Abstract
According to recent epidemiological studies, football has the highest injury rate of any team sport in the United States, at both the high school and collegiate levels. 18,31 The high risk of injury makes it imperative to identify players and environments at risk for specific injuries so that scientifically supported injury prevention programs may be developed and implemented. 36
Muscle strain injuries are a cause of considerable disability in football. In high school football, muscle strains have been reported to comprise between 12% and 24% of all injuries. 8,23,29,34 Muscle strains in a major college football team were reported by Heiser et al 17 to occur at a rate of 7.7% per player-year. Furthermore, muscle strain injuries accounted for 18.9% of spring practice and 22.2% of fall practice injuries in collegiate football across all divisions over the course of 16 seasons from 1988 to 2004. 10 Of these, the thigh accounted for nearly half of all strain injuries during practice.
In a recent 10-year study of preseason injuries of 1 National Football League (NFL) team, Feeley et al 11 reported that muscle strains account for 46% of practice injuries and 22% of preseason game injuries. Hamstring strains were the second most common preseason injury, with an injury rate of 1.79 per 1000 athlete-exposures for practices and 4.07 per 1000 athlete-exposures for games.
A growing amount of research categorizing risk factors associated with hamstring strain injuries at the elite level exists for sports such as soccer 43 and rugby. 4,9 In particular, a recent emphasis on hamstring strains can be found in the literature concerning Australian Rules Football. 12,25,41 Despite their high frequency and potential for prevention, there have been few attempts to characterize the specific relationships between hamstring strains and the factors associated with playing the competitive gridiron football season. Of these, most studies have suffered from small sample sizes, unreliable injury definition, and short surveillance periods. 34,38
The purpose of this study was to characterize the factors that are associated with hamstring injuries in NFL players during a 10-year period. We were particularly interested in whether injuries were incurred in activities related to football’s ballistic collisions, whether they were similar to those experienced in noncontact activity, and whether identifiable periods or activities of higher injury density exist. We hoped that identifying player characteristics and activities associated with a high incidence of hamstring strains would provide foci for preventative interventions.
Materials and Methods
The NFL athletic trainers collected the injury and exposure data prospectively between 1989 and 1998. The data included all players on each team’s roster and focused on activities from the first practice day through the last game. The NFL added 2 teams in 1995, and thus the data reflect 28 teams from 1989 to 1994 and 30 teams from 1995 to 1998. The teams’ data reports were submitted to the central database every 2 weeks throughout the year. The athletic trainers conformed to uniform reporting definitions of injuries and exposures that were established by the NFL Injury Surveillance System. These measures resulted in a high level of compliance, standardization, and consistency of data.
Data regarding recorded injuries included team, date of injury, activity the player was engaged in at time of injury, position played, mechanism of injury, and history of hamstring strain injuries. From 1989 to 1995, a reportable injury in the NFL Injury Surveillance System was any injury that caused the athlete to miss at least 2 days of participation in either practices or games. Beginning in 1996, this definition was modified to reflect 1 day of lost time. For consistency, the data in this article reflect only those cases that met the definition for reporting used before the 1996 season. For the purpose of this study, hamstring strains were defined as any muscle strain injury to the posterior thigh resulting in a minimum 2 days of lost participation in either practice or games. Injury severity was calculated and identified as minor, moderate, and major. Injuries that resulted in less than 7 days of missed participation were defined as minor, those that resulted in between 7 and 21 days missed were defined as moderate, and those resulting in more than 21 days missed were defined as major. Injuries classified as reinjuries reflected injuries in which the athlete reported previously straining the hamstring muscle group on the same side during the current season.
The number of players available during preseason activities is variable by week, with the final team roster set before the first regular season game. To account for this and other seasonal variations in number of players practicing or playing at any one time, the number of players who participated in each session (practice or game) was recorded along with the number of games and practices. The total numbers of games and practices are used to estimate injury rates per session. Aggregating the man-sessions creates an estimate of the number of opportunities for injury and is referred to as athlete-exposures (A-E). This provides for a denominator with which to calculate injury rates at different times during the season and to compare relative injury risk for practice versus games, various positions, and different types of plays. By dividing injury incidence by both the number of years (10) and the number of NFL teams (28-30), we get the mean injury incidence per NFL team in one competitive year, referred to as the injury rate per team-season. In all, 288 team-seasons were studied.
