). As the governing body for interscholastic youth mountain bike competition, NICA is responsible for the administration of all competitive and noncompetitive cross-country mountain bike activities, including implementation of rules, league certification, athlete registration, coach registration, development of training camps and races. In 10 years, NICA has grown from one small league in Northern California to more than 30 leagues across the country. By 2019, there were more than 20,000 student-athletes participating in NICA sanctioned mountain biking in the United States. Although not explicitly reported in the literature, the COVID-19 pandemic may have added to the popularity of mountain biking, as evidenced by nationwide bike shortages (³).

Mountain biking is not without risk (^4,5). Athletes use specially designed bicycles with shock absorbers and wide, knobby tires to negotiate trails in mountain environments that are often narrow and have natural obstacles, such as rocks and tree roots. The recent rise in youth mountain bike competition has been accompanied by a concordant rise in mountain biking related injuries. Little is known about the unique injury risk profile encountered in cross-country mountain biking in the adolescent age group. Aiming to protect participants, NICA collaborated with sports epidemiology researchers to design, test, and implement a prospective, longitudinal injury surveillance system (ISS) to better characterize injuries seen in youth mountain bike racing. The findings from this ISS would ultimately lead to development of injury prevention strategies and interventions. This report briefly describes the NICA ISS, with a primary emphasis on the results from the first 2 years of data collection during competition years 2018 and 2019, collectively comprising more than 40,000 student-athlete-years. To the authors' knowledge, this is the largest mountain biking ISS in existence. The use of natural language processing to expedite data extraction from open text fields within the ISS tool will be highlighted in methodology. Finally, this report will discuss current and future interventions aimed at reducing injury in youth mountain bike competition.

Methods

Data Collection

After the designing and testing phases, the NICA ISS was implemented in January 2018. Details of the NICA ISS are found elsewhere (⁶). Briefly, the NICA ISS is an electronic, web-based sports ISS. It uses REDCap (https://www.project-redcap.org) to collect injury data, along with exposure data (⁷). Because the vast majority of injuries are acute and traumatic in nature, with crashes often resulting in more than one injured body part, we specifically incorporated the term “injury event.” An “injury event” is defined as: “any physical event occurring to a single rider during a NICA-sanctioned practice, race, or other training session that results in physical harm to the participant significant enough to: 1) warrant referral to a medical provider, or 2) lose time from training or competition beyond the day of injury, or 3) miss school or work” (⁶). A single-injury event often results in more than one specific injury within a rider. For example, a rider may crash and have three distinct injuries (a clavicle fracture, a concussion, and a sprained ankle) but this is only one-injury event. Further, a multirider crash could result in multiple-injury events (one-injury event per injured rider).

Injuries were reported via REDCap by a designated reporter on each NICA team, who was an official with each team (often a head coach). A separate injury report form was completed for each injury event, and more than one specific injury was allowed to be reported for each injury event. Within each injury report, open-ended text fields captured verbal descriptions of injury events entered by the designated reporter. The following injury-related information was collected: rider demographics, competition division, trail characteristics at crash location, weather, other factors felt to contribute to the incident (e.g., technical nature of trail, rider inexperience — as reported by the coach), body part injured, diagnosis category, and time loss due to injury. Injury mechanisms were included in the survey. Time loss from the injury was calculated as the date the student-athlete was able to return to practice or competition from the date of injury (e.g., if a student-athlete returned to practice a week after his or her injury, the time loss was 7 d). This was reported by designated reporters. These variables, including those for injuries, were pre-determined and listed in the injury report form (Appendix A, https://links.lww.com/CSMR/A115). Furthermore, the designated reporter identified if the student-athlete had to go to the emergency room or not for their injuries.

Separately, rider exposure also was collected from designated reporters on a weekly basis as part of the NICA ISS, with reports taking less than a minute to fully complete. Exposure was recorded as the number of riders participating in each individual practice ride or race over each week. When each designated reporter received a weekly reminder to complete exposure reporting, they also received a reminder to report any injuries.

