Article Text

Download PDFPDF

Injury rates, types, mechanisms and risk factors in female youth ice hockey
  1. Melissa D Decloe1,
  2. Willem H Meeuwisse2,
  3. Brent E Hagel3,4,
  4. Carolyn A Emery2,3,4
  1. 1Faculty of Kinesiology, University of Calgary, Calgary, Alberta, Canada
  2. 2Sport Injury Prevention Research Centre, Faculty of Kinesiology, University of Calgary, Calgary, Alberta, Canada
  3. 3Department of Pediatrics, Faculty of Medicine, Alberta Children's Hospital, Calgary, Alberta, Canada
  4. 4Community Health Sciences, Faculty of Medicine, University of Calgary, Calgary, Alberta, Canada
  1. Correspondence to Carolyn A Emery, Faculty of Kinesiology, University of Calgary T2N 1N4, Calgary, Alberta, Canada; caemery{at}


Background The objectives of this cohort study were to examine the rate, types, mechanisms and risk factors for injury in female youth (ages 9–17) ice hockey players in the Girls Hockey Calgary Association.

Methods The main outcome was ice hockey injury, defined as any injury occurring during the 2008/2009 season that required medical attention, and/or removal from a session and/or missing a subsequent session. Potential risk factors included age group, level of play, previous injury, ice hockey experience, physical activity level, weight, height, position of play and menarche. Incidence rate ratios (IRR) were estimated with Poisson Regression adjusted for cluster (team). Exposure data were collected for every session for each participating player.

Results Twenty-eight teams (n=324) from Atom (ages 9–10), PeeWee (11–12), Bantam (13–14) and Midget (15–17) participated with 53 reported injuries. The overall injury rate was 1.9 injuries/1000 player-hours (95% CI 1.4 to 2.7). Previous injury (IRR=2.7, 95% CI 1.7 to 4.3), games (IRR=2.1, 95% CI 1.1 to 4.2), menarche (PeeWee) (IRR=4.1, 95% CI 1.0 to 16.8) were significant risk factors. In Midget, the more elite divisions were associated with a lower injury risk (A-IRR=0.2, 95% CI 0.1 to 0.5) (AAA-IRR=0.5, 95% CI 0.2 to 0.9).

Conclusions Injury rates were lower in this study than previously found in male youth and women's ice hockey populations. Previous injury and game play as risk factors are consistent with the literature. Menarche as a risk factor is a new finding in this study. This research will inform future studies of the development of injury prevention strategies in this population.

  • Ice hockey
  • Children's injuries
  • Epidemiology
  • Injury Prevention
  • Concussion

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


In the 2007/2008 ice hockey season, 77 461 female players were registered with Hockey Canada.1 Women represent 15% of all registered players in Canada.1 There has been a 900% increase in female ice hockey registrants in the past 15 years.1 It is acknowledged that ice hockey participation has many social and physical benefits2 ,3; however, ice hockey is associated with a high rate of injury.4 Youth ice hockey injury has received substantial attention in the literature, but primarily in male populations.5–10

Few studies have been conducted examining women ice hockey injury. Dryden et al11 examined women's (ages ≥14) recreational ice hockey injuries in Canada (n=314) and reported an injury rate of 39.8 injuries/100 players/season.11 Schick and Meeuwisse12 found that injury rates (IR) were similar in men and women varsity ice hockey (relative risk (RR)=1.1, 95% CI 0.8 to 1.5).

Most studies examining youth ice hockey injuries have focused on male participants, with IR ranging from 11.7 to 34.4 injuries/1000 player-hours.10 Risk factors previously identified in male youth ice hockey include: policy allowing body checking, higher divisions of play, older age group, lower body weight, fatigue, higher levels of aggression and lower levels of empathy.5–10 It is unclear whether IR and risk factors can be generalised to female youth ice hockey.11 As a result, the objectives of this study were to examine the rate, type, severity, mechanisms and risk factors for injury in female youth (ages 9–17) ice hockey players.



The study sample included female youth ice hockey players (ages 9–17) from the Girls Hockey Calgary Association (GHCA) in the 2008/2009 season. Players were recruited preseason from all levels of play in Atom (ages 9–10), PeeWee (11–12), Bantam (13–14) and Midget (15–17) age groups. (Note: in the GHCA in all age groups, division B is the lowest level of play and A the highest. AAA is offered in Midget only and is the most elite level. Midget is also the only age group that spans 3 years of eligibility.) Each player and their parent/guardian were asked to provide written consent, as per the Office of Medical Bioethics, University of Calgary.