The analysis includes a review of the injury patterns for player position, player activities, onset of injury, time lost from participation, injury rates per session, injury rates per team-season, and injury rates per 1000 A-E. Data analysis used SPSS 10.0 (SPSS, Inc, an IBM Company, Chicago, Illinois).
Results
Between 1989 and 1998, 1716 hamstring strains were reported among players on NFL teams (Table 1). The per season range over the 10-year study was 132 to 210. From 1993 through 1998, the number of injuries remained consistent at about 183 injuries per year (Figure 1) with a range of 160 to 210. During the study years, 1129 different players incurred the 1716 reported injuries. Annually, a mean of 144 different players sustained a mean number of 172 hamstring muscle injuries.
Total Injuries and Exposures Over 10-Year Study Period a
A-E, athlete-exposure; IR, injury rate.
Injury incidence per year over 10-year study period.
Severity
Hamstring strains resulted in a mean of 2222 days lost annually. A mean of 13.2 and a median of 9 days were lost per injury. The average season showed 213 man-games lost because of hamstring injuries with a median of 1.2 man-games per injury. Injury severity was classified as minor in 40.6% of the injuries, moderate in 41.0%, and major in 18.3% of all hamstring muscle injuries.
Type of Session
During the study, a majority (52.7%; n = 904) of the reported hamstring injuries occurred in practices, with 47.3% (n = 812) occurring in games. This translates to an injury rate per 1000 A-E of 0.47 for practices and 2.7 for games and reflects a 5.74 relative risk of hamstring injury during games relative to practice sessions.
Temporal Aspects
The data were analyzed in relation to the month in which the injury occurred. The majority of reported injuries (53.1%; n = 912) occurred during the preseason months of July and August despite the fact that the preseason period is approximately 7 weeks long compared with the 16-week regular season. Of the total hamstring muscle strain injuries, 45.7% (n = 785) occurred in the regular season. Of all practice injuries, 79.8% (injury rate [IR] = 0.82/1000 A-E) occurred in the preseason (Table 2); 56.0% of the practice-related injuries occurred in July, 23.8% occurred in August, and only 20.3% occurred after September 1 (Figure 2). Of all practice-related injuries, 20.1% (n = 182; IR = 0.18/1000 A-E) occurred during the regular season. Of the game-related injuries, 21.8% (n = 191; IR = 2.47/1000 A-E) occurred in the preseason, and 74.3% (n = 603; IR = 2.92/1000 A-E) occurred in the regular season.
Injury Rate by Session Type and Time of Year a
A-E, athlete-exposure; IR, injury rate.
Injury incidence by week of season for both practice and games sessions of 10-year study period.
Reinjuries
In total, 16.5% (n = 283) of total hamstring injuries, including 19.7% of regular-season injuries and 13.8% of preseason injuries, were reinjuries. During the preseason, the reinjury rate was 12.7% (n = 91) for practices and 17.6% (n = 35) for games. For the regular season, the reinjury rate rose to 22.0% (n = 40) for practices and 19.1% (n = 115) for games.
Player Position
Of the 1716 injuries, 0.4% (n = 7) did not report a position. For the remaining 1709 injuries, position played proved to be a significant factor in the incidence of injury (Table 3). Offensive players sustained more injuries (45.1%; n = 771) than defensive players (41.7%; n = 716), whereas the players working on the special teams unit sustained 13.0% (n = 222) of the injuries during the study period. When the units are more clearly defined, the defensive secondary alone accounts for 23.1% of the hamstring injuries and shows an injury rate per team-season of 1.37. Offensively, the wide receivers represent 20.8% of the injuries, with a corresponding injury rate per team-season of 1.23.
Injuries by Position
Estimated injuries across 30-team National Football League season.
Total injuries for which player position was reported.
An estimate of the annual number of hamstring injuries expected in the NFL equals the injury rate per team-season multiplied by the 30 teams and is 178. By position, the defensive secondary group in the NFL has a mean hamstring injury frequency of 41 per season, the wide receivers 37, the running backs 22, and the linebackers 20.