Data Analysis

The current study focused on summarizing injury data using descriptive statistics. Specifically, mean and standard deviation were used for continuous variables, while categorical variables were analyzed with frequency and proportion. An injury event proportion, analogous to epidemiologic incidence proportion and (injury) incidence proportion in literature (^8,9), was calculated as a total number of injury events divided by a total number of student-athletes registered for NICA. As subanalysis, we examined the following aspects of the data: common injury types and locations, medical evacuation from the crash site, time loss due to injury, injuries in practice versus race, trail conditions, and any other common characteristics associated with injuries. Lastly, proportions of total injuries and specific injury types were compared by gender and riding class, using Pearson χ² test. The statistical analysis was done using Stata/MP 16.1 for Windows (StataCorp LLC, College Station, TX). As the research team set a cutoff of 80% exposure reporting for inclusion in analysis, exposure reporting did not meet that threshold and therefore, was not included in the calculations.

Description of Analysis of Text Data

All text data from every entry were reviewed by NICA. The investigators additionally reviewed the text of a subset of entries (100) to assure the text entry matched the survey fields; they did in all cases. Analysis of the text data was completed using R (Version 3.5.1) (¹⁰) and its associated packages. After importing the data, the text field containing the injury mechanism was normalized, meaning that contractions were expanded, special characters and numbers were removed, and capitalization was standardized using DataCamp (https://www.datacamp.com/community/tutorials/R-nlp-machine-learning). Next, the tidytext package (¹¹) was used to tokenize the text into three-word phrases with noncoding stop words (such as “the” and “and”) removed. To visualize the most common phrases, the resulting phrases and their occurrence frequency were used to generate a word cloud using the word cloud package (Fellows I, 2018). The data also were used to generate a Sankey diagram of word flow using the networkD3 package (Allaire JJ, et al., 2017). The most common phrases were investigated further as a subanalysis by sex (men vs women), using the same process.

Results

In the first 2 years of data collection (2018 and 2019), the NICA ISS recorded 1155 injury events in 41,327 student-athlete-years (32,926 or 79.7% men and 8401 or 20.3% women), resulting in an injury event proportion of 2.8% (see Table for the distributions of divisions). There was an average of 1.52 unique injuries per injury event (1750 unique injuries in 1155 injury events). Figures 1 and 2 summarize injuries by body part and type, respectively, in student-athletes. The most commonly injured body parts were head/brain (i.e., concussion or possible concussion, 273 or 23.6%), followed by wrist/hand (275 or 22.3%), and shoulder (180 or 15.6%). Contusions and abrasions accounted for 38.7% (547 injuries) of all nonconcussion injuries. Fractures and dislocations accounted for 26.2% (371 injuries) of all nonconcussion injuries.

Table - Distributions of divisions by student-athlete-years and by injury events.

Division	Student-Athlete-Years (n = 41,327)		Injury Events (n = 1155)
	Male (n = 32,926)	Female (n = 8401)	Male (n = 885)	Female (n = 270)
Middle school	12,671	3472	300	93
Freshman	6763	1458	200	56
Sophomore	3123	649	121	32
Junior varsity	9196	2309	215	65
Varsity	1173	512	43	22
Other/not reported	0	1	6	2

Figure 1: Student-athlete injuries by body part, n = 1155 injury events and 1750 injuries. The most commonly injured body parts were head/brain (i.e., concussion or possible concussion, 23.6%), wrist/hand (22.3%), and shoulder (15.6%).

Figure 2: Student-athlete injuries by type, excluding concussion, n = 1415 injuries. Contusions and abrasions accounted for 39% of all nonconcussion injuries in men and women collectively, while fractures and dislocations accounted for 26% of all nonconcussion injuries.

