Article Text
PDF
Abstract
Background Combat sports involve body contact through striking, kicking and/or throwing. They are anecdotally referred to as ‘dangerous’, yet long-term investigation into specific injury rates is yet to be explored.
Objective To describe incidence and prevalence of injury and illness within Olympic combat sports and to investigate risk of bias of prospective injury and illness research within these sports.
Methods We systematically searched literature published up until May 2016. We included prospective studies of injury/illness in elite combat athletes lasting more than 12 weeks. Risk of bias was assessed using a modified version of the Downs and Black checklist for methodological quality. Included studies were mapped to the Oxford Centre for Evidence-Based Medicine levels of evidence.
Results Nine studies were included, and most (n=6) had moderate risk of bias. Studies provided level 1/2b evidence that the most frequently injured areas were the head/face (45.8%), wrist (12.0%) and lower back (7.8%) in boxing; the lower back (10.9%), shoulder (10.2%) and knee (9.7%) in judo; the fingers (22.8%) and thigh (9.1%) in taekwondo; and the knee (24.8%), shoulder (17.8%) and head/face (16.6%) in wrestling. Heterogeneity of injury severity classifications and inconsistencies inexposure measures prevented any direct comparisons of injury severity/incidence across combat sports.
Conclusions There is currently a lack of consensus in the collection of injury/illness data, limiting the development of prevention programmes for combat sport as a whole. However, sport-specific data that identify body areas with high injury frequency can provide direction to clinicians, enabling them to focus their attention on developing pathologies in these areas. In doing so, clinicians can enhance the practical elements of their role within the integrated combat sport performance team and assist in the regular update of surveillance records.
- injury
- illness
- boxing
- judo
- martial arts
Statistics from Altmetric.com
Background
Injury and illness surveillance is the cornerstone of prevention and is the first step towards designing an prevention programme.1 2 To further extend the capabilities and knowledge of combat sport medical support, it is important to understand which injuries and illnesses are likely across these sports. Published reviews about injury/illness in combat sport exist3–9; however, reviews typically investigate one of the four combat sports rather than investigating combat sports collectively. Practitioners typically work across all four sports, and consolidated evidence will inform prevention practices.
Combat sports involve body contact through striking, kicking and/or throwing. They are anecdotally referred to as ‘dangerous’ sports,10 yet long-term investigation into the specific injury rates is absent in the literature. To collect high-quality injury and illness data, both injury and illness must be defined and reported accurately using validated methods.11 The Injury Definitions Concept Framework (IDCF)12 highlights that the use of different injury definitions may lead to the capture of only certain types of injuries. For example, if injury is defined as ‘a musculoskeletal symptom which causes an athlete to cease training or competition for more than 24 hours’ (time loss/sports incapacity definition), then the burden of chronic injuries that an athlete can train through may be missed entirely.13 14 We aimed to systematically report the incidence and prevalence of injury and illness in combat sports with respect to the definitions and study design used, risk of bias and level of evidence. Previously published reviews on injury/illness in boxing,3 15 judo,7 taekwondo4 9 and wrestling5 8 16–18 failed to address the risk of bias and failed to discuss the level of evidence about injury and illness in athletes.
Therefore, the objective of this systematic review was to describe the incidence and prevalence of injury and illness within the four Olympic combat sports as reported in prospective epidemiological and longitudinal cohort studies of elite athletes.
This review informs practitioners with three perspectives:
An overview of the common injuries and illnesses in combat sports;
A summary of the level of evidence to fulfil stage 1 (surveillance data) of the TRIPP model of prevention3;
The promotion of shared decision making around prevention priorities for the Olympic combat sports, such that training availability and injury/illness prevention can be maximised.
Methods
We followed the Preferred Reporting Items for Systematic Reviews and Meta-analysis guidelines19 when reporting this systematic review, and the protocol was prospectively registered with the PROSPERO International prospective register for systematic reviews on 2 August 2016 (registration no CRD42016027057).
Selection criteria
The lead author (SJB) applied the selection criteria, which were (1) a prospective component to data collection, (2) a collection of injury and/or illness data and (3) cohorts of elite boxing, judo, taekwondo or wrestling athletes. Owing to the university programme regularly offering scholarships for high-level athletes, studies of university level athletes were also considered elite. The exclusion criteria were (1) club, junior, cadet, youth or subelite athletes, or athletes of non-Olympic combat sports; (2) articles published before 2005 and (3) theses, review papers, case reports or case series, case–control studies and cross-sectional cohort studies. The search was limited to English language publications between 1 January 2005 and 12 April 2016, inclusive. We restricted the search date to understand injuries and illnesses that have occurred in the last decade. Owing to the frequency of rule changes in combat sports, a broader search may have returned results that have little application to today’s athletes and practitioners. All publications not in English were assessed to confirm that incorrect language indexing had not occurred and whether an English version was available through the university library. To ensure all potential studies were identified, the lead author (SJB) screened article reference lists during full-text assessment and contacted experts in the fields of combat sports and injury prevention for additional references.
