Middle School Injuries: A 20-Year (1988–2008) Multisport Evaluation

Participants and Setting

Punahou School, located in Honolulu, Hawaii, is a private, nonsectarian school with 3700 students in kindergarten through grade 12. Middle school athletes range in age from 12 to 15 years old and are in grades 7 to 9. The average enrollment for grades 7 and 8 is 740 students, with 57.4% of boys and 51.7% of girls participating in athletics. Although interscholastic league rules allow ninth graders (400 students) to participate at the middle school level, internal school policy limits ninth-grade participation in middle school sports (6.2% of ninth-grade girls, 21% of ninth-grade boys; Table 1). The ninth-grade boys may participate in middle school or high school football, depending on ability. Over 20 years, a total of 8081 boys participated in 14 sports and on 18 teams, with 5960 girls participating in 15 sports and on 19 teams (Table 2). Softball, baseball, basketball, soccer, and volleyball fielded 2 teams each for both boys and girls. During this study, 13 new sports were added to the athletic program. These consisted of boys' and girls' swimming (1994–1995), boys' and girls' diving and girls' water polo (1996), cheerleading and boys' and girls' tennis (1997), boys' and girls' golf (1998), girls' wrestling (1999), and boys' and girls' judo (2005). Gymnastics for girls was introduced in 1993 but dropped in 2007 (Table 2). This study was approved by the University of Hawaii Institutional Review Board.

Table 1.  Middle School Enrollment and Athletic Participation, 1988–2008

View Fullscreen

Table 2.  Injury-Risk Sex Comparisons by Sport Among Middle School Athletes, 1988–2008

View Fullscreen

Data Collection

Five certified athletic trainers were on staff during the study period, ensuring consistency of evaluation and treatment. Of the 3 full-time athletic trainers, 1 had been on staff for the full 20 years of this study and the others for 15 and 5 years, respectively. Two sports medicine physicians (physiatrists) were contracted by the school as team physicians for all levels and sports during the time of the study. They provided coverage for all football games and evaluated students referred by the athletic trainers in a weekly clinic. Athletes were then further referred to their primary care physicians for follow-up evaluation and treatment.

Injuries were evaluated by the athletic training staff and recorded daily into a database program designed by the principal author (G.B.). The data were initially recorded using AppleWorks integrated software (versions 2 and 3; Apple Inc, Cupertino, CA), but all injury data are now recorded and stored on FileMaker Pro (versions 5 to 11; FileMaker, Inc, Santa Clara, CA). The physician's diagnosis and time lost from activity were updated as the athlete returned to play. In 2001, the Vienna concussion guidelines²⁷ were implemented and then revised to align with both the Prague²⁸ and Zurich²⁹ concussion guidelines.

The Athletic Training Room Evaluation Form was used to collect all injury data, including name, grade, sex, sport, coach, date of injury, date of return to activity, team session, body part, category (eg, strain, sprain), evaluation results, severity of injury, activity status, and the evaluator's impressions and proposed treatment and rehabilitation. The player's position, activity at the time of injury, and field surface and conditions were also recorded.

Injuries

Throughout the study, injury was defined as any athlete complaint that required the attention of the athletic trainer, regardless of the time lost from activity.³ Evaluation forms were completed for all athlete complaints. Lacerations requiring sutures or advanced medical care were recorded, but a nosebleed or simple abrasion was not recorded. Five injury-severity classifications were used: (1) minor, no time lost from activity; (2) mild, 1 to 7 days lost; (3) moderate, 8 to 21 days lost; (4) severe, 22 or more days lost; and (5) catastrophic, permanent disability, dismemberment, or death.³

Athlete-Exposures

An AE was defined as 1 athlete participating in 1 practice or game. Participation was not recorded daily. Exposures were estimated using team rosters and number of practice and game dates. The average middle school athletic season was 11 to 12 weeks of training and competition.

Data Analysis

The total injury rate was calculated as the total number of injuries divided by the total number of AEs and was expressed as injuries per 1000 AEs (injuries/1000 AEs). Total injury rates were calculated separately for all injuries and for time-loss (at least 1 day) injuries.

Sex-Matched Sports

The rate ratio (RR) was used to compare the injury rates in girl and boy athletes competing in the same sport (Table 2):

If the girls' to boys' sport RR was above 1.0, the girl was more likely to sustain an injury than the boy. If the girls' to boys' sport RR was below 1.0, the boy was more likely to sustain a higher injury rate.

