Risk of head injury associated with distinct head impact events in elite women's hockey
Show all authors
G. Kosziwka, L. Champoux , J. Cournoyer, ...
First Published December 1, 2021 Research Article
https://doi.org/10.1177/20597002211058894
Article information
Abstract
Head injuries are a major health concern for sport participants as 90% of emergency department visits for sport-related brain injuries are concussion related.1 Recently, reports have shown a higher incidence of sport-related concussion in female athletes compared to males.3 Few studies have described the events by which concussions occur in women's hockey,2,7,8 however a biomechanical analysis of the risk of concussion has not yet been conducted. Therefore, the purpose of this study was to identify the highest risk concussive events in elite women's hockey and characterize these events through reconstructions to identify the associated levels of peak linear and angular acceleration and strain from finite element analysis.
44 head impact events were gathered from elite women's hockey game video and analyzed for impact event, location and velocity. In total, 27 distinct events based on impact event, location and velocity were reconstructed using a hybrid III headform and various testing setups to obtain dynamic response and brain tissue response. A three-way Multivariate Analysis of Variance (MANOVA) was conducted to determine the influence of event, location and velocity. The results of this study show that head- to-ice impacts resulted in significantly higher responses compared to shoulder-to- head collisions and head-to boards impacts however, shoulder and boards impacts were more frequent. All events produced responses comparable to proposed concussion threshold values.21 This research demonstrates the importance of considering the event, the impact characteristics, the magnitude of response, and the frequency of these impacts when attempting to capture the short and long term risks of brain trauma in women's hockey.
Keywords
Sports-related concussion, concussion, mTBI
Introduction
Head injuries are a major public health concern for sport participants. According to the Canadian Institute for Health Information, 9 out of every 10 emergency department visits for sport-related brain injuries are concussion related.1 Concussion is also known as mild traumatic brain injury (mTBI) and is the most common form of traumatic brain injury. Participants of contact sports, such as ice hockey, experience a higher risk of head injury, due to increased exposure to head impacts in addition to players moving at fast speeds.2 Recently, reports have indicated a higher incidence of sport-related concussion in female athletes compared to their male counterparts.3–5 Women's ice hockey is a high velocity sport, which includes many situations that involve impacts to the head. The NCAA Injury Surveillance System (ISS) reported women's hockey to have the highest rate of concussions (0.91/1000 A-Es) of 16 males and female collegiate-level sports,6 despite body checking being illegal. Thus, It has been reported that head impact events occur approximately half as frequently in women's hockey as in men's hockey.7,8 As a result, it is understood that the higher incidence of concussion in women's hockey when compared to men's hockey is not a result of increased head impact exposure. Researchers have identified concussive events in ice hockey in an attempt to explain gender differences.2,3 The most common event causing concussion in women's hockey occurs when players fall, hitting their head against the ice or onto the sideboards of the arena. However, the non-concussive head impact event that is most frequent in women's hockey is player to player collisions accounting for 50% of all head impacts; and falls to the ice or boards resulting in a head impact only accounts for 30% of total head impacts.8 The differences in rules between males and females does not support why there is an increasing concussion reporting in female players. To further understand the discrepancies in the incidence of concussion diagnoses and further understand how the risk of head injuries is created in women's hockey, the frequency and magnitude of common head impacts needs to be analyzed. The purpose of this study was to document the most common head impact events in elite women's hockey players and compare the peak linear acceleration, peak rotational acceleration, and peak maximum principal strain between the events to determine the type of head impacts creating risk of head injury in elite women's ice hockey.
Methods
Fifteen games of elite women's ice hockey were analyzed to catalogue head impact events. The videos included games from online archives of the Canadian Women's Hockey League (WHL), the National Women's Hockey League (NWHL), and International Ice Hockey Federation (IIHF) between 2015 and 2017. Head impact event inclusion criteria consisted of head impacts where resulting head motion could be clearly observed. A total of 35 head impacts met the inclusion criteria including 13 head-to-shoulder, 18 head-to-board, and 3 head-to-ice events.