Literature search
Before finalising search terms, we performed pilot searches to refine our search strategy. Additionally, this study referred to the systematic review of methodological element recommendations of the British Journal of Sports Medicine 20 and used the Assessing the Methodological Quality of Systematic Reviews tool21 prior to publication to ensure accurate reporting of study quality.
To ensure that injury and illness were considered equally, terms were combined as follows: (sport terms) AND ((injury terms) OR (illness terms)). We systematically searched PubMed (with Medline), SCOPUS (Elsevier), Web of Science and CINAHL Complete (with SportDISCUS) electronic databases using combinations of the following terms:
Sport terms: combat sport, taekwon-do, taekwondo, judo, judoka, judokas, wrestler, wrestlers, wrestling, grappling, grapplers, grappler, grapple, boxing, boxer, boxers, martial arts;
Injury terms: injury, injured, injure, injuring, injuries, trauma, tendinopathy, tendonitis, sprain, strain, rupture, fracture, stress fracture, overuse, inflammation, bursitis, hurt, wound, bruise, cork, tear, dislocation, concussion;
Illness terms: illness, infection, contagion, sickness, complaint, ailment, virus, cold, influenza, sick, medical.
Medical Subject Headings (MeSH) terms were used in PubMed searches, along with combinations of the search terms to capture a broader number of articles: athletic injury (MeSH), wrestling (MeSH), boxing (MeSH), martial arts (MeSH). Wrestling, boxing and martial arts terms were also exploded when searching the Web of Science database. A single researcher (SJB) performed the search (see online supplement 1).
Supplementary file 1
Risk of bias assessment
Articles were assessed in full using a modified version of the Downs and Black22 checklist for methodological quality by two authors (SJB and MKD) independently (online supplement 2). Articles were summarised according to data collection methodology, statistical analysis and reporting of outcomes. A third author (ST) adjudicated when consensus could not be reached. A score of ≥75% was deemed to indicate low risk of bias, 60%–75% moderate risk of bias and ≤60% high risk of bias.22 Agreement between reviewers was determined by a weighted Cohen’s kappa for two assessors to measure the inter-rater agreement for qualitative (categorical) items.
Supplementary file 2
Data extraction and synthesis/management
Data were extracted by one author (SJB) and confirmed by another (MKD). Extracted information included author(s), year of publication, study design, demographics (age, sport, level of competition, number of athletes and total team/individual seasons included), definition of injury and/or illness (mapped to the IDCF),12 main statistical approach and significant findings in terms of injury and illness characteristics, incidence and prevalence. The IDCF classification is a system for categorising injuries and illnesses according to clinical examination (injury or disease), athlete self-report (trauma or illness) and/or sports performance (incapacity or sickness).12 Studies were considered for pooling if injury/illness definitions and measures of exposure were homogenous. Where pooling was precluded, data were reported descriptively.
Level of evidence
The Oxford Centre of Evidence-based Medicine—Levels of Evidence23 was used to assess the final pool of articles. The highest level of evidence for prospective studies with good follow-up is level 1b and the lowest level is level 2b, being a prospective study with poor follow-up. The quality of follow-up was determined by the modified Downs and Black criteria22 in relation to items 9, 17 and 26. We judged ‘good follow-up’ as a combined score across these three items of at least two points.
Results
The initial key word search identified 46 712 articles for screening. Following the title and abstract screening, all duplicates, articles not related to the topic and investigations with non-human subjects or in a language other than English were excluded. One hundred and eight articles were assessed in full text, and nine were included in this systematic review (figure 1).
Flow chart of the selection process for inclusion in the systematic review.
Characteristics of included studies
There was one judo,24 three boxing,25–27 two taekwondo28 29 and two wrestling studies.30 31 There was also one multisport study that encompassed all four sports.32 Extracted data are presented in table 1. The majority of studies included in this review evaluated injuries (n=8),24–30 32 and one study reported both injury and illness.31 A variety of injury/illness definitions were used with the majority using a sports incapacity (time loss) definition25 27 32 followed by clinical examination (n=3).24 28 29 One study used a definition that included both sports incapacity and clinical examination.31 No studies used athlete-self reports, and two studies did not define injury or illness.26 30 Methodological detail and key injury/illness findings for each sport can be found in online supplements 3–6 .
