Article Text

Download PDFPDF

Epidemiology of injuries in hurling: a prospective study 2007–2011
  1. Catherine Blake1,
  2. Edwenia O'Malley1,
  3. Conor Gissane2,
  4. John C Murphy3
  1. 1UCD School of Public Health, Physiotherapy and Population Science, Health Sciences Centre, Belfield, Dublin, Ireland
  2. 2School of Sport, Health and Applied Science, St Mary's University College, London, Twickenham, UK
  3. 3Medfit Proactive Healthcare, Blackrock, Ireland
  1. Correspondence to Dr Catherine Blake; c.blake{at}


Objectives Hurling is a stick handling game which, although native to Ireland, has international reach and presence. The aim of this study was to report incidence and type of injuries incurred by elite male hurling players over five consecutive playing seasons.

Design Prospective cohort study.

Setting Male intercounty elite sports teams participating in the National GAA Injury Database, 2007–2011.

Participants A total of 856 players in 25 county teams were enrolled.

Primary and secondary outcomes Incidence, nature and mechanism of injury were recorded by team physicians or physiotherapists to a secure online data collection portal. Time-loss injury rates per 1000 training and match play hours were calculated and injury proportions were expressed.

Results In total 1030 injuries were registered, giving a rate of 1.2 injuries per player. These were sustained by 71% (n=608) of players. Injury incidence rate was 2.99 (95% CI 2.68 to 3.30) per 1000 training hours and 61.75 (56.75 to 66.75) per 1000 match hours. Direct player-to-player contact was recorded in 38.6% injuries, with sprinting (24.5%) and landing (13.7%) the next most commonly reported injury mechanisms. Median duration of time absent from training or games, where the player was able to return in the same season, was 12 days (range 2–127 days). The majority (68.3%) of injuries occurred in the lower limbs, with 18.6% in the upper limbs. The trunk and head/neck regions accounted for 8.6% and 4.1% injuries, respectively. The distribution of injury type was significantly different (p<0.001) between upper and lower extremities: fractures (upper 36.1%, lower 1.5%), muscle strain (upper 5.2%, lower 45.8%).

Conclusions These data provide stable, multiannual data on injury patterns in hurling, identifying the most common injury problems. This is the first step in applying a systematic, theory-driven injury prevention model in the sport.

  • Athletic Injuries
  • Sports Medicine
  • Epidemiology
  • Incidence

This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 3.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See:

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.

Strengths and limitations of this study

  • This is the first study to report on prospective surveillance of injury incidence over multiple playing seasons in the sport of hurling.

  • The use of consensus definitions for injury enables comparison with incidence rates in other sports.

  • Minor injuries that did not require time-out from play were not captured in the injury definition.

  • The sample was limited to the elite hurling population.


The Gaelic Athletic Association (GAA) governs three amateur sporting codes: Gaelic football, hurling and handball. Of these, hurling is perhaps the most unique, predominantly played in Ireland, but hurling clubs exist in Britain, continental Europe, the USA, Canada and Australasia.1 This reflects the Irish cultural diaspora, and with growing interest and adoption of these games, overseas branches of the governing body continue to expand. In hurling an ash stick, called a hurley or camán, is used to propel a hard leather ball, called a slíotar (diameter 69–72 mm, weight 110–120 g, figure 1). Teams of 14 outfield players and a goalkeeper play on a rectangular grass pitch 145 m long and 90 m wide, for durations of 60–70 min per game.2 The aim is to score by sending the ball between the opposition's goal posts. The ball is propelled through the air, at velocities up to 160 km/h, or along the ground, but kicking and hand passing are also permitted.2 Other core skills include catching, blocking and lifting the ball with the stick, maintaining possession while running with the ball balanced or bouncing on the stick and striking the ball while stationary or running.2 Close player-to-player contact occurs in competing for the ball and in the tackle where dispossession is by means of contesting the opponents’ attempts to strike the ball (blocking and hooking), or through a shoulder-to-shoulder body clash.2

Figure 1

Hurley, slíotar and helmet.