Injured student-athletes were unable to complete their practice ride or race 71.3% (823 athletes) of the time. Injured student-athletes were evacuated from the crash site by ambulance 5.9% (68 cases) of the time and by helicopter 0.7% (8 cases) of the time. While 48.8% (564 cases) of all injury events resulted in an emergency room visit, only 3.6% (42 cases) of injuries resulted in hospital admission. As seen in Figure 3, 36.1% (342 cases) of injury events resulted in time loss from riding of less than 1 wk, whereas 28.9% (274 cases) resulted in time loss of at least 4 wk. Injuries were season-ending for 22.0% (254 cases) of student-athletes.

Figure 3: Student-athlete time loss from injury, n = 947 injury events for those who lost time from riding. Thirty-six percent of injury events resulted in time loss of less than 1 wk while 29% of injury events resulted in time loss of greater than 4 wk.

Injury events occurred during a team practice on mountain bike trails 55.4% (640 injury events) of the time and during a race 29.9% (345 injury events) of the time, as well as during a team practice for skills training 5.1% (59 injury events), a team practice on paved roads 3.9% (45 injury events), NICA on-the-bike skills 0.2% (two injury events), and unreported 5.5% (64 injury events). There were 50.5% of injury events (583 cases) that occurred while riding downhill, 33.2% (384 cases) occurred on flat terrain, and 6.7% (77 cases) occurred while riding uphill. A total of 76% of injury events (878 cases) occurred on a trail with which the student-athlete was familiar. Other common factors reported to be associated with injury events included inexperience of the student-athlete (21.6% or 249 cases), technical nature of the trail (19.9% or 230 cases), and negotiating a turn (19.7% or 227 cases). There were 28.4% of injury events (328 cases) that were not associated with any particular cause.

Analysis of free text entries revealed “front tire washed”, “single track trail”, and “front wheel washed” were the most commonly occurring three-word phrases in the text descriptions used to describe the cause of injury events in student-athletes (Fig. 4). In male student-athletes, “single-track trail”, “front tire washed”, and “front wheel washed” were the three most commonly occurring phrases used to describe injury events. In female student-athletes, the three most common phrases to describe the injury event were “front tire washed”, “rider lost control”, and “athlete lost traction”. From this analysis, it appears that loss of control of the front wheel also may be associated with events leading to injury. This loss of control, commonly referred to as “washing out”, is the result of loss of traction of the front tire, causing the bike to slide sideways away from the rider when cornering.

Figure 4: Three-word phrases with at least three occurrences (all student athletes), font size correlating with frequency. Free text description of 80/1155 injury events (7%) included some variation of the word “wash”.

The injury event proportion was slightly but significantly higher in female riders than in male riders (3.21% [270 injury events] vs 2.69% [885 injury events], P = 0.009). Some injury patterns in student-athletes differed significantly by sex; 37.8% of injuries (102 cases) in female riders were to the lower limb compared with 28.3% (250 cases) in male riders (P = 0.003). Student-athletes entered in the sophomore and varsity divisions experienced a significantly higher rate of injury (4.06% and 3.86%, respectively) in comparison to the other divisions (middle school, freshman, junior varsity with rates of 2.43%, 3.11%, and 2.43%, respectively; P < 0.001).

Discussion

The successful implementation of the NICA ISS has been made possible by a number of factors, including the accelerating work of the sports epidemiology community over the past 25 years, technology that enables large-scale data collection from geographically dispersed locations, and a strong desire of many within the mountain biking and sports medicine communities to protect the health and safety of young riders. It is important to highlight that an efficient nationwide ISS for mountain biking has been implemented with early results that will inform future interventions to prevent injury in this young and promising group of athletes.

The quality of injury data is considered good with formal data validation anticipated to be completed after 3 years of data collection. The injury definition is aligned with similar systems (^12,13) and designed to exclude inconsequential injuries. All discrete entry fields on the injury report form are required for submission, so there are no missing data in these fields. The designated reporters provided quite informative accounts in the open-ended text entry fields. Analysis of the text entry fields using R has enabled identification of additional common injury characteristics and patterns, namely, injury resulting from the front wheel washing out.