Additional video analysis was performed to obtain head impact characteristics necessary for laboratory reconstructions such as impact velocity, and impact location/angle. Impact velocity from collisions events was calculated using Kinovea (version 0.8.20) to determine the distance between the head of the player and the impacting surface over the time needed for the impact to occur.9 The time prior to the impact was set at a maximum of 0.2 s to minimize the error associated with a change in velocity prior to the impact. This method of video analysis requires additional inclusion criteria to minimize errors in velocity calculations: 1) the head must be visible in the few frames (<5) prior to and at the moment of impact and 2) the presence of markings near the impact location on the ice must be visible to allow for calibration of the distance. Impact velocities from head-to-ice events were determined using MADYMO (TASS International, The Netherlands) where the limbs and positioning of the model was manipulated to replicate the condition observed on the video. No other metrics were obtained from MADYMO. Impact locations were catalogued using the reference grid presented in Figure 1.
Figure 1. Transverse and vertical reference grids used for determination of head impact location.
Ten head-to-shoulder, six head-to-boards and four head-to-ice impacts could be analyzed for velocity and location and were included for reconstructions. The ranges of velocity for each event type are presented in Table 1.
Table 1. Impact velocities of head-to-shoulder, head-to-boards, and head-to-ice included for laboratory reconstructions.
Table 1. Impact velocities of head-to-shoulder, head-to-boards, and head-to-ice included for laboratory reconstructions.
View larger version
Reconstructions were performed using the three most common impact locations for each event type (Figure 2) and impact velocities consisting of the average for each event and an upper and lower boundary of ±1 m/s. In the case of head-to-ice events, only two impact locations were recorded. An additional impact location to the side of the head was reconstructed as it represents a common impact location for concussive impacts in women's ice hockey as reported by.2
Figure 2. Impact locations selected for reconstructions for head-to-shoulder, head-to-board, and head-to-ice events.
Figure 3. Pneumatic linear impactor consisting of a steel frame (A), compressed air tank (B), piston chamber (C), impacting arm (D), time gate (E), impacting cap (F), sliding table (G). (Meehan MSc Thesis, 2019).
Figure 4. Monorail drop rig shown with vertical rail (A), motorized release system (B), drop carriage (C), 45° steel anvil (D), time gate (E), concrete base (F).
Figure 5. Mean peak resultant linear acceleration for the board, ice, and shoulder events.
Figure 6. Mean peak resultant angular acceleration for the board, ice, and shoulder events.
Figure 7. Mean MPS for the board, ice, and shoulder events.
Equipment
Linear impactor
A pneumatic linear impactor was used to represent shoulder-to-head impact conditions. The system consists of a pneumatically accelerated impacting arm encased in a standing frame of mass 13.1 kg. When engaged the arm moves towards a helmeted headform mounted on a sliding table allowing it to slide with little resistance after impact. This sliding table that supports the Hybrid III headform can be adjusted to five degrees of freedom including fore-aft (x-axis), lateral (y-axis), up-down (z-axis), fore-aft (y-axis), and axial (x-axis) rotation of the neckform. This allowed the impact location and direction described from the video analysis to be accurately reflected by the headform test set up.10 An electronic time gate measures impact velocity just prior to impact and was used to match the impact velocities obtained from video analysis.
A shoulder pad cap consisting of a nylon disc covered with 142 mm of vinyl nitrate R338 V foam material under a Reebok 11 K shoulder pad was mounted on the end of the impacting arm to represent shoulder-to-head collisions.11 A shoulder-to-head event is considered a high mass event where the striking mass has been calculated to be approximately 15% of the striking player's mass.12 The average body mass of a women's hockey player on the Canadian national women's hockey team is listed as 70 kg.13 Effective mass for women's elite hockey players can be calculated as approximately 10.5 kg. In this study, a striking mass of 13.1 kg was used as it is the lowest mass that can be used with the linear impactor. Karton et al. demonstrated that the effect of striking mass had little effect above 10 kg when using an MEP striker.14 It is unlikely it would have a different effect using a compliant surface such as the shoulder.
