Journal Logo

CLINICAL SCIENCES: Clinical Investigations

Concussions in hockey: there is cause for concern


Author Information
Medicine & Science in Sports & Exercise: December 2001 - Volume 33 - Issue 12 - p 2004-2009
  • Free


Ice hockey is a sport unlike any other. Players move at a high rate of speed estimated to be approximately 30 mph when skating with sliding speeds of up to 15 mph in senior amateur players (15). What separates hockey from other contact sports is that players traveling at high rates of speed make contact with numerous hard surfaces and objects such as the boards, glass, ice, goal posts, and other players. The situation where speed and rigid obstacles are combined increases the likelihood of acceleration and deceleration of the player’s body, especially the head. Since Denny-Brown and Russell’s (5) seminal work on the subject, it has been known that acceleration/deceleration forces are required for concussion to occur. As such, it is not surprising that the most common athletic injury is concussion and that hockey rates among the highest of all contact sports for concussion rates per player exposure (4). It has also been recognized that although concussion constitutes a considerable medical problem in hockey, it often time receives the least medical attention (8).

An increased understanding about concussion in amateur hockey is required because many of these players will continue to play for several years, and recent evidence suggests that the cumulative effects of multiple concussions can be demonstrated experimentally in young athletes (5,7,9). In the fall of 1998, a 2-yr study was initiated with the British Columbia Junior Hockey League (BCHL) that has addressed the problem of concussion in Canadian amateur hockey. Detailed retrospective and prospective concussion histories were collected over two BCHL seasons (1998–1999 and 1999–2000). The purpose of this paper is to document various aspects of concussion in Canadian Amateur hockey including demographics, causes, treatment, and prevention. Increased awareness as to the nature and causes of concussion in hockey is important as it will have a direct bearing on recommendations intended to protect participants from unnecessary injury. In addition, further information regarding prevention and treatment of these injuries will help reduce injuries and shape treatment strategies to ensure that injury rates of concussion and their subsequent effects do not increase.


The study involved 272 players in year 1 of the study, and 283 players in year 2, 115 of which were players from the previous year. All 14 teams in the league participated fully in the study. All participants were male and ranged from 15 to 20 yr of age. The regular season, not including playoffs lasted approximately 6 months. Practices averaged 1.5 h·d−1 and occurred approximately 4 d·wk−1. Each team played 54 regular season games. The players wore standard hockey pads during games and practices, including helmets with league-mandated half visors. League-wide mouth-guard use was mandated in 1998, in part to protect against concussion, though there was no monitoring of use.

The data collection for both seasons was divided into two sections: baseline (retrospective) and follow-up (prospective). Baseline data were collected in the community of origin for each team by using a mobile laboratory. The baseline collection period took approximately 5 wk to complete. A laminated card provided by the British Columbia Brain Injury Association with a grading system for concussion as well as return to play recommendations were distributed to each team during the baseline in an attempt to standardize referrals. The three-point grading system used was based on those approved by the Quality Standards Subcommittee of the American Academy of Neurology (13) (1997).

After the baseline collection, the follow-up phase of the projects commenced. If a player sustained a concussion at some point during the season, notification of the injury was provided by a team trainer, coach, or physician, and a follow-up commenced within 24 h. All testing was repeated, and details of the most recent concussion were obtained. Players with concussion ranging from grade 1 to grade 3 were included in the study. Prospective concussion data were verified in all cases by the team trainer. The detailed prospective and retrospective concussion history allowed for an analysis of various aspects of how and why concussion occurred over the Midget and Junior careers of these individuals.

Incidence of injury was calculated using an athlete exposure rate that was referenced to player-game or player-practice hours. Incidence of injury was calculated by assuming that six players were on the ice during game situations and 21 players were on the ice during practice. The number of players on the ice [6] was multiplied by the number of game hours [54] by the number of teams [14], and this resulted in a total of 4536 player-game hours. Player-practice hours were calculated by multiplying the number of players on the ice [21], by the actual noninstructional practice time per day (1.25 h), by the number of practices per week [4], by the number of weeks in the 6-month season [32], for a total of 3360 player-practice hours. This is likely an overestimate, as practices were cancelled on occasion due to travel demands.

Only hockey-related concussions were analyzed. For questions regarding the retrospective history, participants from the 1998–1999 and 1999–2000 season were treated as separate cohorts; therefore, there is overlap between seasons for 115 players who were in the league both years. Retrospective information included the player’s entire concussion history, although most concussions reported occurred during Midget (15–16) and Junior (16–20) years. Analysis of factors related to each concussion was performed using descriptive statistics expressed as injury rates and percentages, with summary statistic reported.