Player Activity
The reported injuries were grouped into those that occurred as a result of contact or collision and those in which there was no contact. During the study period, 81.5% (n = 1318) of the hamstring injuries were identified as a noncontact player activity, and this included 88.5% of practice injuries and 73.0% of game injuries (Figure 3). Among the player positions contributing to the noncontact injuries, the defensive secondary accounted for 26.4% (n = 287) of the reported noncontact injuries and the wide receivers 20.8% (n = 298). For those respective positions, noncontact injuries accounted for 92.0% and 93.5% of their total injuries. In contrast, the offensive and defensive linemen cited 69.7% and 43.2%, respectively, of their injuries as having occurred as a result of contact. When analyzing hamstring injuries sustained during noncontact, 981 (74.4%) were attributed to sprinting. Noncontact skill drills accounted for 10.2% (n = 135) of strain injuries.
Injury mechanism by preseason and regular season for practice (A) and games (B). A-E, athlete-exposure.
Team Activity
For injuries occurring during all NFL games, more hamstring muscles were strained during passing plays (n = 295; 36.8%) than during running plays (n = 147; 18.4%). Special teams injuries accounted for 189 (23.3%), with the kickoffs accounting for 140 (74.1%) of those injuries. In the passing play injuries, 225 (77%) of them were noncontact, whereas 69 (45.1%) of the rushing play injuries were noncontact. Of the 140 injuries sustained during the kickoff, 120 (85.7%) were noncontact.
During regular-season games, we found that 230 (42.9%) muscle strains occurred during passing plays, 102 (19.0%) of the injuries occurred during rushing plays, and 204 (38.1%) occurred during special teams play (kickoff and punt). Based on 129 314 running plays, 148 566 passing plays, and 62 341 special teams plays, these represented injury rates per 1000 plays of 0.79, 1.55, and 3.27, respectively.
Of the special teams injuries, 106 occurred during 21 633 kickoffs for an injury rate per 1000 plays of 4.90, whereas 85 occurred during 21 783 punts for an injury rate per 1000 plays of 3.90. Among the kickoff and punt coverage teams, the kicking team incurred a higher injury rate than the kick receiving teams, with the kickoff team incurring 53 of 65 injuries (81.5%) and the punt coverage team incurring 41 of the 54 injuries (75.9%).
Discussion
The most striking finding in this 10-year study is that a disproportionately higher number of hamstring muscle strains occurred during the preseason time period. Our data showed that more than half (53.1%) of all hamstring injuries occurred in the 7-week preseason, before the teams had even played their first regular-season game. These data are striking when compared with the 16-week regular season, in which only 45% of the injuries occurred, and the postseason, in which only 1.1% of injuries occurred. If we consider only primary hamstring strains, we find that an impressive 57.3% occur in this short preseason versus 41.4% in the regular season. Despite the high preseason injury rate in our study, this rate is approximately half of that reported by Feeley et al 11 (0.82/1000 A-E vs 1.79/1000 A-E for practice and 2.47/1000 A-E vs 4.07/1000 A-E for games) in their 10-year report on preseason injuries in 1 team.
Almost 4 of 5 (78.9%) practice injuries occurred in the preseason, with more than 70% of those occurring in July, the first month of football participation. Furthermore, the injury rate during practice is 449% higher, 0.82 per 1000 A-E versus 0.18 per 1000 A-E, in the preseason as compared with the regular season. The first month of NFL preseason games, August, is also the month with the highest incidence of game injuries, 192 or 23.7% of the total. These high preseason injury incidences are devastating not only for their immediate impact but also because primary hamstring strains are associated with decreased performance upon return to competition 39 and have a high risk of more severe 4 reinjury during the competitive season. 12,15,17
The data presented here confirm the findings of Feeley et al 11 that the preseason is a period during which NFL athletes are at particular risk for injury. These findings are also consistent with those of Whiteside et al, 42 who found that the frequency of hamstring muscle strains was heavily weighted toward the spring preseason preparation. We believe that a number of factors contribute to this high hamstring strain incidence during the preseason, but most revolve around the relative deconditioning that occurs in the off-season. Muscle weakness has been implicated as a predisposing factor for both primary 7,26 and recurring 6 hamstring strain injuries. Studies have also cited the role of fatigue in causing muscle strains, 20,43,45 suggesting that fatigued muscles are more vulnerable to injury.