The existing literature on mountain biking injuries is modest but growing. Most reports primarily concern injuries in adult recreational riders (^4,14,15). A smaller number of reports have focused on injuries in adult racers (^16,17). Some reports have focused on disciplines other than NICA-sanctioned cross-country mountain biking, such as downhill mountain biking or injuries from mountain bike terrain parks. Other reports have focused only on injuries that presented to emergency rooms (^15,18,19). Two review articles have focused on youth mountain biking injuries but not specifically in the discipline of cross-country nor during competition (^20,21). To our knowledge, the NICA ISS is the first systematic, prospective study of injuries in the competitive youth mountain biking population, and the largest mountain biking ISS in existence.

The overall injury event proportion was relatively low for student-athletes (2.8%). It is difficult to compare these proportions directly with those of other youth sports as reporting methods and the nature of the sports are different, but they appear to be lower than injury proportions observed in youth basketball, soccer, and American football (²²). Concussion/possible concussion was the leading diagnosis in this cohort of youth mountain bikers. Discussions are underway to study factors leading to concussion and interventions to decrease concussion rates. Not surprisingly, upper limb injuries were more common than lower limb injuries in student-athletes, which is aligned with reports of mountain biking injuries in other populations (^{17,19,20,23–25}). Interestingly, injury patterns differ by sex in student-athletes, with women sustaining more lower limb injuries than men. Nelson and McKenzie (¹⁹) found similar patterns in upper versus lower limb injuries between male and female pediatric recreational mountain bikers treated in the emergency room. Differences in injury characteristics may reflect differences in the manner in which men and women fall or a slight variability in position/station on the bike. Video capture of injury events could elucidate these differences better but presents practical challenges on mountain biking trails that cover significant distances. One option for video capture of injury events is to identify sections of frequently used trail where crashes are common, and place video cameras at those locations during practices or races.

It should be noted that designated reporter-entered comments appear to differ for men and women. Injury events in male riders appear to be attributed to the bicycle or terrain (single-track trail/front tire washed/front wheel washed), while injury events in female riders were more commonly attributed to the rider (front tire washed/rider lost control/athlete lost traction). Put simply, underlying cause of injury was placed on factors external to the rider in men and intrinsic to the rider in women. Whether this finding reflects unconscious gender bias of the designated reporter or differences in self-perception between male and female student-athletes, it is not possible to say due to the possible joint contribution from both parties in completing the injury report form. However, this trend is worth highlighting given the degree of inequity seen in many sports, including discrepancy in support of women's cycling teams and events (^26–28).

Many injuries were relatively minor, including abrasions and contusions, and with less than 1 wk of time loss. However, some injuries were more significant, including concussions/possible concussions, fractures and dislocations, and required four or more weeks of time loss. No catastrophic injuries such as death, spinal cord injury, or severe traumatic brain injury were reported. Efforts to prevent more severe injuries will be prioritized.

Several factors are associated with injury occurrence and will inform future injury reduction interventions. More injuries occurred while riding downhill sections relative to riding on flat or uphill sections of trail. While it seems clear that improving riders' downhill skills is important, close to half of all injuries occur on flat and uphill sections. Therefore, riders need to remain aware that injuries can occur on any trail incline. Injury proportions differed by competition division, with riders in the Varsity and Sophomore divisions sustaining more injuries than riders in the junior high school, Freshman, and Junior Varsity divisions. Although these differences were small, they did meet statistical significance. The reasons for these differences are not known and warrant further investigation.

Analysis of discrete field data, when applied to all racing classes collectively, reveals that rider inexperience, the technical nature of the trail, and negotiating turns were frequently reported to be associated with injury occurrence. This highlights a mismatch between riders' self-perception of ability versus the skill required for the terrain and speed they are riding as a primary causative factor in crashing. Injury prevention efforts should emphasize the need for additional skills training for riders. The frequency with which washing out or losing traction of the front wheel was implicated in injury occurrence in the text analysis further emphasizes the need for additional skills training. The interactions between rider, bicycle, speed, and trail characteristics have important implications for injury prevention (²⁹) and warrant further study.