Of the 379 concussions reported, 341 (90%) occurred during games and 30 (7.9%) during practice with 8 subjects (2.1%) reporting other. The position of the players translated into a greater number of concussions for forwards and fewer concussions for goaltenders. Figure 1 shows the number of concussions per position for the 1998–1999 and 1999–2000 cohorts. Because there was six players on the ice per team at any given time, the three forwards were considered to account for 50% of players, defensemen accounted for 33%, and goaltenders 16.6%. As shown in Figure 1, the proportion of concussion for forwards is slightly higher than would be expected for 1998–1999 (58.1%) and 1999–2000 (56.1%) seasons. The percentage of concussions for defensemen is approximately equal to their on ice representation for both seasons (34.7% and 32.6%, respectively). The number of concussions experienced by goaltenders was far below their on-ice representation (4.8% and 6.1%, respectively). Table 1 shows the age at which players experienced their first, second, and third concussion. As shown in the table, the mean age for the first concussion is in the 15th year, the second concussion in the 16th year, and the third concussion in the 17th year.

Percentages of concussion by position.
Table 1
Table 1:
Mean and standard deviation (SD) for the age of first, second, and third concussion.

Figure 2 shows the percentage of players who experienced a concussion versus those who did not at some time during their amateur careers. Concussed players were divided into groups based on whether their concussion resulted in loss of consciousness (LOC) and no LOC. If more than one concussion was reported players could check both items. Consequently, there is a separate category for players who experienced more than one concussion, one with LOC and one without LOC (other available choices included “do not remember” and “other”). Of all concussed players from both years of the study, 32.5% (1998–1999) and 43.5% (1999–2000) reported LOC of less than 1 min, 8.4% and 5.8%, respectively, for 1–5 min, and 3.2% and 0.6%, respectively, for 5–30 min.

Percentages of individuals never concussed and concussed with and without LOC.

For the 1998–1999 season, 29 of 272 players were referred to us for follow-up constituting an annual concussion rate of 10.66% for the entire league. Because 27 players were injured during games, the total of concussions per 1000 player-game hours was 5.95. The number of players injured during practice was 2, resulting in a player-practice injury rate of 0.6 per 1000 h. For the 1999–2000 season, the number of players referred for follow-up was 21/283 making the annual rate of concussion for this season 7.42%. Because all 21 players were injured during games, the total of concussions per 1000 player-game hours was 4.63 with no injuries reported during practice. However, because some of the players from the first season also were members of the league for the second season, data were available regarding how many players had a concussion during the first season compared to how many concussions we were notified about. Of the 20 players that had hockey-related concussions during the 1998–1999 season that filled out the retrospective concussion history form again for the 1999–2000 season, only 10 (50%) were referred to us for follow-up. Therefore, it appears that there was up to 50% underreporting of concussion. This would elevate the annual rate of concussion in the BCHL to as much as 20% per year.


Figure 3 shows the causes of all concussion reported for both the 1998–1999 group and the 1999–2000 group. The greatest reported cause of concussion for both groups was the head striking the boards including the end glass. This was followed by elbow, shoulder, or ice contact. Being hit with a stick or puck resulted in relatively few concussions, as did fighting at 7.3% for the 1998–1999 group and 4.5% for the 1999–2000 group. For players who responded to the 1999–2000 questionnaire, 22.1% of concussions resulted in a minor penalty, 13.7% in a major penalty, with 54.2% resulting in no penalty at all.

Causes of concussion.


During the 1998–1999 season, 91.9% (113/123) of subjects reported wearing a helmet at the time they sustained the concussion. In the 1999–2000 season, 94.8% (127/134) of subjects reported wearing a helmet at the time they sustained the concussion. Because helmet use was mandatory for the duration of each player’s amateur career, it was assumed that those who sustained concussions without helmets were not wearing them properly or were not wearing them during practice. For the 1998–1999 season, subjects were asked whether their helmet was equipped with a full-face or half visor. Of 113 players with concussion, 33.6% (N = 38) were wearing a half visor, 57.5% (N = 65) were wearing a full visor, and 8.8% (N = 10) reported wearing no facial protection. For the 1999–2000 season, 41.4% (55/133) of players reported wearing a mouth-guard at the time of the incident.


Immediately after the injury, 37.1% of concussed players were taken to the hospital after concussion in the 1998–1999 group and 44.4% for the 1999–2000 group. In the later group, 32.3% reported that a team physician was present during the game they were injured in, 63.2% said no physician was present, and 3.8% did not know. Before returning to play, 16.4% of the 1998–1999 group stated that they were tested with exertional activity before returning to play, and this increased to 20.6% for the 1999–2000 group.