Protection against muscle strain is provided not only by strong muscles but also by the appropriate timing and magnitude of neural control. Activated muscle complexes are able to absorb larger eccentric forces, 13 and high-intensity sprinting involves intricate neuromuscular coordination in the hamstrings. 37 Recent biomechanical studies of sprint mechanics suggest that the posterior thigh is most susceptible to such strains near the end of the swing phase, when the hamstrings reach maximal length and undergo eccentric contraction just prior to heel strike. 16,33,44 It has also been shown that fatigue alters neuromuscular activity in the hamstrings during sprinting. 28 Although Pinniger et al 28 hypothesized that hamstring fatigue-related changes in sprint mechanics may serve as a protective mechanism to reduce stress on the hamstrings during sprinting, changes such as decreased hip and knee flexion could also be associated with a temporary loss of athleticism, decreasing the ability of the athlete to safely and efficiently change direction or accelerate. Fatigue has been shown to significantly delay activation of the lateral hamstrings during a landing task, thereby compromising knee control strategies. 14 It is possible that fatigue-induced dysynergistic neuromuscular activation, resulting in suboptimal absorption of eccentric forces, 13,20 may be responsible for many hamstring strain injuries in sports that involve not only high-velocity sprinting but also multidirectional acceleration. Although the role of these complex sports-specific neuromuscular firing patterns that occur during athletic activities has not been fully elucidated, we believe that following the off-season, sports-specific neural deconditioning, relative muscle weakness, and fatigability make some athletes particularly vulnerable to hamstring strain in the NFL preseason.
This study also showed that despite the ballistic, contact nature of professional football, the majority of hamstring strains (81.5%) were sustained by players during noncontact activities, with 71.0% citing sprinting as the primary activity at the time of injury. Although most previous attempts to describe the injury setting in football dealt with all types of injuries, our findings are in stark contrast to those earlier reports, which found that most injuries in football are due to contact. 8,24 At the college level, contact has been found to account for 77.9% of fall game, 56.5% of fall practice, and 69.2% of spring practice injuries. 10 A recent report on high school football injuries in California found that more than 78% of all injuries occurred during player contact. 30 By contrast, within player positions most frequently injured in our study, the defensive secondary and wide receivers, 92.0% and 93.5% of injuries were noncontact. Because a wide majority of the injuries we studied occurred during sprinting, independent of the contact element frequently present in football injuries, our findings and injury prevention strategies may be applicable to other less contact-intense sports with similar activity patterns.
Hamstring muscle strains were distributed with striking inequity among players who played different positions. Wide receivers had significantly more hamstring strains than did offensive linemen (n = 355 vs n = 89, respectively) and injury rates per team-season (1.23 vs 0.31). Thus, despite the fact that the standard formation in football uses 5 offensive linemen and only 2 wide receivers, those wide receivers sustain 403% more injuries annually than the offensive linemen. This difference was even more striking in view of previous studies that reported offensive lineman to be among the most frequently injured players and wide receivers among the lowest. 8,30,38
We found that all of the “speed positions” (receivers, defensive backs, and running backs) had high injury rates for hamstring muscle strains compared with the “strength positions” (offensive and defensive linemen). It is likely that this disparity is primarily because of the increased exposure to the injurious activity—namely, sprinting—by receivers and other members of the “speed positions.” However, we cannot conclude that this is solely the case because perhaps some intrinsic physical quality—for instance, absolute hamstring muscle strength—may be protective in these generally larger strength position athletes. This possibility becomes apparent when we consider that all football players do a substantial amount of sprinting, especially during early season practices and conditioning, and that many of the “strength position” athletes are capable of high-velocity sprinting, thus exposing them to a similar environment in which the speed position players incurred most of their injuries. It is hoped that future studies will continue to delineate the intrinsic and extrinsic factors that are responsible for this considerable variability in muscle injury rates.