Players were included if they (1) were registered with GHCA in Atom, PeeWee, Bantam or Midget, (2) were participating fully at the beginning of the 2008–2009 ice hockey season and (3) signed a consent form along with their parent/guardian. Teams were included if (1) the coach consented to participate and (2) a team designate was identified. Players were excluded if they had an injury or chronic disease that prevented full participation at the start of the 2008–2009 ice hockey season.


A previously validated injury surveillance system was used to measure risk factors and injury among female youth ice hockey players.5 The independent variables included: age group, division of play, session type (games vs practices), relative age (first or second year of age group) and the following self-reported variables: player position, height, weight, menstrual history, previous ice hockey experience (years), previous concussion, previous injury and physical activity level. All variables are operationalised table 1. The Preseason questionnaire was completed at baseline by the players themselves unless assistance was required from their parent/guardian.

Table 1

Risk factors for injury (adjusted for cluster)

Each team was assigned a team therapist (physiotherapist, athletic therapist or athletic therapist candidate) who attended one game or practice session every 2 weeks to assess all ice hockey-related injuries. Therapists attended a 2 h training session with the study coordinator prior to the season, at which time the injury definition and data collection materials were explained in depth. If injuries occurred prior to the next scheduled 2 week visit, the team designate contacted the therapist and made an effort to visit the team sooner such that all injuries would be assessed within 1 week. Team designates (coach or parent) were trained during the first session by the team therapist to record individual exposure data (player attendance) on the Weekly Exposure Sheet (WES). The team designate recorded on the WES whether each player participated fully (100%), partially (<75%) or not at all (0%) in each game and practice, as well as the length of each session (in hours). We recorded individual exposure for each player in hours. Of note, players participating partially in the session were given credit for 50% of the session (ie, 30 min of an hour-long session) towards their total exposure. The other role of the team designate was to liaise with the team therapist to ensure that they were aware of all injuries that required assessment and that all study forms were completed and collected.

Outcome measures

The primary outcome was ice hockey injury defined as any injury occurring in ice hockey in the 2008/2009 season that required medical attention (beyond that of the team therapist), removal from a session and/or resulted in a player missing a subsequent session.5 Medical attention was recommended when appropriate by the study therapist. Injury Report Forms (IRF) were used to collect information regarding the time and date of the injury, mechanism of injury, injury type, severity and other injury-related details. Injuries were described by type, mechanism and severity based on time loss from hockey. Injuries were categorised as slight (0–1 day), minimal (2–3 days), mild (4–7 days), moderate (8–28 days) and severe (>28 days) based on consensus guidelines.13 The completion of the administrative components of the IRF was initiated by the team designate or the team therapist at the time of the injury. The team therapist assessed all injuries and completed the injury-specific portions of the IRF prior to return to play. Injured players were referred to a study sport medicine physician if the injury resulted in time loss >1 week from ice hockey or immediately if a concussion was suspected. This physician (or any other healthcare practitioner who assessed the injured player) was asked to complete a diagnosis and treatment plan form. Telephone follow-up to complete any missing information was done by the research coordinator.


We report IR (number of injuries/1000 player-hours) with 95% CI and incidence proportion (IP) (number of injuries/100 players). To examine the risk factors for injury, univariate Poisson regression analyses, adjusted for cluster by team and offset for exposure hours, were used to estimate incidence rate ratios (IRRs; 95% CI). This model was selected for its appropriateness for count data and because it allows for multiple outcomes by a single participant.14 All analyses were exploratory.


Twenty-eight of the 33 teams approached (85%) agreed to participate. There were four Atom teams, six PeeWee teams (three Division A, three Division B), nine Bantam teams (five Division A, four Division B) and nine Midget teams (four Division A, two Division B and three AAA). In total, 324/476 (68.1%) potential study participants consented to participate. Baseline characteristics are presented in table 2.

