Article Text
PDF
Abstract
Objective The physical demands of professional tennis combined with high training/match loads can contribute to musculoskeletal injury. The objectives of this study were to (1) describe the type, location and severity of injuries sustained during a 12-month tennis season in a cohort of professional female tennis players on the Women’s Tennis Association (WTA) tour and (2) prospectively investigate associations between training/match loads and injury.
Methods 52 WTA players competing at the Australian Open (2015) consented to participate. Injuries reported to WTA medical staff were classified using tennis-specific guidelines. Individual match exposure data were collected for all matches played at international level in 2015 and expressed per 1000 hours of WTA competition matchplay (MP) and 1000 match exposures (MEs). Variables associated with the number of injuries in the season and loss of time from competition were identified with regression analysis.
Results The injury incidence rate (IR) was 56.6 (95% CI: 49.5 to 64.6) per 1000 hours of MP or 62.7 (95% CI: 54.8 to 71.6) per 1000 MEs, although the IR of injuries resulting in loss of time from competition was lower (12.8 per 1000 hours of MP, 92 injuries/100 players). Lower limb (51%) and muscle/tendon (50%) injuries were the most common site and type of injury. Common specific injury site subcategories were the thigh, shoulder/clavicle, ankle and knee in order of frequency. Various measures of match load were significantly associated with injury.
Conclusion This study prospectively analysed injury profiles, including severity across an entire season of professional tennis, and investigated the relationship between training/match loads and injury. These data may help medical professionals develop injury risk identification and prevention programmes.
- tennis
- risk factor
- injury
- load
- epidemiology
Statistics from Altmetric.com
Introduction
Tennis is a global sport with over 75 million participants1 in more than 200 countries.2 Professional women’s tennis is played at 59 tournaments across 36 countries and on varying surface types.3 Like many sports, playing professional tennis poses a risk of musculoskeletal injury due to its physical demands and high training/match loads. The reported injury incidence in tennis players varies in the literature. A review by Pluim et al4 reported incidence rates (IRs) ranging from 0.04 to 3.0 injuries per 1000 playing hours in players of all ages and participation levels.
Few studies have described injury profiles in elite tennis. The injury incidence in professional athletes within single Grand Slam tournaments (US Open and Wimbledon) has been published in longitudinal retrospective cohort studies.5 6 To the best of our knowledge, no prospective data on injury IRs in women’s professional tennis have been published previously. To allow development of evidence-based injury prevention programmes, injury surveillance is an important preceding step.7
Previous epidemiological studies of tennis players have identified that injuries to the lower limb are more common than injuries to the upper limb, trunk or head/neck.5 6 8–12 In these studies, some variability exists in the type and location of injuries, dependent on the level of tennis being played. In professional tennis, little research reports the location or severity of injury.
The objective of this study was to report on the incidence of injuries reported to the Women’s Tennis Association (WTA) medical professionals (physiotherapists and tournament physicians) during a season. The study aimed to describe the type, location and severity of injuries in professional female tennis athletes and investigate relationships between training/match loads and injury which to our knowledge has not been published previously.
Methods
Participants
WTA players playing in the singles or doubles event in qualifying or main draw in the first Grand Slam tennis tournament in 2015 were invited to participate in the study via flyer advertisement in places frequented by athletes, for example, player locker rooms. Players volunteered to participate as part of a pilot study to investigate a tennis screening tool.13 Convenience sampling from approximately 200 female players was conducted to obtain sufficient participants to power the pilot study. Junior (age <18 years) or wheelchair tournament athletes were excluded from this study. Players provided informed consent. Due to legal and ethical restrictions to protect participant anonymity, ranking was categorised (1–20, 21–50, 51–100, 101–150, 151–200, 200+) for singles and doubles.
Procedure
The study methods used the framework and recommendations outlined in a consensus statement on epidemiological studies of medical conditions in tennis.14 Recommendations included reporting injury IRs as the number of injuries/1000 hours played and recording match and training exposure separately.