Exposure Setting (Practices Versus Games)

Practice and game time-loss injury-rate comparisons were based on the following RR (Table 3):

If the practice/game RR was above 1.0, the athlete was more likely to sustain an injury in practice than during a game. If the practice/game RR was below 1.0, the individual was more likely to sustain an injury during a game than in a practice.

Table 3.  Injury-Risk Estimates During Practices and Games by Sport and Sex, 1988–2008

View Fullscreen

For time-loss injuries only, RR comparisons were also calculated for body location injured, type of injury, severity of injury, and specific injury types. The 95% confidence interval (95% CI) was used to determine the statistical significance for all rate ratios. Confidence intervals that did not include 1.0 indicated statistical significance (P ≤ .05).

All analyses were conducted with the STATA (version 5.0; STATA Corporation, College Station, TX) and SPSS (version 18.0; SPSS, Inc, Chicago, IL) statistical packages.

Sport Participation

Over the 20-year period, 14 038 athletes were included in this analysis, with a greater number of boys (n = 8078) than girls (n = 5960; Table 2). By sport, track had the greatest number of athletes participating (n = 3126; 1537 girls, 1589 boys), followed by cross-country (n = 1466; 756 girls, 710 boys), and football (n = 1435). Boys had a higher participation level (n = 500 626 AEs) than girls (n = 366 963 AEs; Table 2). By sport, track had the highest participation level (n = 146 670 AEs) followed by soccer (n = 106 800 AEs) and football (n = 99 810 AEs).

Injury Rate

Boy athletes reported 4279 injuries, with 1930 injuries (45.1%) resulting in at least 1 day lost from activity (Table 2). Girl athletes reported 2781 injuries, with 1077 injuries (38.7%) resulting in at least 1 day lost from activity. Most injuries (4053) were classified as minor (no time lost from participation) and accounted for 54.9% and 61.3% of all injuries for boys and girls, respectively.

The injury rate for girls was highest in 1990 and 1998 and gradually declined over the following 10-year period (Figure 1), even with the addition of 10 new sports for girls. The injury rate for boys remained relatively the same (Figure 2). The time-loss injury rate for girls rose gradually until 1998 and then gradually declined (Figure 1), whereas the time-loss injury rate for boys has gradually increased over the 20-year period (Figure 2).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Sex Comparisons

In sex-matched sports, middle school girls were more likely to sustain any injury (RR = 1.15, 95% CI = 1.1, 1.2) or a time-loss injury (RR = 1.09, 95% CI = 1.0, 1.2) than middle school boys (Table 2). Although middle school girls were more likely to sustain an injury in practice (RR = 1.97, 95% CI = 1.7, 2.4) than during a game, the risk of injury was similar during practices and games for middle school boys (RR = 0.99, 95% CI = 0.9, 1.1; Table 3).

Middle school girls were twice as likely to sustain a mild injury during a practice than in a game (RR = 2.05, 95% CI =1.7, 2.5), but no differences were noted during practices or games for moderate or severe injuries. For boys, the risk of a severe injury was 55% less during practices than during games (RR = 0.45, 95% CI = 0.3, 0.6; Table 4). We found no differences between practices and games for mild or moderate injuries.

Table 4.  Injury by Severity in Practice and Game Settings by Sex, 1988–2008

View Fullscreen

Specific sex-matched sport differences and specific injury differences are reported under these topics.

Exposure Setting (Practices Versus Games)

For the total sample, practice injury rates were higher than game injury rates (RR = 1.24, 95% CI = 1.1, 1.4; Table 3). For all girls' sports, a higher injury rate was recorded in practices (3.297/1000 AEs) than in games (1.673/1000 AEs) (RR = 1.97, 95% CI = 1.7, 2.4). For boys, the risk of injury was nearly identical during practices and games (RR = 0.99, 95% CI = 0.9, 1.1).

For softball (RR = 3.10, 95% CI = 1.7, 6.4), baseball (RR = 1.69, 95% CI = 1.1, 2.6), boys' cross-country (RR = 2.50, 95% CI = 1.2, 6.0), girls' track (RR = 2.62, 95% CI = 1.8, 4.0), boys' basketball (RR = 1.72, 95% CI = 1.1, 2.8), and girls' (RR = 9.71, 95% CI = 2.6, 82.3) and boys' volleyball (RR = 4.92, 95% CI = 2.0, 15.7), the injury risk was greater during practices than in games. Only in boys' wrestling (RR = 0.50, 95% CI = 0.3, 0.8) and football (RR = 0.49, 95% CI = 0.4, 0.6) was the likelihood of sustaining an injury greater during a game than during a practice.