Table 2

Baseline characteristics of the study sample

Injury rates

There were 53 injuries reported during the study period. Four players had two injuries and two players had three injuries. The overall IR was 1.9 injuries/1000 player-hours (95% CI 1.5 to 2.7) and IP was 16.3 injuries/100 players (95% CI 10.8 to 24.9). Incidence rates by age group and division of play are summarised in table 3. There was no statistically significant increase in risk of injury by age group; however, point estimates could suggest an increased risk in Midget (IRR=2.7, 95% CI 0.9 to 8.0). There was also no statistically significant increase in risk by division of play; however, the point estimates in Bantam B also suggest an increase in risk (IRR=5.6, 95% CI 0.9 to 33.4). In Midget, players in A (IRR=0.2, 95% CI 0.1 to 0.5) and AAA (IRR=0.5, 95% CI 0.2 to 0.9) had a decreased risk of injury compared with players in division B (the lowest level of play).

Table 3

Incidence rates by age group and division of play (adjusted for cluster)

Mechanism/type and severity of injury

Body contact with other players (intentional and incidental) was attributed to 58% of all injuries, a trend consistent across the age groups. Body checking (a subset of intentional contact) was attributed to 20.7% of all injuries. No contact was reported in 20% of injuries, while contact with the environment was reported in 15% of injures. The remaining injuries were recorded with unknown mechanisms. Penalties were reported for 7/53 injuries (13.2%). With respect to severity, 44/53 (83%) of the injuries resulted in at least 1 day of time loss and 15/53 (28.3%) resulted in more than 1 week of time loss. Muscle strain was the most commonly reported type of injury (28.3%), followed by contusion (16.9%), ligament sprain (16.9%), concussion (15.1%), joint swelling (9.4%), dislocation (5.7%) and fracture (3.8%). Eight concussions were reported in six players. Only 3/8 (37.5%) concussions resulted in time loss greater than 1 week.

Other risk factors for injury

The additional risk factors examined are summarised in table 1. There was an increased risk of injury in games versus practices (IRR=2.1, 95% CI 1.1 to 4.2). At the PeeWee level only, players who had begun menstruating at preseason had a fourfold increased risk of incurring an injury than those who had not (IRR=4.1, 95% CI 1.0 to 16.8). Previous injury was a risk factor for injury (IRR=2.7, 95% CI 1.7 to 4.3; table 1).


This is the first cohort study to examine IR, mechanisms, types, severity and risk factors for injury in female youth ice hockey. This study contributes to the ice hockey injury literature as female youth ice hockey IR and risk factors were previously unknown.

Our first objective was to examine the rate of injury in female youth ice hockey players. The overall injury rate in this study (IR=1.9 injuries/1000 athlete exposure-hours, 95% CI 1.4 to 2.7) was lower than the rates reported in women's ice hockey (IR=2.5 to 12.6 injuries/1000 athlete-exposures).15 IR in this study were also lower than what was found in male youth ice hockey in the same age groups using similar injury surveillance (IR=4.1 injuries/1000 exposure-hours, 95% CI 3.6 to 4.6).5 While no other studies to date have examined female youth ice hockey players specifically, Dryden et al11 included Midget players in their sample and reported an injury rate of 6.7 injuries/1000 player-hours in Midget players, which is higher than the rate we report (IR=2.8, 95% CI 1.8 to 4.3) in Midget players in our study. There are many possible reasons why the rates in female youth ice hockey are lower than in boys’ or women's ice hockey. In boys’ ice hockey, body checking is permitted at most levels of play considered in this study population and body checking is associated with a twofold to fourfold increased risk of injury.6 ,10 ,16–19 Adult female players may have a higher rate of injury due to their likelihood of having a previous injury, which is an established risk factor for injury.20 ,21 With respect to differing IR in Midget, it must be noted that Dryden et al11 utilised self-reported injuries, compared with the therapist reporting in our study. We can hypothesise that the different methodology contributed to the twofold higher injury rate in the Dryden et al11 study due to the over-reporting of injuries by participants, or possible under-reporting in our study by the therapist as an intermediary. However, it is not known how the injury assessment methodology affects results, making comparisons between these studies crude.