The biomechanical demands of this game include jumping, landing, sprinting, rapid acceleration, deceleration, torsional movements and directional changes, as well as evasion through planting and cutting manoeuvers. Such actions pose risks for lower limb injury in particular, while the speed, intensity and force of the stick-to-stick, or stick-to-player contact give rise to direct traumatic injuries to the upper and lower limbs and trunk. Protective helmets and face guards have been mandatory for all grades of players since January 2010.2

Until now, research into the epidemiology of injury in hurling has primarily focused on one body region or on injuries presenting to the hospital emergency department.35 An early prospective study profiled injury in 74 players,6 while a more recent paper describes hurling injury in 127 elite male players over one playing season.7 Such a single snapshot view, however, does not account for season-to-season variation, thus the focus of this report is to extend the prospective surveillance period to five complete competitive seasons, as the first phase in the TRIPP8 and Van Mechelen sports injury prevention models.9 These models define epidemiological research as the first step in injury prevention, allowing quantification of injury and associated risk factors. This then provides a platform for development, implementation and evaluation of injury prevention interventions, in the context of controlled research and real-world sport environments. This study highlights key injuries and provides direction for future research into risk factors and prevention strategies. The aim was therefore to describe incidence, mechanism, nature and severity of injury in elite male hurling over a 5-year time span. Differences between subgroups of players based on age and playing position were also explored.


The men's senior grade county representative hurling competition starts in January, running through to September. The season includes preliminary cup and shield competitions, followed by the National Senior Hurling League and culminating in the All Ireland Senior Hurling Championship. This study focuses on teams enrolled in the National GAA Injury Database during the years 2007–2011. The data collection system opened from 1 January each year, with teams prospectively followed until eliminated from the competition. Data collection ceased for the off-season, following the All Ireland Hurling Final, restarting in January the following year.

The only inclusion criterion for participation was that the team had a qualified professional that is, a medical practitioner or a chartered physiotherapist present at every match and training session who could verify injury diagnosis and classification as well as game and training exposure hours. Injury data were entered weekly by the team personnel through a dedicated secure web portal, recording the information onto the National GAA Injury Database. The participants were male players selected for their representative county team and the total sample recruited was determined by the number of teams who volunteered to participate.

Ethical approval

Only de-identified player data were recorded. Players were given an opportunity to decline inclusion of their data in the team reports. Anonymity was maintained and data protection assured in accordance with ethical approval received from the University Research Ethics Committee (LS-E-11-91-OMalley-Blake).


Consensus injury definitions for hurling were agreed with the GAA Medical, Scientific and Player Welfare Committee, following a review of international literature. These have been applied to hurling and Gaelic football, its sister sport, and have already been described in detail.7 ,10 An information pack regarding injury definition and classification was distributed to each participating medical team and these were also embedded into the user interface in the online data collection tool.

Injury was defined as a time-loss injury, that is, ‘any injury that prevents a player from taking a full part in all training and match play activities typically planned for that day, where the injury has been there for a period greater than 24 h from midnight at the end of the day that the injury was sustained’. This definition mirrors that employed by Brooks et al,11 and conforms with consensus time loss injury definitions proposed for soccer and rugby union.12 ,13 Recurrence of injury was defined as ‘a reinjury to a previously injured region’. This was subclassified according to duration since original injury into early recurrent (within 2 months), late recurrent (2–12 months) and delayed recurrent (>12 months). Return to full fitness was deemed to be when the player was able to take part in full training activities and was available for match selection. Other agreed definitions included classification into acute injuries, overuse injuries or chronic injuries similar to the description used by Van Mechelen et al.9 Severity of injury was classified as mild (lasting up to 1 week), moderate (up to 4 weeks) or severe (>4 weeks), with these times similarly relating to absence from training or match play.9