Injury Severity

Minor injuries accounted for 57% of all reported middle school injuries. Middle school girls reported that 61.2% of all injuries involved no time loss, compared with 54.8% of all boys' injuries. Although mild injuries were more likely to occur during practices than during games (RR = 1.35, 95% CI = 1.2, 1.5), severe injuries were more likely to occur during games (RR = 0.54, 95% CI = 0.4, 0.7; Table 4). Girls were twice as likely to report a mild injury during practice than during a game (RR = 2.05, 95% CI = 1.7, 2.5), whereas boys were less likely to incur a severe injury in a practice than in a game (RR = 0.45, 95% CI, 0.3, 0.6; Table 4).

Body Part

Ankle, knee, thigh, shin/calf, and wrist/hand/finger injuries were the most frequently reported time-loss injuries by middle school athletes. Ankle and knee injuries were the time-loss injuries reported most overall (n = 581, rate = 0.669/1000 AEs, and n = 384, rate = 0.442/1000 AEs, respectively) and during practices (0.695/1000 AEs and 0.459/1000 AEs, respectively) and games (0.576/1000 AEs and 0.380/1000 AEs, respectively; Table 5).

Table 5.  Injury by Sex and Body Part in Practice and Game Settings, 1988–2008

View Fullscreen

Shin/calf (RR = 4.48, 95% CI = 3.0, 9.2) and foot/toe (RR = 2.73, 95% CI = 1.6, 4.9) injury rates were greater during practices than during games. Only the injury rate for head/face injuries (RR = 0.57, 95% CI = 0.4, 0.9) was higher during games than during practices.

For girl athletes, higher injury rates for the ankle (RR = 1.59, 95% CI = 1.1, 2.3), knee (RR = 1.85, 95% CI = 1.2, 3.1), thigh (RR = 1.62, 95% CI = 1.1, 2.7), shin/calf (RR = 7.18, 95% CI = 3.4, 18.1), and foot/toe (RR = 4.50, 95% CI = 1.5, 22.6) were reported during practice sessions than during games. For boys, although the risk of sustaining a head/face (RR = 0.42, 95% CI = 0.3, 0.7), elbow/forearm (RR = 0.51, 95% CI = 0.3, 0.9), or trunk/abdomen (RR = 0.46, 95% CI = 0.3, 0.8) injury was higher during games than during practices, shin/calf (RR = 3.46, 95% CI = 1.7, 8.2) and foot/toe injuries (RR = 2.20, 95% CI = 1.2, 4.4) were more likely to be reported during a practice than a game.

Injury Type

Overall, strains, sprains, and tendinitis were the injury types reported most frequently by middle school athletes (Table 6). The rates for strains (RR = 1.46, 95% CI = 1.3, 1.9), sprains (RR = 1.13, 95% CI = 0.9, 1.3), and tendinitis (RR = 7.14, 95% CI = 4.6, 11.5) were higher during practices than during games, whereas contusions (RR = 0.58, 95% CI = 0.4, 0.7) and fractures (RR = 0.77, 95% CI = 0.5, 0.9) were more likely to occur during games.

Table 6.  Injury Rates by Sex and Injury Type in Practice and Game Settings, 1988–2008

View Fullscreen

For girl athletes, the injury rate of reporting tendinitis was 12 times higher (RR = 12.39, 95% CI = 5.8, 30.5) during practices than during games. Strains (RR = 1.72, 95% CI = 1.1, 1.9) and sprains (RR = 1.40, 95% CI = 1.2, 2.4) were also more likely to occur during practices than games. Tendinitis (RR = 4.67, 95% CI = 2.7, 8.7) and strain (RR = 1.48, 95% CI = 1.2, 1.9) rates were higher for boys during practice than during games, but the rates for contusions (RR = 0.55, 95% CI = 0.4, 0.7), fractures (RR = 0.54, 95% CI = 0.4, 0.8), and neurologic injuries (RR = 0.41, 95% CI = 0.2, 0.6) were greater during games than during practices.