Although passing plays resulted in almost twice as many hamstring muscle strains per thousand plays as running plays (1.55/1000 A-E vs 0.79/1000 A-E, relative risk [RR] = 1.96), the risk of being injured during a special teams play was considerably higher than either (3.27/1000 A-E), with an RR of 2.11 and 4.14 as compared with the passing and rushing plays. Within the special teams data, we found the highest injury rates for the kickoff and punt plays, with injury rates of 4.9 and 3.9 injuries per 1000 plays. Of these, the vast majority occurred in the kick coverage teams, with the coverage teams sustaining 81.5% and 75.9% of the injuries. It is not unexpected to have high injury rates in the kick coverage teams, as most players in this setting are exposed to high levels of full-speed sprinting. Special teams plays also frequently use players who are not regularly involved in offensive or defensive plays and thus are only intermittently involved in game activity. Proper warm-up is critical in injury prevention because an increase in muscle temperature results in increased muscle elasticity. 32 Irregular participation allows muscles to cool down, which decreases muscle elasticity and ability to absorb energy. In lieu of rule changes, we recommend greater attention to warm-up of kick coverage teams before entering the game. This can take the form of progressively more intense sideline sprinting or progressive warm-up on a stationary exercise machine (eg, cycle ergometer, elliptical trainer) that uses the hamstring muscle group adequately.
It has been reported that muscle strain reinjury occurs at a high prevalence in the hamstring group. 12,15 Reinjuries accounted for 16.5% of total injuries in our study, with almost one-fifth (19.7%) of the regular-season injuries being reinjuries. Thus, primary injuries in the preseason appear to account for a large number of the injuries sustained during the regular season. Comparatively, these reinjury rates are similar to those reported at the elite level in other sports. Reinjury rates have been reported at 12% in soccer, 43 23% in rugby, 4 and as high as 34% in Australian Rules Football. 27
On the basis of these findings, we strongly recommend an increased effort by athletes to enter the NFL preseason in a high state of sport-specific conditioning, particularly with regard to strengthening and full-velocity sprinting. An injury prevention program in Australian Rules Football had success by integrating sport-specific high-intensity interval running and acceleration drills into preseason conditioning programs. 40 Also, movement-specific neural training with an emphasis on sprint technique has been shown to enhance neuromuscular control in elite athletes. 5 As activated muscles are able to absorb greater amounts of force, 13 increased neural control should lead to an increase in absorbed energy without tearing. By exposing an athlete to progressively intense bouts of sprinting and sport-specific neuromuscular training during the off-season and early in the preseason, injury risk may be mitigated. Neural conditioning may be even more important than previously thought in preventing reinjuries. A study by Sherry and Best 35 demonstrated that a program of progressive agility drills and trunk stabilization proved to be significantly more effective at protecting an athlete from reinjury when compared with a program focused on static stretching and isolated resistance training. Such preventative measures are especially important for those groups determined to be at high risk for injury, specifically the “speed positions” (ie, wide receivers and defensive secondary) and special teams players.
Evidence suggests conditioning should include not only concentric strength training for the hamstring muscle group but also an emphasis on eccentric loading. 1,22 Eccentric exercise has been shown to produce greater strength gains than similar concentric hamstring movements. 19 Also, eccentric hamstring movements were shown to result in an increase in the optimum angle for torque generation after just 1 session. 3 Such a change may make the hamstrings stronger in a more functional position during sprinting and increase the ability of the hamstrings to absorb repeated eccentric loads prior to and during heel strike. The authors successfully reduced muscle injuries by an order of magnitude in an NFL team by implementing a program focused on high-force eccentric loading of the hamstrings, as well as identifying and correcting muscle weakness and progressive sprint training for identified high-risk athletes.
The athlete’s hamstring muscle group needs to be not only strong with appropriate neural timing but also fatigue resistant because fatigued muscles appear vulnerable to strain injury. 20 Hamstring fatigue correlates with decreased sprint speed and an alteration in sprint mechanics. 28 Also, hamstring fatigue has been associated with increased anterior tibial translation and a mechanical loss of knee stability, 21 thus compounding the risk of strain injuries with risk of traumatic injuries to the knee.
We have provided a number of foci for preventive intervention in this large descriptive study of hamstring injuries in football. Coupled with recent advances in our understanding of muscle function and injury mechanism, we hope this will support the continued development of hamstring injury prevention programs in sport. It is important to note that the benefits of a preseason training program with a focus on preventing hamstring strains go above and beyond injury prevention. Many of the factors associated with reducing the risk of hamstring strain injuries are also closely associated with overall athletic performance. 2,13 Thus, in addition to reducing the risk of hamstring strains, the principal side effect of a well-orchestrated muscle injury prevention program is generally faster, stronger, and more athletic athletes.
Footnotes
One or more of the authors has declared the following potential conflict of interest or source of funding: A grant from NFL Charities was used to support this study.