Concussion rates in varsity women's ice hockey have been estimated to range from 1.8/1000 player-hours in practices to 3.6/1000 player-hours in games.15 These rates are considerably higher than the concussion rate found in this study (IR=0.2, 95% CI 0.1 to 0.6). However, speed of play, power and aggression, and a previous history of concussion in women's varsity athletes versus youth ice hockey players all make this comparison unreasonable. On the other hand, concussion rates in male youth ice hockey players from a study with similar injury surveillance methods were consistent with this study. They reported (IR=0.2 concussions/1000 player-hours, 95% CI  0.0 to 0.7) in Atom (where there is nobody checking) and (IR=0.9 concussions/1000 player-hours, 95% CI  0.6 to 1.6) in Bantam (the highest rate of concussion in that study).5

Our second objective was to examine the mechanisms and types of injury in female youth ice hockey players. The mechanisms reported in this study are consistent with the literature in women's and male youth ice hockey. Body checking is not permitted in female youth ice hockey; yet, it was reported as the mechanism of injury for 20.7% of the injuries. This is consistent with the study by Dryden et al11 who found body checking to be the most common mechanism of injury in women's ice hockey, accounting for 21.6% of injuries. This finding has important practical implications as young female ice hockey players and their coaches should be aware that, despite being illegal, body checking does occur and emphasis should be placed on teaching young players how to safely receive a body check as a way to prevent injury.

The injury types reported in this study are consistent with those reported in female adult ice hockey players; however, the distribution of injury types differed.12 Schick and Meeuwisse12 found concussion to be the most common injury type (25%), while in our study muscle strains (28.3%), contusions (16.9%) and ligament sprains (16.9%) were the most common and concussions accounted for 15.1% of the injuries. Concussions were also the most common injury type in a similar study conducted among male youth ice hockey players.5

Our third objective was to examine the severity of injury. The majority of the injuries reported in this study (74%) resulted in less than 1 week of time loss, consistent with the literature in women's and male youth ice hockey.11 ,12 ,15 In Bantam, however, there were a higher proportion of injuries resulting in greater than 1 week of time loss (71.4%) than in other age groups. Emery et al5 also found slightly more injuries that resulted in greater than 1 week of time loss (52%) in Bantam only. The reasons behind this trend are not clear, particularly since factors that one could hypothesise would impact injury severity (such as size differentials and previous experience) were not identified as risk factors for injury in this study.

Our final objective was to examine risk factors for injury in this population. Previous injury is identified as a risk factor in our study and is also consistently a risk factor for injury across all populations.20 ,21 This may be due to incomplete rehabilitation, decreased physical fitness, persistent instability, premature return to play and underestimation of injury severity.22 ,23 A greater risk of injury in games versus practices is consistent with the literature.5 ,24–26 A unique finding in this study was that players who had begun to menstruate by ages 11/12 (PeeWee) were at an increased risk of injury over those who had not yet begun to menstruate. It is possible that the hormone changes that have been shown to increase risk of ACL injury in women influence other types of injury as well.27 ,28 While this is a non-modifiable risk factor in itself, it is possible that menarche is a marker of other (possibly modifiable) changes that we have been unable to elucidate in this study.

Age group has been shown to be a risk factor in male youth ice hockey5 and point estimates suggest that age group may be a risk factor for injury in this study. Overall, Midget players were at threefold greater risk of injury; however, this was not statistically significant. Also consistent was the finding that more elite Bantam players (level A) demonstrated a greater than fivefold greater risk than non-elite (level B) players in this age group, though again it was not statistically significant.5 Interestingly, Midget B players in this study were found to have a higher risk of injury than the more elite divisions (A, AAA). This was unexpected and contrary to the previous findings in male youth.5 One possible explanation for this may be that players new to ice hockey in that age group would play in the B level, but this was not examined specifically in this study. Unexpected findings such as these, however, highlight the importance of studying female youth separately to discover these trends, which will hopefully inform injury prevention strategies in the future.

Other risk factors examined that were not found to be significant included height, weight, player position, relative age and ice hockey experience. Risk factors including age,5 ,8 ,23 ,29–31 relative age,32 level of play,5 ,32 player position,8 ,31 weight and height7 ,24 ,25 have all been found to be significant risk factors in male youth ice hockey and as such deserve further investigation in a larger sample of female youth ice hockey players.


One team was lost to follow-up and four teams decided not to participate. It is unlikely, however, that the participating teams differed from the non-participating teams with respect to injury risk, given the reasons for non-participation (ie, no volunteers to act as the team designate). Many of our independent variables were self-reported at the preseason and therefore may be associated with bias or, given the age of the youngest players in the study, may not have been truly self-reported if the parent/guardian had to assist the player complete the forms. It is also quite possible that the measured baseline characteristics changed throughout the season and we would not have captured these changes with our study (ie,: player position as reported at baseline might have changed several times throughout the ice hockey season), and therefore our results examining these variables as risk factors for injury are limited by this fact. Our analyses were exploratory; however, given the paucity of studies in female youth ice hockey, this investigation sets the stage for future research.