Specific Injury Types

Concussions

The overall rate for concussion was 0.067/1000 AEs, with the rate almost 3 times higher (RR = 2.83, 95% CI = 1.5, 5.9) for boy athletes (0.092/1000 AEs) than girl athletes (0.033/1000 AEs) (Table 7). Middle school athletes were at higher risk of concussion during games than during practices (RR = 0.33, 95% CI = 0.2, 0.6). Football accounted for 35 of the 58 concussions (2.2% of all football injuries, rate = 0.35/1000 AEs) with 28 resulting in time lost from activity. The remaining concussions were spread among 6 sports, with basketball (n = 7, rate = 0.085/1000 AEs), soccer (n = 7, rate = 0.066/1000 AEs), and wrestling (n = 6, rate = 0.15/1000 AEs) participants recording the most injuries. The boys' concussion rate was 0.063/1000 AEs for practices and 0.206/1000 AEs for games (RR = 0.31, 95% CI = 0.2, 0.6). Concussions were more likely to occur during competition for boys. Girls participating in soccer (n = 6) and basketball (n = 5) accounted for the majority of concussions. Girls reported a higher rate of concussion during games (0.061/1000 AEs) than during practices (0.025/1000 AEs; RR = 0.40, 95% CI = 0.1, 1.6), but the difference was not significant.

Table 7.  Specific Injuries by Sex in Practice and Game Settings, 1988–2008

View Fullscreen

Main Findings

Our results indicate that (1) the overall risk of injury was greater for boys for non–time-loss or time-loss injuries, (2) football had the highest rate of injury for both non–time-loss and time-loss injuries, (3) the risk of time-loss injuries varied for sex-matched sports, (4) the overall risk of injury was greater during practices than in games, especially for girls, (5) most injuries were mild in nature, (6) the ankle and knee had the highest incidences of injury, (7) strains, sprains, and tendinitis were the most common injury types, (8) ankle sprains were the most frequent specific injury, and (9) boy athletes were more likely to suffer concussions than girl athletes.

To our knowledge, this is the largest prospective, longitudinal study of middle school sport injuries to date. The compilation of 20 years of injury data at the same site and with the same athletic trainers is unique for both middle and secondary school programs. The duration of this study is also substantially longer than other studies at the secondary or middle school levels.^3,5,7,30,31

All Injury Versus Time-Loss–Injury Rates

The reporting of non–time-loss versus time-loss injury rates remains debatable. Presently, rates based on time-loss injuries^6,7,9,11 have been advocated because injuries that cause time loss may be more accurately recalled.^30,32 However, a time-loss injury definition that requires a coach, parent, athletic trainer, or physician report may underestimate the true injury burden.^3,30,32 Even though many injuries are not serious in nature, injuries considered minor in nature and that do not cause immediate impairment may still have long-term consequences.³⁰ Additionally, reporting non–time-loss injuries may provide a better perspective of the daily workload of the athletic trainer. In reporting both non–time-loss and time-loss injuries, we noted that the rates for non–time-loss injuries were higher than for time-loss injuries. These findings support those of other youth^3,30 and collegiate³² studies in which rates have been reported using both injury definitions. Our results suggest that non–time-loss injury rates should be used to reflect the true extent of the problem, but the difficulty of collecting accurate data in most settings (eg, lack of enough athletic trainers at the middle school level to identify and record the non–time-loss injuries) may preclude this recommendation. Given that our study may be the first to describe injuries in most middle school girls' and boys' sports, additional research is needed to confirm these findings.

Although middle school sport participation levels increased over the 20-year period, the total number of injuries decreased. However, the number and percentage of day-loss injuries increased over the same time period. The greater participation by girls in athletics, both interscholastic and club sports, may explain the change in the occurrence of injuries. Over the 20-year period, the girl athletes may have become more competitive and less likely to report a minor injury to the athletic trainer or coach in order to continue participation in the sport. The reasons for these changes in severity are not clear. We speculate that the athletes may be reporting only those injuries they consider more problematic, rather than the nuisance injuries that may have little effect on their participation level.

Because data for middle school sport injuries are limited, we used reports from secondary school and some collegiate studies for comparison. Based on our data, middle school girls were more likely to sustain an injury during athletic participation than their male counterparts in the same sport. This finding is consistent with secondary school injury surveillance reports by Comstock et al,¹¹ Rechel et al,³³ and Powell and Barber-Foss.⁷ However, similar to reports at the high school and collegiate levels, injury rates varied by sport.^{11,14,32,34–36}