Monorail drop Rig
In this study, the monorail drop rig was used to represent head to ice falls and head to board impacts. The headform and neck were attached to a drop carriage affixed onto a monorail drop rig of maximum height of 4.7 m rail. A pneumatic piston is responsible for releasing the drop carriage and attached headform when It reaches the appropriate height for the desired impact velocity. The impact anvil can be adjusted to represent various surfaces including ice or hockey boards.
Hybrid III headform and unbiased neck
In this study, a Hybrid III 5th-percentile female headform with a circumference of 21.2 inches and composed of steel covered in a vinyl skin, was used to recreate head impact events in women's hockey (Humanetics, Plymouth, MI, USA). An unbiased neckform designed to allow similar movement in all planes of motion without directional bias was used for the reconstructions. This unbiased neckform was scaled from the 50th-Hybrid III neckform15 with four centred and unarticulated rubber butyl disks (radius 27.5 mm; height 18.0 mm) recessed slightly (3.2 mm) and serially inside aluminum disks (radius 34.5 mm; height 12.5 mm) (Figure 2). Nine single-axis accelerometers positioned in a 3-2-2-2 array captured the three-dimensional acceleration-time curves for linear and rotational acceleration.
Finite element model
Linear and rotational acceleration time histories were used as an input to the University College Dublin Brain Trauma Model (UCDBTM) to calculate maximum principal strain (MPS). The UCDBTM was developed using CT and MRI scans and validated using cadaver data16 and is modeled as viscoelastic to represent the shear behavior of brain tissue as elastic for the compressive behavior’s of brain tissue. Results are presented as a gradient of deformation from low to high. In terms of maximum principal strain, results are presented as a percentage of strain. Additional brain material properties are described by Horgan & Gilchrist (2003). The compression of the brain tissue was defined as elastic. The shear characteristic of the viscoelastic brain was expressed:
G(t) = G∞ + (G0 − G∞)e−βt
with G∞ representing the long term shear modulus, G0 the short term modulus, and β the decay factor.
Statistical analysis
Three one-way ANOVAs were used to determine the difference in peak linear acceleration, peak rotational acceleration and MPS between head-to-shoulder, head-to-board, and head-to-ice events. Tukey's post hoc tests were then conducted to determine mean comparisons between variables. Significance was accepted at p < 0.05.
Results
The mean peak linear acceleration, peak rotational acceleration and MPS are presented in Table 2. Significant differences between head impact events were detected for each of the variables. Differences in peak linear accelerations were detected between head-to-ice and both head-to-shoulder (p < 0.01) and head-to-board events (p < 0.01), where head-to-ice events were significantly greater than the other two types of events. There was no difference in peak linear acceleration between head-to-shoulder, and head-to-board events (p = 0.1). Peak rotational acceleration was statistically greater in head-to-ice events compared to head-to-shoulder (p < 0.01), and head-to board (p < 0.01). There were no significant differences between head-to-shoulder and head-to-board for peak rotational acceleration (p = 0.376). The maximum principal strain was significantly different between all three impact events. Head-to-ice was greater than both head-to-shoulder (p < 0.01), and head-to-board (p < 0.01). The MPS associated with head-to-shoulder events was also significantly greater than head-to-board events (p < 0.01).
Table 2. Mean peak linear acceleration, peak rotational acceleration, and MPS for head impact events. () = z scores.
Table 2. Mean peak linear acceleration, peak rotational acceleration, and MPS for head impact events. () = z scores.
View larger version
Discussion
The purpose of this study was to identify the most common head impact events in elite women's ice hockey and compare dynamic response and maximum principle strain between each event type to determine how risk of injury occurs in women's hockey. The most frequent event type documented in this study was head to boards, accounting for 51% of the impacts (18 impacts). The shoulder to head represented 37% of the impacts (13 impacts) and the head to ice accounted for 12% (4 impacts). This distribution of events is different from previously reported distribution of concussive or non-concussive events. Concussive injury events reported in women's ice hockey, are caused primarily by falls to the ice or boards representing approximately 50–60% of diagnosed concussions; but represent only 30% of overall head impact frequency.2,3,8 Despite penalties awarded for body checking in women's hockey, collisions cause 27–42% of concussions in elite women's hockey.2,3 The differences between concussive event types and the distribution of impact events presented in this study can be explained by our methodology and player sample. The diagnosis associated with the events of this study were not available publicly and may include a mix of both concussive and non-concussive events and were limited to what could be observed clearly on video. In addition, the population used in this study is composed of professional women's hockey players as opposed to high school and college players. These players are highly skilled and may have learnt to protect themselves better over time than high school and collegiate players. In contrast, men's professional players sustain concussion primarily via shoulder to head impacts accounting for 53% of concussions,17,18 which is likely a result of allowing body checking in men's hockey.