Organized youth mountain biking is a relatively new sport; the rapid growth in participation has resulted a dearth of highly qualified adult coaches. All NICA coaches are volunteers, and in contrast to more established sports, a majority of coaches did not grow up mountain biking. NICA is rapidly expanding its skills training program for coaches, which will improve skills training for youth riders. In addition, as more graduates of NICA teams become coaches themselves, the available pool of skilled coaches is growing.

As discussed in the Methods section, exposure and injury data were tracked via an electronic injury reporting system with data entered weekly by a designated reporter. For the purpose of this study, exposure was defined as one student-athlete participating in one practice or race. Alternative ways to report exposure could include riding time, distance ridden, vertical distance ridden, and speed at time of crash. To acquire this information, riders would need GPS units with GPS data synchronized to the NICA ISS. Future research may involve beta testing the acquisition of GPS data to obtain more granularity of exposure and risk. Unfortunately, during the first 2 years of data collection, compliance with completing exposure reports was highly variable from team to team and was insufficient to calculate injury incidence. Therefore, only injury proportions and injury event proportions have been reported. The reason for the variability in exposure reporting is under review. Recognizing the importance of capturing exposure data (^8,13), additional efforts by NICA are underway to increase compliance with exposure reporting.

Limitations

The NICA ISS faces some of the same limitations as other, large sports epidemiology investigations in youth athletes. One limitation is that data are not always entered by medical personnel, but rather by a volunteer designated reporter on each team who may or may not be medically trained. Unfortunately, it is not practical in a study of this size to obtain formal medical records for all injuries or have a medically trained individual complete the injury report forms for all teams. A second important limitation is that teams did not provide adequate exposure data during the first 2 years of data collection to allow for calculation of injury incidence rate. NICA is making extensive efforts to increase compliance with exposure reporting from all teams. A third important limitation is that it cannot be assured that all injuries that meet the inclusion criteria were reported, potentially causing an underestimate in the total number of injuries. However, teams are mandated by NICA to report all qualifying injuries for the purposes of insurance reporting as well as for the ISS, and frequent reminders are provided to coaches, designated reporters, parents, and student-athletes about complying with this mandate. Another limitation is that data validation has not yet been completed, but is planned for the future, after enough data have been collected. Finally, the NICA ISS focuses on acute traumatic injuries, which by far account for the majority of injuries in this sport. The inclusion of overuse injuries and medical illness remains under consideration for the future.

Conclusions

An effective, nationwide ISS for injuries sustained in youth mountain bike racing has been successfully implemented, now with 2 years of data collection completed. To our knowledge, this is the largest mountain biking ISS in existence. This NICA ISS also tracks injuries in coaches, but only student-athlete injuries are reported here. The first year of data collection has provided early insights into injury characteristics in this sport; these preliminary data have been affirmed in the second year of data collection, now with more than 40,000 student-athlete-years of data. Overall, student-athlete injury proportions appear to be lower than those seen in some other high school sports. Many injuries are relatively mild, including abrasions and contusions. However, some injuries are more significant, including concussions or possible concussions, fractures, and dislocations, and result in trips to an emergency room and sometimes hospital admission. A mismatch between rider experience and terrain being ridden at the time of a crash is reported to be an associated factor for more than half of all injury events, emphasizing the importance of additional skills training for riders. Injury proportions are slightly higher in women than in men, and female student-athletes sustain lower limb injuries more often than male student-athletes. Further research is needed to confirm these findings and look for trends as this relatively young sport evolves and changes. We plan to continue conducting longitudinal injury epidemiology research in this athlete population. Future directions with an emphasis on injury prevention, informed by data collected from the ISS, include development of an athlete-specific injury prediction algorithm. The first prospective, controlled, multiarm injury prevention study is planned for the fall 2021 season.