The initial enrolment of players required that anthropometric and demographic details, position of play, involvement in other levels of competition, past injury and use of protective equipment were recorded. Age was defined in years, as on 1 January of that year. Players joining or leaving the squad were added or deactivated as required throughout the season. Thereafter, the database was open for entry of new or updating of existing injury diagnosis at any time, but weekly injury data entry was required at minimum, as were details of training and match play exposure hours. For new injuries, the team doctor or physiotherapist recorded: player code, position of play, ground conditions, date of injury, mechanism of injury, body region, main tissue injured, side of injury, whether recurrent or new injury and clinical diagnosis. Progression details for current injured players were also required weekly, including update of the status that is, whether still injured or date of return to partial or full fitness.


The data were analysed by calculating injury rates per 1000 h, with 95% CIs, using the substitution method.14 ,15 Percentages with 95% CI were derived using Wilson's method.15 Risk ratios (RR) and 95% CI were used to compare injury rates. χ2 Tests compared the observed with the expected proportions of injured players according to playing position and age group. Computations were made using PASW Statistics, Release V.18.0.0,16 VRP Injury Analysis Software17 and the CI Analysis Package V.


Twenty-five male hurling teams were recruited; 4 for 2007, 5 for 2008, 7 for 2009, 5 for 2010 and 4 for 2011, so between 12.5% and 22% of competing teams per year were enrolled over this period. A total of 856 player seasons were followed and no player declined participation. The median squad size was 31 players (IQR = 28–33), with 15 of these taking to the field for each game. The mean age was 24.3±3.6 years (range 18–36; n=820). The duration of team enrolment ranged from 21 to 32 weeks (median=28) per season and so variable lengths of injury exposure were included in the incidence rate analysis.

Injury incidence

In total, 1030 injuries were recorded, giving a rate of 1.2 injuries per player registered. These injuries were sustained by 71% (608/1030; 95% CI 67.9% to 74.0%) of the player cohort. Ninety two per cent (948/1030) of injuries were attributed to single incidents occurring in the course of a match or training session, with the remaining 8% (82/1030) occurring insidiously, so these could not be identified as specific training or match injuries. Proportionately more injuries occurred in match play (56.9%) than training (35.1%) and the incidence rates per 1000 h match play and training exposure are illustrated in table 1. The RR for match play injury was 20.7 times that for training injury (96% CI 18.2 to 23.5).

Table 1

Injury incidence

Where injuries were sustained on the field (n=952), the playing surface conditions were predominantly dry (80%, 95% CI 77.3% to 82.4%). For injuries incurred during a competition match or structured training games, (n=702), the majority occurred in the third (32.3%, 95% CI 29.0% to 35.9%) or fourth (33.2%, 95% CI 29.8% to 36.8%) quarters of the session. In some cases of lower limb fracture and severe ligament disruption, the injured player did not return to the squad in that same season and so their time lost from play extended past the surveillance period. For those players whose injuries occurred and resolved in the same season, the median time loss for injury was 12 days, ranging from 2 to 127 days.

Playing position, age and injury

When observed injury classified by position of play was compared with expected injury, no statistically significant difference in injury proportions was noted between positions (χ2=4.93, df=3, p=0.177), but it was interesting to see that the risk of injury was lower in goalkeepers (0.62/player registered) compared with outfield players, where the ratio of injury to players enrolled ranged from 0.99 to 1.04 (table 2). Age distribution was also compared between the injured players and the total group and similarly no significant difference in proportion of injury by age group was found (χ2=4.10, df=3, p=0.25). In this case, however, a stepwise increase in injury incidence proportion was seen with ascending age. The youngest age group (18–20 years) had a ratio of injury to players of 0.63, rising to 1.16 in the over 30 age group (table 2).