Injury definition and collection
All musculoskeletal injuries sustained by players as a result of tennis match play or training in the 2015 season were recorded in the WTA medical documentation system. An injury was defined as any condition that a player presented with to a WTA affiliated health professional (medical practitioner or physiotherapist/athletic trainer). Injuries were assessed and diagnosed by WTA medical personnel and entered into the WTA injury database; Athlete Health Medical System. Non-tennis-related injuries/illnesses were excluded. Data were extracted and categorised by one investigator.
Exposure
Two methods of calculating injury rate were performed to allow comparison with previous studies. First, match data were collected by SAP for all international-level matches played at either a WTA or International Tennis Federation (ITF) event and expressed per 1000 match exposures (MEs). Injury IR was also calculated per 1000 hours of match play (MP) at WTA or Grand Slam tournament level only. Data on match time at ITF events were not available. Due to the individual nature of tennis, collecting prospective training volume data from players in multiple locations is challenging. Instead, we collected self-reported training volume by asking the question "in the last two weeks please detail your participation in training activities’’.
Injury classification
The consensus statement14 was used to classify injuries according to type and location. The type of injury was classified according to (1) bone, (2) joint (non-bone) and ligament, (3) muscle and tendon, (4) skin, (5) central/peripheral nervous system or (6) other. The injury location was categorised according to the groupings (1) head and neck, (2) upper limb, (3) trunk, (4) lower limb and (5) other. Injury type and location categories were further subdivided into subcategories. For example, under the injury type joint (non-bone) and ligament, subcategories of dislocation/subluxation/instability, ligament injury, lesion of meniscus/articular cartilage and synovitis were identified.
Severity
Any injury that required retirement or medical withdrawal from an event either onsite or offsite was recorded. Severity of injury was then categorised into three groups according to the loss of time from competition. Mild was defined as no loss of time from competition, moderate as withdrew/retired from a tournament but was available to play at the next scheduled event and severe as retired/withdrew, missing greater than one tournament.
Data analyses
Injury data are reported as mean (SD), median (IQR) or frequencies. Injuries are presented as IR per 1000 MEs and per 1000 hours of MP and injuries per 100 players. Poisson regression was used to determine rate ratios (RRs) to compare the rate of injury at each classified type and location and to identify variables associated with the number of injuries sustained in a season (2015). Logistic regression was used to determine variables associated with sustaining a severe injury during the season that resulted in loss of time from competition with match volume (hours per season) and past history of severe injury in the last 12 months included in a multivariable regression model. ORs, regression coefficients and 95% CIs were reported. Data were analysed using IBM SPSS Statistics for Windows, V.24.0. Armonk, New York: IBM Corp. Statistical significance was assigned when p<0.05.
Results
Players (n=52) competed in a total of 3447 international matches for a total of 3819 hours in the 2015 season. All players (n=52) were followed throughout the entire 2015 season with no dropouts. Ranking was widely distributed across all categories. The most frequent ranking category for singles and doubles was 51–100 (both n=17) (table 1).
Player characteristics
During 2015, a total of 216 injuries were recorded (415 injuries/100 players). The overall injury IR was 56.6 per 1000 hours of MP (95% CI: 49.5 to 64.6) or 62.7 per 1000 MEs (95% CI: 54.8 to 71.6). Players sustained an average of four injuries per season (range 0–12). Ninety-six per cent reported at least one injury during the 2015 season. More than half (55%) sustained an injury resulting in loss of time from competition (retirement/medical withdrawal), and nearly a quarter of players reported more than one injury resulting in loss of time from competition.
Injury location
The lower limb was the most common site of injury (41.3 per 1000 MEs, 95% CI: 28.6 to 54.0) followed by the upper limb (21.4 per 1000 MEs, 95% CI: 11.1 to 31.6) and the trunk (16.1 per 1000 MEs, 95% CI: 7.5 to 24.7). There was over twice the rate of lower limb injuries compared with upper limb injuries (RR=2.2, 95% CI: 1.5 to 3.2) and nearly three times the rate of lower limb injuries compared with trunk injuries (RR=2.8, 95% CI: 2.0 to 3.8) (table 2). Common injury sites were the thigh (17.1% of injuries), shoulder/clavicle (9.3%), knee (8.3%) and ankle (7.9%) (table 2).