Head to ice impacts only represented a low proportion of head impacts but resulted in significantly higher peak resultant linear acceleration and peak resultant rotational acceleration when compared to head to board and shoulder to head impacts. Head to ice impacts resulted in an average peak linear acceleration of 123.6 g and an average peak rotational acceleration of 9518 rad/s2. Both measures fall far above reported values resulting in concussion in ice hockey.18 The high average velocity (4.8 m/s) of head to ice impacts observed in this study with the addition of the low compliant surface of the ice, may be the reason for high magnitudes of dynamic response.19 Although not frequent, these high values demonstrate the continued need for hockey helmets to be developed to protect against falls to the ice, as this is a high-risk event for concussion. To avoid falls to the ice, skating and balance skills should be developed at an early age. The head to boards impacts and shoulder to head collisions did not have significantly different peak linear and rotational accelerations.
Maximum principal strain (MPS) was the only measure to indicate significant differences between the all three head impact events highlighted in the study. Head to ice falls resulted in the highest average MPS of 0.51 compared to the head to board (0.19) and shoulder to head events (0.27). Similarly to dynamic response, MPS identifies head to ice has the riskiest events in terms of magnitude. The mean magnitude of MPS for the three events represent approximately 50% or above risk of concussion.20,21
While peak linear and rotational acceleration reported significantly higher values in the head to ice events than head to boards and shoulder to head; MPS demonstrated differences between all three event types. This suggests that MPS may be more sensitive to change in direction, slope, and duration of acceleration as opposed to only using magnitude of head acceleration.22 For example, the longer duration of the shoulder to head compared to the head to boards impacts may have contributed to the higher values of MPS.
While head impacts to the ice produced the highest MPS, it was far less frequent than the shoulder to boards accounting for only 11% of the head impacts documented in this study. The head to boards (51%) and shoulder to head (37%) impacts occurred much more frequently and the MPS associated with these events represents above 50% and 80% risk of concussive injury, respectively. Currently, the standards test for ice hockey helmets include only a flat drop test; a similar mechanism to a fall to the ice.23 Therefore, in order to properly protect against concussive injury in elite women's ice hockey, head impacts to the boards and shoulders of opposing players need to be considered when designing and testing protective equipment. Standards organizations could aim to decrease MPS and test using moderate magnitude, long duration impacts.
Several researchers have demonstrated that a higher number of concussions are reported by female athletes in comparison to their male counterparts competing in the same sport.4,24,25 Female athletes also experience concussions differently than males reporting more symptoms of greater severity and needing longer recovery times.26–29 Interestingly, the dynamic response and brain tissue deformation values reported in this study have similar or lower magnitudes than those reported in elite male hockey,18,30 with the exception of MPS for head to ice events which can be attributed to the smaller mass of the head form used to conduct women's reconstruction. The frequency of head impact events observed in this study is much lower than what is seen for professional men's hockey players.17,18 This suggests that the difference in prevalence of concussion and recovery may not be related to how the game is played but may be due to internal factors such as willingness to report injury. While elite men's ice hockey is a full-time job for the players, women who participate in ice hockey at the elite level are typically not paid as well as the men. The highest paid women in the NWHL in 2016 earned $25,000 a year and the CWHL did not pay its athletes until 2017 with a salary ranging from $2000 to $10,000. This suggests that the athletes in these leagues need an additional source of income such as another occupation and may be less willing to play injured. While there have been no known studies connecting salary and willingness to report injuries, the difference in pay between men and women could be a reason for the difference in incidence of concussion due to willingness to report the injury. Other biological factors beyond the scope of this study may be able to further explain the differences in injuries incidence between male and female hockey players.