Table 2

Distribution of injured players by position and age range

Main type of injury and regional distribution of injury

Injuries classified by location are presented in table 3, while table 4 illustrates the main type and tissue injured, with subclassification for upper and lower limbs. Overall, the majority of injuries occurred in the lower limbs (68.3%), with an incidence rate of 5.4/1000 h (95% CI 5.2 to 6.0). The thigh was the most common injury location (22.9%). The knee (11.3%) and ankle (9.3%) were the joints most frequently injured in the lower limb, while the foot, toes and hip had a lower incidence. Injuries to the pelvis/groin region constituted 10.3% of injuries overall. Upper limb injuries accounted for 18.6% of the total, with an incidence rate of 1.5/1000 h (95% CI 1.3 to 1.7). Collectively, injury to the distal part of the upper limb (wrist, hand, fingers and thumb) constituted 10.3% of all injuries (95% CI 8.6% to 12.3%), while the shoulder and upper arm sustained 7.1% of injuries.

Table 3

Injury details

Table 4

Main type of injury

Soft tissue injuries were foremost overall, with muscle (36.9%), ligament (17.6%), tendon (7.6%), general joint trauma (7.6%), contusions (7.5%) and haematomas (3.8%) accounting for over 8 in 10 of all injuries. Close to 9% of injuries were bone fractures (table 4). When the type of tissue injury was compared between the upper and lower limbs, a significantly (p<0.001) different distribution was evident (table 4). There were proportionately more fractures (36.1%) in the upper limb than in the lower limb (1.5%). In contrast, muscle strain constituted 45.8% of lower limb injuries but only 5.2% of injuries to the upper limb. The trunk and spine region, including ribs, sustained 8.6% injuries, while just over 4% of injuries were to the head and neck region.

Mechanism and nature of injury

Contact with another player was responsible for 38.6% of all injuries (table 3). Sprinting accounted for almost one-quarter of injuries with landing (13.7%) and turning (7.1%) the other commonly reported injury mechanisms (table 3). While the majority (80.8%) were new injuries, 17.4% were recurrent in nature. The proportions attributed to early, late and delayed recurrence can be seen in table 3. Similarly the majority were acute injuries (85.2%), with overuse mechanisms reported in 9.2%.


The results presented here provide for the first time a comprehensive injury profile of elite hurlers, through prospective surveillance over five consecutive seasons. The key findings are that the incidence rate of injury for match play is 61.75/1000 h, which is almost 21 times that of training (2.99/1000 h), mirroring previous research over one season.7 This excess risk of injury from match play is commonly seen in field sports, including Gaelic football (61.86/1000 h match vs 4.05/1000 h training)10 and soccer (26.6/1000 h match vs 4.0/1000 h training),19 reflecting the intensity of competition. However, it is notable that hurling demonstrates the highest relative risk of injury for matches versus training among these sports.

The current data show that aggregated hurling injury rates are lower than those recorded in the single year snapshot taken in 2007, 102.5/1000 h (84.4 to 123.2) for matches and 5.3 (4.2 to 6.5) for training; RR=19.5 (14.8 to 25.6),7 suggesting that incidence standardised for hours exposure has declined in subsequent years. Similarly, the incidence proportion of players is lower in the current pooled data than recorded in a single season (71% vs 82%).7 Multiyear surveillance is thus recommended to account for such variability, providing more precise population estimates with tighter CIs.

An injury rate of 1.2 per player registered was recorded and it can be extrapolated that for the average squad of 31 players, competing for the average duration of 28 weeks, 37 time loss injuries would occur. Of these 6 (17%) would be recurrent injuries, and each injury would on average necessitate 12 days absence from training or match play. These injuries would be sustained by 22 players, while 9 would escape uninjured. Outfield players are at higher risk of injury than the goalkeeper, but no notable position-related differences were seen. Injury risk increases with age, a factor noted in other sports as well.20 ,21 This may be due to greater susceptibility to injury with ageing, while the increased likelihood of an older player having prior injury due to increased sport exposure may be a confounding factor.