Ice hockey is a sport where players move at high rates of speed estimated to be approximately 30 mph when skating with sliding speeds of up to 15 mph in senior amateur players (15). High speed, coupled with the fact that rigid obstacles exist to confine the playing area, combine to make the likelihood of acceleration/deceleration injury high in these athletes. Rates of concussion in hockey per player exposure are among the highest for all contact sports (4). It has also been recognized that although concussion constitutes a considerable medical problem in hockey, it often receives the least medical attention (8). We, therefore, performed a combined retrospective and prospective analysis of various factors related to concussion including demographics, causes, treatment, and prevention with the intent of shaping future decisions aimed at reducing concussion rates of incidence and severity in hockey.

Consistent with numerous studies documenting all ice hockey injuries, concussion occurred most often during games (90%) compared with practice (7.9%). The proportion of players concussed during games in the present study was similar to rate for all injuries in the National Hockey League (NHL) between 1984 and 1990 (86–89%) (3). The rate of concussion that occurred in practice for the present study was substantially lower than the 19.2% occurring in practice games and practices reported by Tegnor and Lorentzon (19). However, for the Tegnor and Lorentzon data, when practices alone are considered, the rates of concussion drop to 6% of all concussions and are similar to those observed in this study. This pattern of results suggested that factors related to game situations such as forcible body contact may lead to higher concussion rates, whereas respect for a team mates’ well-being over an opposing player’s may lead to lower rates.

The position of the player appeared to be an important factor regarding concussion susceptibility. Forwards reported a higher proportion of concussions (56–58%) compared with their proportional representation on the team (50%). Defensemen, on the other hand, sustained a number of concussions (32–34%) consistent with their proportional representation on the team (33%). The number of concussions experienced by goaltenders (4–6%) was far below their on-ice representation (16%). This pattern of results is consistent with studies documenting all types of ice hockey injuries. Most of these studies suggest that forwards are the most vulnerable players, whereas goaltenders are least likely to be injured (3,12,16). However, not all studies are consistent with this pattern of injury susceptibility. Equal rates of injury susceptibility for forwards and defensemen (8) and greater rates of injury for defensemen compared with forwards (11) have been reported. Concussion rates for goaltenders may be far higher in other countries as a Finnish study documented. Of nine concussions sustained in the Finnish National League during a season, five occurred in goaltenders (10). The primary difference between the studies indicating susceptibility of forwards versus other positions is that the former were performed using North American players. As such, there may be an aspect of the North American game that leads to greater vulnerability of forwards to injury. These may include the reliance on a “dump and chase” system of play, “cycling the puck” near the end boards with players that are smaller in stature, and differences in the icing rule (“no-touch” vs touch).

The annual rates of concussion for the league were 10.66% for the 1998–1999 season and 7.42% for the 1999–2000 season with 5.95 and 4.63 concussions per 1000 player game hours, respectively. These figures most likely represent the lower bound for concussions rates in the BCHL because there was evidence of underreporting of concussions based on the retrospective histories of players participating in both years of the study. The results for concussions per player-game hours were very similar to those in the Swedish Elite League (SEL). Tegnor and Lorentzon (19) reported that the risk of sustaining a concussion during SEL play was calculated at 6.5 per 1000 player game hours or approximately one concussion per team. In addition, Cantu (4) reviewed a report suggesting that the rate of concussion per 1000 athlete exposures is 0.27 for hockey which is slightly higher than football at 0.25/1000 athlete exposures. Concussion has been reported to account for between 2 and 14% of all hockey-related injuries (3,11,14,18,19) and 15 and 30% of all hockey-related head injuries (3,15). Therefore, the current available information from this study and others suggest that concussion has been, and continues to be, a substantial problem in the sport of ice hockey.

Novel demographic information obtained in this study was that the mean age of the first concussion was in the 15th year, the second concussion in the 16th year, and the third in the 17th year. This is also the range where most spinal injuries occur in Canadian hockey (17). Therefore, in Canada, the first concussion occurs on average during 1st year of Midget hockey in the majority of those who have had at least one concussion. However, the standard deviation for the first concussion was high compared with the second and third concussions (see Table 1). As such, the age when the first hockey-related concussion was sustained was in many cases much younger than 15 yr of age. The age when most players experience their first concussion is important information for coaches and trainers who work with youth in this age group. Proper assessment and treatment of the first concussion may allow for more complete recovery because a player who returns to play before the succession of acute postconcussive symptoms is at risk for further injury (6). Although this has not been demonstrated experimentally, a player who experiences dizziness, attention deficits, and slowed reaction times is at greater risk of further injury because balance, the ability to attend and anticipate play development, and the ability to react appropriately are necessary to avoid potentially dangerous body contact. As such, it is very important to ensure that a player is asymptomatic, at rest and during physical exertion, before returning to a full contact practice.