Sex Comparisons

Overall, we observed a difference between practice and game injury rates for girls but no such difference for boys. This is consistent with findings of secondary school studies.^8,15,37 Middle school cross-country girl athletes reported a higher injury rate than their male counterparts, which is also consistent with other studies.^7,8,37 Rauh et al⁸ suggested that girl runners might be more apt to report minor pain complaints than boys. Girls' and boys' soccer players reported higher injury-incidence rates in games than in practices, which is consistent with previous studies.^15,38

Practices Versus Games

Our findings indicated that the injury risk was greater during practices than during competition for all sports except football, gymnastics, and male wrestling. This finding is in contrast to previous observations at the secondary school level,^{7,9,10,16,33,39} where higher rates occurred during competitions than during practices in all sports. Our findings are also noteworthy because practices are conducted daily in the competitive interscholastic middle school setting, compared with community-based programs in which practices may take place only twice per week. Dompier et al³⁰ noted that youth football players sustain more game injuries, which is consistent with our findings. In a study of school children, Backx et al⁴⁰ reported that basketball and soccer players were at 3 times greater risk of injury during games than during practices. This result is in direct contrast to our finding that middle school boy basketball players were at more risk of injury during practices than during games. Thus, except for a few sports, such as football, gymnastics, and wrestling, in which coverage is known to be needed at competitions, we suggest that middle school administration and athletic departments also ensure that appropriate medical coverage is available at practices for those middle school sports with greater injury risks.⁴¹

Team Contact Sports

Individual Sport: Wrestling

Contemporary published data on middle school–aged boy wrestlers are limited to tournament competition data,⁴⁶ with no practice injury data available. To our knowledge, injury data for female wrestlers at any level have not been reported. In our study, only 1 match injury was reported, which likely reflects the low number of interscholastic participants and matches. The overall injury rate was nearly identical for boys' and girls' wrestling. Girl wrestlers reported higher practice injury rates than boy wrestlers, and boy wrestlers' injuries were more likely to occur during matches, but the limited participation numbers for girls did not allow for appropriate statistical comparisons. The higher rate during competition for middle school boy wrestlers is consistent with that of high school^7,33 and collegiate⁴⁷ male wrestlers. Knowles et al³⁴ and Rechel et al³³ noted that game injuries were more frequent but reported lower injury rates in secondary school wrestlers. Agel et al⁴⁷ reported a much higher match injury rate for collegiate male wrestlers.

Noncontact Sports: Team

Noncontact Sports: Individual

Severity of Injury

The reporting of minor, no–time-loss injuries may provide a more accurate picture of the athletic trainer's workload. Minor injuries accounted for more than half (57%) of all reported middle school injuries for both girl and boy athletes. To date, only 3 published studies^3,30,32 have reported no–time-loss injuries, and those percentages are similar to our findings. In the only secondary school study,³ 60% of injuries did not involve time loss. Youth football³⁰ (58.6%) and collegiate³² (78%) studies have also reported similar estimates for no–time-loss injuries. Although mild injuries (1–7 days lost) for girls and for the overall sample were more likely to occur during practices, severe injuries (22 or more days lost) were more likely to occur during games, especially for boys.

Injury Location

Lower extremity injuries accounted for 70% of all days-lost injuries for middle school athletes. This finding is comparable with data reported at the high school^7,11 and collegiate levels.³⁵ Ankle and knee injuries have been the most commonly reported injuries in published literature,^3,5,7,34 and our results indicate that these body parts are also the most likely to be affected at the middle school level. Ankle injuries had the highest injury rate for both sexes in practices and games. Knee injuries were the second highest rate for boys and third highest rate for girls. Shin/calf injuries were the second most frequently reported injury for girls and third most frequently reported for boys. Wrist/hand/finger injuries were more common for boys. In studies assessing multiple high school sports, Beachy et al³ observed that shin injuries were the third most reported injury for secondary school athletes, with girls reporting a higher injury frequency than boys. At the collegiate level, Lanese et al²⁵ noted that 7% to 8% of injuries involved the shin/calf. In summary, the high incidence of shin/calf injuries reflects the need to monitor conditioning and running programs for the middle school–aged athlete.

Specific Injury Types

Small injury numbers may warrant cautious data interpretation, regardless of injury site and sex. This is true for the rotator cuff and ACL injuries reported in this study because the small numbers limit the comparative value for middle school athletes. We will try to retain that objectivity when discussing these injuries.