The development of consensus statements for injury definition, methods for expressing risk and assurance of professional diagnosis of the injury sustained,12 ,13 allow direct comparison of time loss injury incidence between sporting codes, which formerly would not have been possible due to methodological heterogeneity. We previously contrasted 1-year injury data in hurling with that available for football field sports,7 finding that injury in hurling match play was close to that of rugby union. The findings reported here, however, clearly place the full-body contact sports of rugby sevens and rugby union higher in the overall match play injury-risk hierarchy (106 and 91 injuries per 1000 h, respectively),11 ,22 with soccer match play incidence lower than hurling, as reported in the long-standing UEFA study (26.6–27.5 injuries per 1000 h).19 ,23 It is difficult to make a direct comparison with rugby league,24 Australian Football League25 and American Football League26 due to different injury definitions and reporting methods, but Gaelic football, the sister sport to hurling, has a match play incidence rate (61.86/1000 h) directly equivalent to that of hurling.10

The most closely related field sports to hurling are shinty, another Gaelic stick handling game played in Scotland, lacrosse and field hockey. Research in shinty is limited to reports of injuries presenting at the emergency department27 and there is a modest body of published descriptive epidemology in lacrosse and field hockey, with a focus on head, eye and hand injuries.2830 In men's intercollegiate lacrosse, a total incidence of 12.58 match injuries per 1000 athlete exposures (AE) was reported,28 but no comparable data for men's field hockey is available. The nature and demands of bandy and ice hockey differ considerably from hurling, however, it is interesting to make comparisons between these stick handling sports. Timpka et al31 reported 7.3 injuries/1000 player game hours in 16 elite male bandy teams over one season and one study from an elite Japanese team, where a time loss injury definition was used, reported an injury rate of 11.7 injuries per 1000 game hours over 3 seasons.32

Upper limb injury

When comparing regional injuries as proportions of overall injury sustained, more reliable contrasts between sports can be made, although it must be acknowledged that more minor injuries, which did not require absence from training or games are not counted here due to the time loss injury definition used. Thus it is not surprising that time loss incidence rate here (1.5/1000 h) is markedly lower than the 19 injuries/1000 game hours recorded in a field hockey tournament where contusions were the most common injury.33

Almost 19% (95% CI 16.3% to 21.0%) of all injuries in hurling were to the upper limb, the majority of which were fractures or sprains. The proportion of upper limb injury is comparable to National Collegiate Athletic Association (NCAA) men's lacrosse,28 where the upper extremity accounted for 26.2% of game injuries and 16.9% of practice injuries, so it would appear that the frequency of upper extremity injury is similar in these sporting codes. Bowers et al29 compared hand injury in intercollegiate women's field hockey, women's lacrosse, men's ice hockey and men's lacrosse, finding significantly higher injury incidence in the ungloved hockey players compared with gloved athletes, a finding also reported by Mukherjee in field hockey.33 Like field hockey, most hurling players play ungloved, so they too are at particular risk of contusion and laceration as well as carpal, metacarpal and phalangeal fracture as previously reported.3 ,4 With all stick contact sports, the hand is susceptible to trauma due to direct blows from the opponent's stick or ball, but repetitive gripping, combined with forces transmitted to the hand and wrist may also contribute to injury. The use of an adhesive padded grip on the hurley stick itself is common, with the objective of enhanced grip, protection of the palm from friction and provision of some shock absorbency. Hurling gloves designed with dorsal padding and either low profile or absent palmar cover are also available to protect the player, without impeding grip, but unlike lacrosse and ice hockey, these are not prescribed as mandatory protective equipment. Hurling players can often continue to compete despite minor-to-moderate hand injury and so the extent and consequences of hand trauma may be under recognised. Further evaluation of these injuries is warranted and only longer term follow-up can identify the functional sequelae.