Finally, it appears that most concussions were not severe although a large proportion of players had sustained at least one concussion. Approximately 60% players reported having sustained at least one concussion, which is considerably higher than the 22% who in the Swedish Elite League sustained a concussion at some point in their entire career (19). Approximately 85–90% of all concussions either did not result in LOC or resulted in LOC less than 1 min in duration. Taken together, this pattern suggests a high proportion of players sustain at least one concussion that does not cause prolonged LOC. This may be good and bad news; the bad news is that the number of players who sustain a concussion is quite high and that the possibility of limited brain damage exists with every concussion. The good news is that improved rule enforcement and equipment may be protecting players against more severe injuries. For instance, time lost from practices and games as a result of concussion was significantly longer for players wearing partial, compared with full, facial protection (2).

The fact that hockey is a game played at a high rate of speed with many obstacles to mediate on ice is reflected in the causes of concussion in this study. An overwhelming number of head injuries occurred due to collisions with either the boards and/or the ice, accounting for between 40 and 50% of all concussions reported. This pattern of results is consistent with Tegnor and Lorentzon’s (19) study where body checking or boarding was the most common reason for a concussion. Impact with the boards also accounted for 77% of all cervical injuries in Canadian hockey (17).

What was somewhat surprising was the degree to which illegal plays, such as elbowing, were responsible for concussion. Over 20% of all concussions reported for the 1999–2000 cohort were the result of elbow contact to the jaw or another part of the head. The fact that elbowing alone accounted for a high percentage of concussions is consistent with the fact that 45% of concussions that occurred in the 1999–2000 cohort resulted in a major (13.7%) or minor (22.1%) penalty being called. The rates of injury due to illegal play are consistent with a study on all ice hockey injuries sustained by players wearing full-face visors (41.3%) and half visors (32.2%) even though penalties were called only 10% of the time (2). However, most studies reported lower rates of injury due to illegal play for concussion (19) and all ice hockey injuries (9,12,18).

It is interesting to note that fighting accounted for fewer concussions than elbowing alone, and far fewer concussions than either striking the head on the ice or on the boards. As an illegal action, fighting in hockey has been targeted by some as a major cause of injury, including concussion. The results of this study do not support this position. Those interested in concussion prevention through rule changes and enforcement should focus their efforts on actions where a clear intent to injure another player is demonstrated, especially when the injured player is in a vulnerable position. Examples of such plays include elbowing an opposing player in the jaw or head, charging, boarding, “slew foot,” check when vulnerable (check from behind or when off-balance, near boards traveling past net, etc.), and a high stick to the head (cross-checking or a “ two-hander” to the head). In the 1999–2000 NHL season, players who performed these types of illegal plays were suspended more severely than in previous seasons, suggesting that this league is beginning to recognize the importance of eliminating this type of activity from the game. Unfortunately, there continued to be a number of well-publicized incidences where players clearly intended to injure other players, for instance, by swinging the stick at the head of the opponent, all of which resulted in concussions.

Other efforts to reduce the number of concussions include developing better helmets and facial protection, as well as having players wear additional mouth protection such as mouth-guards. Having the helmet secured in a proper manner is of course important, although a large number of players continue to wear the chin-strap in a loose manner. This may have accounted for the small percentage of individuals (5–8%) who reported not wearing a helmet at the time of their concussion. Most players reported wearing a full (57.5%) or half shield (33.6%) at the time of their concussion. Previous studies have reported no significant difference in protection from concussion based on the type of facial protection worn by a player. However, both types significantly reduced the number of lacerations to the face and the full visor may have reduced the severity of the concussion (2). The results of the current study are consistent with these findings in that very few players sustained severe concussions. The continued use of a mouth-guard, especially in players using half visors, may also reduce the proportion and severity of individuals who sustain a concussion. Although the protective benefits of mouth-guard use have not been demonstrated experimentally, use of these devices has been implemented in numerous contact sports for their potential protection against concussion, as well as their obvious protection from dental injury (1).