Despite individual team designate training regarding study injury definition and identification of any suspected concussion, it is possible that team designates did not identify all suspected concussions if a player was not removed from play. Further, it is acknowledged that there is a subjective element to injury reporting as the player or parent may under or over report an injury not identified by the team designate. Recording individual exposure data inherently overestimates the injury rate denominator, thereby possibly underestimating IR, particularly in games where players are not on the ice for the entire session time. However, this method is consistent with the literature and therefore allows us to make comparisons with other studies.

Time loss as an indirect measure of injury severity is limited. Other factors, such as willingness to play with pain or under-reporting pain, can impact the time to return to play. Finally, we have stratified by age group, and therefore some of our analyses are based on small numbers of outcomes (eg, menarche in PeeWee), which does limit our conclusions; however, we do provide 95% CI's, so this is clear to the reader.


This is the first study in which female youth ice hockey IR and risk factors have been examined, and it utilised prospective injury surveillance methods. All study questionnaires were previously validated. Clustering by team was considered in the analysis, producing more conservative 95% confidence limits.


The extent of the problem of injury in female youth ice hockey players was previously unknown. IR reported are lower than those for youth male or women's ice hockey, a finding that will hopefully work to promote participation so that girls may reap the many benefits of participation.2 ,3 However, with growing rates of participation in female youth ice hockey, injury remains a public health concern and deserves future attention in the literature.

Future studies should further examine risk factors (both physical and behavioural) for injury in female youth ice hockey to facilitate the development and evaluation of interventions aimed at decreasing injury in this population. For example, given the finding that strains and sprains were the most frequent injury types in this study, a future study may include an intervention component with a warm-up with dynamic stretching or a neuromuscular training programme. Player education sessions regarding how to safely receive body contact or body checking could also be studied as an intervention to see if there is any effect on IR and concussion rates. Future studies in this population would also benefit from having a larger sample size to increase the number of observed outcomes. This study has laid a foundation for future research by identifying risk factors for injury and injury trends that can be targeted in future studies in female youth ice hockey players.

What are the new findings?

  • The overall injury rate was 1.9 injuries/1000 player-hours (95% CI 1.5 to 2.7) and incidence proportion was 16.3 injuries/100 players (95% CI 10.8 to 24.9). These rates are lower than those found in adult female and male youth ice hockey.

  • Previous injury (IRR=2.7, 95% CI 1.7 to 4.3), session type (games) (IRR=2.1, 95% CI 1.0 to 4.2) and menarche (in PeeWee only) (IRR=4.1, 95% CI 1.0 to 16.8) were risk factors identified in female youth ice hockey.

  • In Midget only, a decrease in the risk of injury was seen in the more elite divisions of play (IRR=0.2, 95% CI 0.1 to 0.5) in A and (IRR=0.5, 95% CI 0.2 to 0.9) in AAA.


Dr Emery and Dr Hagel are supported by Population Health Investigator Awards from the Alberta Heritage Foundation for Medical Research and New Investigator Awards from the Canadian Institutes of Health Research. Dr Emery holds a Professorship in Paediatric Rehabilitation funded through the Alberta Children's Hospital Foundation. Dr Hagel holds a Professorship in Child Health and Wellness funded by the Alberta Children's Hospital Foundation, through the support of an anonymous donor and the Canadian National Railway Company. Melissa Decloe's MSc was supported by the Social Sciences and Humanities Research Council of Canada (SSHRC), a Department of Paediatrics Child Health Research Group Scholarship and the Joanne A. Vincenten Scholarship from the Alberta Centre for Injury Control.


View Abstract


  • Contributors MDD, CAE, BH and WHM participated in concept and design. MDD and CAE participated in data collection. MDD, CAE, BH and WHM participated in the analysis and interpretation of data. MDD and CAE participated in drafting of the manuscript. MDD, CAE, BH and WHM participated in a critical revision of the manuscript. MDD and CAE obtained funding. MDD, CAE were involved in administrative work. CAE, BH, WHM were involved in study supervision.

  • Funding This research project was funded through an Investigator-driven grant from the Alberta Centre for Child, Family and Community Research (grant number 08SM-EMERY).

  • Competing interests None.

  • Ethics approval University of Calgary, Conjoint Health Research Ethics Board.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Data can be requested by contacting the corresponding author at

  • ▸ References to this paper are available online at