Lower limb injury

As previously reported in hurling,6 ,7 lower limb injuries were dominant, with over two-thirds of injuries being to the lower extremity in this cohort. It is also interesting to note that over 45% of injuries were attributed to non-contact mechanisms of sprinting, landing and turning. The thigh (22.9%) was the most common site and muscle (47.5%) the most common tissue injured in the lower limb. The proportion of lower limb injury appears greater in hurling (68.3%) by comparison with approximately 48% of game and 59% of practice injuries recorded to the lower limb in men's collegiate lacrosse.28 Dick et al28 reported that ankle ligament sprains ranged between 11.3% game and 16.4% practice injuries in men's lacrosse, contrasting with 9% (95% CI 7.7% to 11.3%) injuries overall to the ankle for hurling in the current study and 9% in an earlier report.6 In hurling, the knee joint sustained just under 12% of injuries (95% CI 10.1% to 14.1%), while for men's ice hockey, it has been reported that 10% (284/2828) of all injuries involved the knee ligaments34 and that the knee was the most common game-related (13.5%) lower extremity injury.35 The data suggest that knee and ankle injuries do not vary widely between these sports.

Head, neck and trunk injury

Injury to the trunk, thoracic or lumbar spine represented 8.6% injury in hurling with most of these being soft tissue injuries and just over 4% of injuries were to the head and neck. When comparing with the NCAA statistics, it was not unexpected that the proportion of head injury would be lower in hurling than in ice hockey (9% game, 5.3% practice),34 given the chance of contact with another player, ice or boards in the ice rink. It was, however, surprising to see that hurling which allows overhead blocking and fielding of the ball, has a lower head injury incidence proportion than ladies’ field hockey (13% game injury, 3.4% practice injury),36 while lacrosse also had a higher proportion (8.6% game injury, 3.6% practice injury) of head injury than our hurling cohort.28 Further data collection over additional seasons is planned to allow systematic evaluation of subtypes of head injury and trends over time. However, it is of note that no ocular injury was recorded in this elite cohort since the introduction of mandatory helmets with faceguards in 2010.

One limitation of this study is that the results here pertain to elite teams, so the generalisation of these findings to other levels of participation should be performed with caution.

Owing to the voluntary nature of the injury reporting system, there was limited ability to accurately follow-up the more severe injuries that were ongoing at the end of the season, since not all teams re-registered for subsequent seasons. Tracking of individual players rather than teams would have facilitated complete follow-up, but due to the de-identification of players as part of the approved ethical procedures, this was not possible in this timeframe. For more complete data collection and better response rates, greater engagement with the medical teams is a priority and a mandatory injury recording system could be considered.


This study, while reporting on an indigenous game provides novel and relevant findings for the wider sports medicine community, since hurling is represented in sporting populations outside of the country of origin. The data presented here are the results of a commitment by the GAA, an amateur sporting organisation, to adopt a strategic and theory-based approach to the problem of player injury starting with measurement of incidence. This report identifies the extent and nature of injury in hurling through a long-term prospective design, providing reliable information on the incidence and type of injuries sustained. We can contrast these with other sports internationally and learn from injury prevention strategies employed in other codes. Ongoing injury surveillance, as described here, provides a platform to identify predictive factors and vulnerable subgroups of players, along with capacity to monitor changing trends in injury and to evaluate the effects of injury-prevention interventions.



  • Contributors CB was responsible for conception, design, acquisition, analysis and interpretation of the data; drafting, revision and approval of the manuscript. EO made substantial contribution to acquisition, analysis and interpretation of the data; revision and approval of the manuscript. CG made substantial contribution to the conception of the work and data analysis, revision and final approval of manuscript. JCM was responsible for conception, design, acquisition and interpretation of the data, revision and approval of the manuscript.

  • Funding This work was supported by an unrestricted educational grant from the Medical, Scientific and Player Welfare Committee of the Gaelic Athletic Association. It would not have been possible without the support of the participating teams.

  • Competing interests None.

  • Ethics approval University College Dublin, Human Subjects Research Ethics Committee and University Research Ethics Committee (LS-E-11-91-OMalley-Blake).

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

  • Data sharing statement Summary data from the injury database can be accessed through contacting the corresponding author.