Limitations
Only 15 games were available for analysis, which resulted in a low number of impacts that could be analyzed. Some impacts were excluded from this analysis due to obstruction of the view during the impact. This could have skewed the distribution towards impacts of higher magnitudes that resulted in replays from the broadcast and allowed for better analysis. In addition, the 5th percentile Hybrid III head form may not represent all head sizes of the observed players but they produce highly reproducible results, which make them useful for impact testing. Finally, the finite element brain model is based on a 50th percentile male and was scaled to calculate the maximum principal strain of female athletes. No model representing female athletes currently exists.
Conclusion
The most common head impact events in elite women's ice hockey based on video analysis of 15 games were: head to boards (51%), shoulder to head (37%), and the head to ice events (12%). Falls to the ice demonstrated the highest dynamic response values compared to head to boards and shoulder to head. MPS was highest in head to ice followed by shoulder to head, and finally head to boards, however, shoulder to head and head to boards impacts occurred more frequently and still resulted in a high-risk event for concussive injury. This suggests that protective equipment should be designed to protect against these events to reduce the risk of head injury in elite women's hockey.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship and/or publication of this article.
ORCID iD
L. Champoux
https://orcid.org/0000-0003-2954-6050
References
1. Canadian Institute for Health Information , Head’s Up on Sport Relates Brain Injuries,
https://www.cihi.ca/en/heads-up-on-spor ... injuries-0 (2018, accessed 25 November 2020).
Google Scholar
________________________________________
2. Delaney, JS . Mechanisms of injury for concussions in university football, ice hockey, and soccer. Clin J Sport Med 2014; 24: 233–237.
Google Scholar | Crossref | Medline
________________________________________
3. Agel, J, Dick, R, Nelson, B, et al. Descriptive epidemiology of collegiate women’s ice hockey injuries: national collegiate athletic association injury surveillance system. 2000-2001 through 2003-2004. Journal of Athletic Training (National Athletic Trainers’ Association) 2007; 42: 249–254.
Google Scholar | Medline | ISI
________________________________________
4. Dvorak, J, McCrory, P, Kirkendall, DT. Head injuries in the female football player: incidence, mechanisms, risk factors and management. Br J Sports Med 2007; 41: i44–i46.
Google Scholar | Crossref | Medline | ISI
________________________________________
5. Gessel, LM, Fields, SK, Collins, CL, et al. Concussions among United States high school and collegiate athletes. J Athl Train 2007; 42: 495–503.
Google Scholar | Medline | ISI
________________________________________
6. Hootman, JM, Dick, R, Agel, J. Epidemiology of collegiate injuries for 15 sports: summary and recommendations for injury prevention initiatives. J Athl Train 2007; 42: 311–319.
Google Scholar | Medline | ISI
________________________________________
7. Brainard, L, Beckwith, J, Chu, J, et al. Gender differences in head impacts sustained by collegiate ice hockey players. Med Sci Sports Exercise 2012; 44: 297–304.
Google Scholar | Crossref | Medline | ISI
________________________________________
8. Wilcox, B . Head impact exposure in male and female collegiate ice hockey players. J Biomech 2004; 47: 109–114.
Google Scholar | Crossref
________________________________________
9. Post, A, Koncan, D, Kendall, M et al. Analysis of speed accuracy using video analysis software. Sport Engineering 2018; 21: 235–241.
Google Scholar | Crossref
________________________________________
10. Walsh, E . The influence of impact location and angle on the dynamic impact response of a hybrid III headform. Sports Eng 2011; 13: 135–143.
Google Scholar | Crossref
________________________________________
11. Rousseau, P, Post, A, Hoshizaki, TB. A comparison of peak linear and angular headform accelerations using ice hockey helmets. J ASTM Int 2009a; 6: 1–11.
Google Scholar
________________________________________
12. Rousseau, P, Hoshizaki, TB. Defining the effective impact mass of elbow and shoulder strikes in ice hockey. Sports Biomech 2015. In press.
Google Scholar | Crossref | Medline
________________________________________
13. Hockey Canada , Canadian National Women’s Team Roster 2018-2019,
https://www.hockeycanada.ca/en-ca/team- ... teamid=179 (2019, accessed 25 November 2020).