Incidence data for injury location (category and subcategory) for study population (ordered according to decreasing frequency of % of cases)
Injury type
Collectively, 50% of recorded injuries involved muscle or tendon (muscle injuries 24.3 per 1000 MEs (95% CI: 15.2 to 33.5), tendon injuries 18.7 per 1000 MEs (95% CI: 8.9 to 28.5)). Thirty-eight per cent of injuries involved joints (33.0 per 1000 MEs, 95% CI: 18.9 to 47.0) and 5% involved bone (3.2 per 1000 MEs, 95% CI: 1.1 to 5.3). There were nearly twice the rate of muscle injuries to tendon injuries (RR=1.9, 95% CI: 1.2 to 2.8). Joint synovitis (17.6 per 1000 MEs, 95% CI: 4.8 to 30.5) and ligament injuries (9.3 per 1000 MEs, 95% CI: 2.7 to 15.8) were also commonly recorded (table 3).
Incidence data for injury type (category and subcategory) for study population (ordered according to decreasing frequency of % of cases)
Severity
Nearly a quarter (22%) of injuries resulted in loss of time from competition. Compared with the overall IR (62.7 per 1000 MEs), the IR of injuries resulting in loss of time from competition was lower (12.8 per 1000 hours of MP (95% CI: 9.7 to 17.0), 14.2 per 1000 MEs (95% CI: 10.8 to 18.8), 92 injuries/100 players. The most common site for severe injuries was the upper limb (46% of severe injuries) compared with the lower limb (36%); however, the knee was the most common location of severe injury (comprising 43% of severe joint injuries). In contrast, in the mild category, the rate of lower limb injuries was twice the rate of upper limb injuries (56.6% vs 19.6%, RR=2.2, 95% CI: 1.5 to 3.2).
When categorised according to the type of injury, tendon (32%), joint (32%) and bone injuries (18%) were the most commonly recorded severe injuries. Tendon injuries occurred at a similar frequency in the shoulder, elbow, wrist and knee. Joint injuries (41%) were the most common type of mild injury. In contrast, muscle injuries had a higher IR in mild injuries than severe injuries (20.5 per 1000 MEs vs 0.4 per 1000 MEs). When combining injury location and subcategory type to investigate the diagnosis resulting in greatest loss of time from competition (severe/moderate injury), abdominal muscle strain occurred most frequently (8% of injuries resulting in loss of time from competition), followed by knee ligament sprain, thigh muscle strain, wrist bony injury, shoulder tendinopathy and non-specific low back pain (all 6% of injuries resulting in loss of time from competition). Individual injury location and type IRs are presented in table 4.
Incidence and descriptive data for severity of injury described by injury location and type for study population
Variables associated with the number of injuries per season
Variables with a significant association with the number of injuries per season were singles ranking, total number of matches, total number of singles matches, match volume, singles match volume and doubles match volume (table 5). Greater ranking, number of matches and match hours per season were associated with a greater number of injuries per season.
Variables associated with the number of injuries per season (2015)
Variables associated with loss of time from competition due to injury
Variables significantly associated with sustaining a severe/moderate injury during the season that resulted in loss of time from competition (ie, retirement or withdrawal from one or more matches) included singles ranking, total number of singles and doubles matches, match volume (hours per season) and singles match volume (hours per season). For each additional hour of match play per season, in 2015, the odds of having an injury resulting in loss of time from competition increased by 2% (OR 1.02, adjusted for past history of severe injury in the last 12 months: table 6).