While it is apparent that concussions in hockey are of considerable concern, there is now encouraging information with respect to the treatment of concussion. Between 37 and 45% of players were assessed in hospital after their injury, and games where concussions occurred were attended by physicians in one third of the cases. This represents a substantial increase in treatment from previous studies that reported involvement of a physician for only 8% of hockey-related concussions (8). This trend is important as it may represent an increased awareness by players, trainers, coaches, physicians, and parents of the proper assessment of concussion, as this partially dictates when an injured player can safely return to play.

The authors would like to thank the British Columbia Hockey League president, Ron Boileau, players, coaches, trainers, and managers. We thank Anthony Brough and Fannie Hansen for the assistance in the data collection and Robert Schutz for his help with the statistical analysis. This research was made possible by a grant from the B.C. Rick Hansen Neurotrauma Foundation. Northern Lite provided equipment support by way of donation of a camper used as a mobile lab. There were no professional relationships with companies or manufacturers who will benefit from the results of the present study. The results of the present study do not constitute endorsement of any product by the authors or ACSM.

Address for correspondence: David Goodman, Ph.D., School of Kinesiology, Simon Fraser University, Burnaby, B.C., Canada, V5A-1S6; E-mail: [email protected]


1. Barth, J. T., J. R. Freeman, and J. T. Winters. Management of sports related concussions. Dent. Clin. North Am. 44: 67–83, 2000.
2. Benson, B. W., N. G. H. Mohtadi, M. S. Rose, and W. H. Meeuwisse. Head and neck injuries among ice hockey players wearing full face shields vs half face shields. JAMA 282: 2328–2332, 1999.
3. Biasca, N., H. P. Simmen, A. R. Bartolozzi, and O. Trentz. Review of typical ice hockey injuries: survey of North American NHL and hockey Canada versus European Leagues. Unfallchirurg. 98: 283–288, 1995.
4. Cantu, R. C. Head injuries in sport. Br. J. Sports Med. 30: 289–296, 1996.
5. Denny-Brown, D., and R. Russell. Experimental cerebral concussion. Brain 64: 93–164, 1941.
6. Collins, M. W., S. H. Grindel, M. R. Lovell, et al. Relationship between concussion and neuropsychological performance in college football players. JAMA 282: 964–970, 1999.
7. Gaetz, M., D. Goodman, and H. Weinberg. Electrophysiological evidence for the cumulative effect of concussion. Brain Inj. 14: 1077–1088, 2000.
8. Jorgensen, U., and S. Schmidt-Olsen. The epidemiology of ice hockey injuries. Br. J. Sports Med. 20: 7–9, 1986.
9. Matser, E. J. T., A. G. Kessels, M. D. Lezak, B. D. Jordan, and J. Troost. Neuropsychological impairment in amateur soccer players. JAMA 282: 971–973, 1999.
10. Molsa, J., O. Airaksinen, O. Nasman, and I. Torstila. Ice hockey injuries in Finland: a prospective epidemiological study. Am. J. Sports Med. 25: 495–499, 1997.
11. Pettersson, M., and R. Lorentzon. Ice hockey injuries: a 4-year prospective study of a Swedish elite ice hockey team. Br. J. Sports Med. 27: 251–254, 1993.
12. Pinto, M., J. E. Kuhn, M. L. V. H. Greenfield, and R. J. Hawkins. Prospective analysis of ice hockey injuries at the junior A level over the course of one season. Clin. J. Sports Med. 9: 70–74, 1999.
13. Practice Parameter. The management of concussion in sports (Summary Statement). Report of the Quality Standards Subcommittee. Neurology 48: 581–585, 1997.
14. Ruchinskas, R. A., J. P. Francis, and J. T. Barth. Mild head injury in sports. Appl. Neuropsychol.. 4: 43–49, 1997.
15. Sim, F., W. T. Simonet, L. J. Melton, and T. A. Lehn. Ice hockey injuries. Am. J. Sports Med. 15: S-86–96, 1987.
16. Stuart, M. J., and A. Smith. Injuries in junior A ice hockey: a three-year prospective study. Am. J. Sports Med. 23: 458–461, 1995.
17. Tator, C. H., J. D. Carson, and R Cushman. Hockey injuries of the spine in Canada, 1966–1996. CMAJ 162: 787–788, 2000.
18. Tegnor, Y., and R. Lorentzon. Ice hockey injuries: incidence, nature and causes. Br. J. Sports Med. 25: 87–89, 1991.
19. Tegnor, Y., and R. Lorentzon. Concussion among Swedish elite ice hockey players. Br. J. Sports Med. 30: 251–255, 1996.


© 2001 Lippincott Williams & Wilkins, Inc.