Google Scholar
________________________________________
14. Karton, C, Hoshizaki, TB, Gilchrist, M. The influence of impactor mass on the dynamic response of the hybrid III headform and brain tissue deformation. ASTM International 2013; Mechanism of Concussion in Sports: 23–40.
Google Scholar
________________________________________
15. Walsh, K . Comparative analysis of hybrid III neckform and an unbiased neckform. Sports Eng 2018; 21: 479–485.
Google Scholar | Crossref
________________________________________
16. Horgan, TJ, Gilchrist, MD. The creation of three-dimensional finite element models for simulating head impact biomechanics. Int J Crashworthiness 2003; 8: 353–366.
Google Scholar | Crossref | ISI
________________________________________
17. Hutchison, MG, Comper, P, Meeuwisse, WH, et al. A systematic video analysis of national hockey league (NHL) concussions, part II: how concussions occur in the NHL. Br J Sports Med 2015; 49: 552–555.
Google Scholar | Crossref | Medline | ISI
________________________________________
18. Post, A, Hoshizaki, TB, Karton, C, et al. The biomechanics of concussion for Ice hockey head impact events. Comput Methods Biomech Biomed Engin 2019; 22: 631–643.
Google Scholar | Crossref | Medline
________________________________________
19. Post, A , et al. Peak linear and rotational acceleration magnitude and duration effects on Maximum principal strain in the corpus Callosum for sport impacts. J Biomech 2017; 61: 183–192, Elsevier td,.
Google Scholar | Crossref | Medline
________________________________________
20. Kleiven, S . Predictors for traumatic brain injuries evaluated through accident reconstruction. Stapp Car Crash J 2007; 51: 81–114.
Google Scholar | Medline
________________________________________
21. Zhang, L, Yang, KH, King, AI. A proposed injury threshold for mild traumatic brain injury. J Biomech Eng 2004; 126: 226–236.
Google Scholar | Crossref | Medline | ISI
________________________________________
22. Post, A, Oeur, A, Walsh, E, et al. A centric/non- centric impact protocol and finite element model methodology for the evaluation of American football helmets to evaluate risk of concussion. Comput Methods Biomech Biomed Engin 2014; 17: 1785–1800.
Google Scholar | Crossref | Medline | ISI
________________________________________
23. ASTM F1045-16 , Standard performance specification for Ice hockey helmets. West Conshohocken, PA: ASTM International, 2016.
Google Scholar
________________________________________
24. Covassin, T, Swanik, C, Sachs, M. Sex differences and the incidence of concussions among collegiate athletes. J Athl Train 2003; 38: 238–244.
Google Scholar | Medline | ISI
________________________________________
25. Dick, RW . Is there a gender difference in concussion incidence and outcomes? Br J Sports Med 2009; 43: i46–i50.
Google Scholar | Crossref | Medline | ISI
________________________________________
26. McLeod, TCV, Leach, C. Psychometric properties of self-report concussion scales and checklists. J Athl Train 2012; 47: 221–223.
Google Scholar | Crossref | Medline | ISI
________________________________________
27. Broshek, DK, Freeman, JR. Psychiatric and neuropsychological issues in sport medicine. Clin Sports Med 2005; 24: 663–679.
Google Scholar | Crossref | Medline | ISI
________________________________________
28. Kutcher, JS, Eckner, JT. At-risk populations in sports-related concussion. Curr Sports Med Rep 2010; 9: 16–20.
Google Scholar | Crossref | Medline | ISI
________________________________________
29. Covassin, T, Schatz, P, Swanik, B. Sex differences in neuropsychological function and post-concussion symptoms of concussed collegiate athletes. Neurosurgery 2007; 61: 345–351.
Google Scholar | Crossref | Medline | ISI
________________________________________
30. Rousseau, P . Analysis of Concussion Metrics of Real-World Concussive and Non-injurious Elbow and Shoulder to Head Collisions in Ice Hockey. PhD Thesis. University of Ottawa, Canada, 2014.
Google Scholar
View Abstract
https://journals.sagepub.com/doi/full/1 ... 2211058894