Concussions

Reports on the incidence and effects of concussion have primarily focused on football players at all competitive levels. We found that 2.2% of middle school football players incurred a concussion, a value consistent with that noted by Dompier et al³⁰ (2.7% of youth football players) but unlike that of Turbeville et al²⁰ (none among middle school–aged football players over 2 years). Our concussion injury rates of 0.067/1000 AEs and 4.0 per 100 football players are lower than those reported at the high school and collegiate levels. Gerberich et al⁵⁰ reported 19 injuries per 100 secondary school football players (24% of all injuries reported). Guskiewicz et al^51,52 stated that 5.6% of secondary school and 6.3% of collegiate football players surveyed sustained at least 1 concussion. Our lower rate may be attributed to the lack of physical maturity and the associated reduced collision intensity, but football continues to have the highest concussion rate at the middle school level.

In this study, concussions accounted for 0.4% of all injuries to girls and 1.0% of all injuries for boys. Overall, boy athletes were 3 times more likely to incur a concussion than girl athletes. In a study of youth soccer players reporting to the emergency room, Adams and Schiff ⁵³ also found that boys were more likely (7.8%) to sustain a concussion injury than girls (0.5%). However, when we excluded concussion injuries that occurred in football players, the risk was similar. At the high school level, the risk of concussion injury by sex appears equivocal, with some studies indicating a greater risk for girls^9,39 and others⁵⁴ showing a greater risk for boys. At the collegiate level, however, the risk of concussion for sex-matched sports appears to be higher for female than male athletes.³⁵

We noted that the incidence of concussion injuries was 3 times higher during games than during practices. This finding is in agreement with studies of middle school,²⁰ high school,^7,11,32 and collegiate athletes^35,55 and suggests that the intensity of impact and risk taking may be greater in games.

Concussion guidelines changed during the 20-year period of this study. During the 1980s and 1990s, the Colorado Medical Society⁵⁶ and the American Academy of Neurology guidelines⁵⁷ allowed return to activity if symptoms resolved within 15 minutes of the injury. The 7 athletes with no–time-loss concussions were in this category.

Anterior Cruciate Ligament Injuries

Injuries to the ACL are concerning due to the severity and the time lost from participation and because females are at greater risk of injury than males in matched sports. Only 8 ACL injuries (0.6% of all injuries) occurred during the 20-year period. The overall injury rate of 0.009/1000 AEs was much lower than the rates reported in high school studies^7,9,11 and the 0.15/1000 AEs observed by Hootman et al³⁵ for all collegiate sports. In collegiate football, ACL injuries accounted for 3% of all collegiate sport injuries, with an injury rate of 0.018/1000 AEs.³⁵ We found similar risks of ACL injuries for girl and boy middle school athletes but different risks than those reported at the high school,³⁶ military,⁵⁸ and collegiate⁵⁹ levels. It may be that the female middle school athletes had not reached the pubertal or maturational levels that have been suggested to affect their hormonal, structural, and neuromuscular traits and possibly put them at greater risk of injury.^36,58

Medial Tibial Stress Syndrome

Our finding that the risk of MTSS was higher for girl athletes than boy athletes is consistent with values reported at the high school level.^60,61 For both girl and boy athletes, MTSS was more likely to be reported in practices than in games. We are unaware of any studies of MTSS by participation setting. The higher occurrence of MTSS during practices may be related to training errors, including overly repetitive activities or training regimens with short recovery times.^31,60 Further research is needed to help identify these factors.

Limitations

The primary limitation of our study was that all injury and participation data were collected from only 1 private school that has sufficient resources to employ more than 1 full-time athletic trainer. Thus, the extent to which our findings are generalizable is unknown. Therefore, these results await comparison with the results of future investigations among middle school athletes. Because only a few athletic trainers were involved in this study, evaluating and reporting of the data were performed consistently and with little variation. An additional potential limitation of our study was the possibility of nonreporting bias. Again, the injury-reporting system at Punahou School is longstanding, and the procedures for middle school athletes to report any injury, regardless of severity, have been fostered throughout the 20-year period. Some sports had too few injuries or participation numbers by sex for ample statistical comparisons. These sports may be new to the middle school population, resulting in limited participation, or the sport may have a lower injury risk. Some athletes may have had self-treated injuries (eg, tendinitis, minor contusions) that were not reported to the athletic training staff. Finally, most middle schools in the United States do not allow ninth-grade participation in middle school sports.

We hope that our initial findings for these sports will provide the impetus for all middle schools to report their data in these sports for comparative purposes. We used an injury-surveillance design, so detailed data on each injured athlete were not available. Thus, efforts should be focused toward implementing epidemiologic studies designed to identify risk factors in these athletes at the middle school level.