Variables associated with loss of time from competition due to injury (2015 season)
Discussion
This study prospectively analysed injury profiles, including severity across an entire season of professional tennis, and investigated the relationship between training/match loads and injury. We found that the thigh was the most commonly injured location, although a majority of these injuries were mild and did not result in loss of time from competition. Knee, shoulder and tendon injuries were the most common severe injuries, with abdominal muscle strain being the most frequent diagnosis resulting in loss of time from competition. Various measures of match load, including match volume (hours) and total number of matches, were significantly associated with the number of injuries and loss of time from competition.
Previous studies have focused largely on competition at a junior elite level10 11 15–19 or at singular events at professional level.5 6 Due to the global and individual nature of the sport, detailed data collection for epidemiology studies across an entire season of professional tennis is challenging. This study is an important step in describing detailed injury profile in elite female professional tennis players inclusive of severity.
Injury IR
Injury IR across all levels of tennis has previously been reported as between 0.04 injuries/1000 hours and 3.0 injuries/1000 hours.4 At the elite junior level, injury IRs vary from 6.9 to 21.5 per 1000 MEs compared with 0.06 per 1000 MEs at the amateur junior level.10 11 19 In comparison, our study reported higher IRs of 56.6 per 1000 hours of MP and 62.7 per 1000 MEs. Our higher IR is likely attributable to differences in participant population (ie, junior vs senior athletes) and definitions of injury, severity and incidence.4
At the professional level, two longitudinal retrospective cohort studies at the US Open and Wimbledon reported injury IRs in female professional athletes as 40.6 per 1000 MEs and 23.4 per 1000 sets played, respectively.5 6 These findings are again lower than those reported in our study. This difference may be attributed to differences in injury reporting. Studies conducted by Sell and McCurdie defined injury according to those injuries reported to tournament physicians. On the WTA tour, a number of injuries that do not require loss of time from competition or warrant the opinion of a physician may be managed entirely by the expert physiotherapists and hence injury rates in other studies may be under-reported. Other potential influences for the differences in reported injury rates may be reporting across an entire season with environmental challenges; court surface and match volume are likely to affect injury incidence compared with injury rate reported at any single event.
Injury IR described across an entire professional international-level tennis season has been reported in one study as 40.03 per 1000 MEs.12 In this study, IR was determined using publicly available website data obtained on retirement and withdrawal rates. While loss of time from competition is an important factor, the incidence of injuries that allows players to continue to compete but may impact on performance ability is also worth determining. Lower IRs for loss of time from competition in our study are likely due to methods of data collection, that is, as reported to health professionals versus online data.
Injury severity/location
Previous reviews on injury location have suggested that lower limb injuries are more common (31–67%) than upper limb (20–49%) and trunk (3–21%) injuries across all levels of tennis.4 8 Our data are consistent with these reviews.
Previous studies have found the wrist and lumbar spine/buttock to have the highest injury IR in tennis, in contrast to our study where thigh injuries were most common.5 6 This discrepancy is likely attributed to the demands placed on the body by different court surfaces in comparison with recording data at only a single event. When examining injury IR across an entire season, thigh injuries have previously been reported as frequent in other professional sports.20 21 Kryger et al12 reported thigh injuries as the most frequently reported injury location for loss of time from competition in professional tennis (IR 4.44 per 1000 MEs). While mild thigh injuries were frequent, we also found that when injury type and location data were combined, both thigh and abdominal muscle strains were found to be a common injury resulting in loss of time from competition.
Injuries sustained to the upper limbs were more severe than other regions, although the knee (22%) was the most frequent isolated site for severe injuries—a majority of these were ligament injuries. Shoulder tendinopathy was the most common upper limb injury resulting in loss of time from competition. Beyond this, upper limb injuries varied widely with no significant trends identified. Knee and shoulder injuries have previously been reported in the literature as the most frequent injuries resulting in loss of time from competition among all sports at the Summer Olympic Games.20 In professional tennis, severe shoulder injuries requiring arthroscopic shoulder surgery have previously reported a prolonged and often incomplete return to previous level of performance post-surgery.22
Variables associated with injury and loss of time from competition
In our study, 22% of injuries resulted in loss of time from competition and therefore understanding the predictors of these injuries is important for athletes and medical staff. Greater number of singles matches played and greater match volume (hours per season) were significantly associated with greater odds of sustaining an injury resulting in loss of time from competition. Since some players did not play doubles matches (and vice versa), and our training volume data were self-reported, match volume (hours per season) was deemed to be the most robust measure of player load over the season. Match volume remained significantly associated with loss of time from competition when adjusted for past history of severe injury in the previous 12 months (included because it is a possible risk factor for future injury). Our findings indicate that a relatively small increase in match volume per season may be associated with greater risk of injury (eg, one additional hour of match play was associated with 2% greater odds of injury). This finding is important because the number of tournaments entered per season (ie, match volume) can be modified by players and their team and is therefore a variable that should be considered when evaluating a player’s overall injury risk across a tennis season. Further well-powered studies are needed to quantify this relationship before guidelines for players, coaches and sports medicine staff can be established.
Limitations
A limitation of this study is that only injuries reported to WTA medical professionals were recorded. A small percentage of WTA players have their own medical staff and do not use the WTA medical team. However, most players with private therapists also use WTA medical personnel due to the continuity of expert care provided across events. All injuries requiring withdrawal or retirement from an event must be reported to WTA medical staff. Therefore, the total burden of injury is unlikely to affect our reported rate of moderate and severe injuries, but may be higher than what we reported for mild and overall IRs. The prospective nature of this study means that players were made aware their injuries would be recorded and may have been more diligent in injury reporting. By keeping methods of reporting consistent with standard WTA procedures, we aimed to minimise this. Interestingly, past history of severe injury and training volume were not significantly associated with injury in this study. Both variables were self-reported via questionnaire and hence may have led to under-reporting. Further studies with utilisation of injury databases and prospective collection of training volume are required to allow reporting of training volume IRs and time loss from training due to injury. Permission for this study was collected as part of a pilot study investigating musculoskeletal risk factors of injury in tennis and involved a convenience sample of players; hence, our sample is not inclusive of the entire population. Furthermore, in some subcategories of injury, incidence was low resulting in limited statistical power; hence, it may be difficult to generalise IRs from these items to the broader population.
Conclusion
This study provides a detailed injury profile of elite professional women’s tennis in season and investigates relationships between training/match loads and injury. Injury IR was found to be higher than previously reported. Lower limb injuries were twice as common as injuries to other areas. Thigh injuries had the highest IR; however, shoulder and knee injuries resulted in greater time lost from competition. Match volume (hours per season) was significantly associated with the number of injuries and loss of time from competition.
What are the findings?
This study prospectively analysed injury profiles across an entire season of professional tennis.
Injury incidence rates were higher than previously reported in the literature with some variability in location of injury.
Classification of injuries according to severity in professional tennis is reported.
A relationship between match volume and injury was established.
How might it impact on clinical practice in the future?
This study may help guide tennis medical professionals in targeted development of injury risk identification and prevention programmes, an area where there is currently a paucity of literature.
Acknowledgments
Thanks are given to the Women’s Tennis Association (WTA) Sports Science and Medicine staff without whom this work would not have been possible and the Australian Open tournament staff for their permission to allow data collection to commence at the event. The authors are grateful to the staff at SAP, a partner of the WTA, for providing match volume data.
References
Footnotes
Handling editor Karim M Khan
Contributors JGD and BS devised the idea and study plan. JGD, BS and CMG developed the study design and ethics application. JGD and BS were responsible for data collection. JGD and LGP were responsible for statistical analysis with early stage assistance from CMG. JGD drafted the manuscript. All authors contributed to editing the original manuscript. JGD and LGP drafted the rebuttal, performed additional statistical analysis and revised the final draft of the manuscript.
Competing interests None declared.
Ethics approval Approval for the study was obtained from the Women’s Tennis Association and Monash University human ethics research committee.
Provenance and peer review Not commissioned; externally peer reviewed.