Characteristics of the studies, including data collection period, statistical analysis, level of evidence, risk of bias, working definition for injury/illness and how this definition relates to the IDCF
The ICDF definitions used in the reviewed literature are displayed in table 1. Three studies (33.3%) used a definition of incapacity during sport performance,25 27 32three studies24 28 29 relied only on clinical examination for reporting injury (33.3%), one study31 focused on both clinical examination and incapacity during sport performance (11.2%) and two studies26 30 did not provide an explicit definition (22.2%). One study21 focused only on injuries to the female reproductive system in order to determine whether the sport would have a negative impact on these structures. Skin infection was explicitly studied alongside injury in one paper31; however, the authors only provided an injury definition.
Reported analysis of injury data
The purpose of all but one study was to record all injuries that met a particular working definition (displayed in table 1) across 12 weeks,32 1 year,17 20 22 31 4 years19 and 5 years.16 Collectively, the studies reported an assortment of statistical methods, with logistic regression analysis,24 Student’s t-test32 and χ2 tests24 25 27 29 32 being the most common. Descriptive statistics such as proportions were reported in four studies26 28–30 that also reported injury incidence. One study reported rate ratios and injury proportion ratios, which related to the percentage of athletes within the cohort that sustained that particular injury during the study period.31
Injury severity classifications of reviewed literature.
Incidence and severity of injury and illness
Table 2 outlines the study populations and provides a brief summary of injury information from each article. The calculation of exposure varied across the studies: training and competition time,25 30 31 training time alone24 27 or competition time alone.26 28 Because of the use of the same exposure measure, injury incidence could only be directly compared in two boxing studies and ranged between 1.5 and 3.0 injuries per year.25 27 Incidence of injury could be contrasted across a number of sports; however, some figures were calculated based on training data only and others on competition data. Injury incidence in training was 4.2/1000 hour in judo24 and 12.8/1000 hour in boxing,27 whereas the incidence per 1000 athlete exposures in training and competition was 7.3/1000 athlete exposures (AEs) in wrestling31 and 19.09–69.5/1000 AEs in taekwondo competition.28 29
Description of the study participants, level of competition, context of surveillance and incidence of injury of the included studies
Injury severity classifications differed, complicating the comparison between injury severities across the combat sports. Figure 2 displays the classification of injury severity in the reviewed articles.
Risk of bias
Overall, item agreement on the risk of bias between the initial assessors was high, κw=0.81. Adjudication by a third reviewer was required to reach consensus for 16 items (6.8%). The modified Downs and Black scores ranged from 14 to 22, with a maximum possible score of 24. Percentage scores for each article and individual item scores are provided in online supplement 2.
Best evidence synthesis
Four studies, three with moderate25–27 and one with high risk of bias,33 provided level 1b/2b evidence that the head/face, wrist and lower back are the most commonly injured body sites in boxing. In both national and elite boxers, head/face injuries accounted for almost half of all injuries.27 32 Analogous injury definitions allowed injury incidence to be directly compared across boxing studies (see online supplement 3).
Supplementary file 3
Injuries in judo were investigated two studies, with low risk of bias in one study24 and high in another.32 Together, extracted information from these studies provide level 1b/2b evidence that lower back, shoulder and knee injuries have the highest incidence and prevalence in judo. Men had higher injury rates than women. Moreover, men generally sustained a greater total number of upper body injuries, whereas women sustained a greater total number of lower body injuries (online supplement 4).25 30
Supplementary file 4
Three studies investigated injury in taekwondo, with moderate risk of bias in two studies28 29 and high in the third.32 These studies provides level 1b/2b evidence that thigh and finger injuries have the highest incidence and prevalence in taekwondo, with finger injuries being of greater concern during competition (online supplement 5).28 29 32
Supplementary file 5
Three wrestling studies, two with moderate30 31 and one with high risk of bias,32 provided level 1b/2b evidence that the knee, shoulder and head are areas of concern for injury27 28 32 and that skin infection is an illness concern.28 Skin infection has been highlighted as a serious issue in wrestling, and despite a lower incidence than injury, it is often more severe and requires a greater amount of time off training (online supplement 6).31
Supplementary file 6
Table 3 highlights bodily areas that are associated with relatively high injury frequency in combat sports. Due to the exposure measure differing across sports, injury incidence was not able to be compared.
Areas of relatively high injury frequency* in combat sport
Discussion
The impact of negative health events, such as injury and illness, on athletic performance has been investigated previously,34 and avoiding injury/illness should be considered a multidisciplinary task, shared between the clinician, coach, sport scientist and athlete.
This study highlights that there is considerable inconsistency across the literature with regard to how injury data in combat sports have been collected and reported in the last decade. This suggests a need for consensus in methodology when investigating injury and illness in combat sport.
Only one prospective epidemiological study with low risk of bias on injury in combat sport currently exists.24 All other prospective studies were rated as being of moderate or high risk of bias, with inadequate descriptions of the studied athlete populations and poor follow-up. The review also highlights that most studies did not report injury severity and/or related functional capacity, for example, return to competition. This is an important omission, when combat sports are considered ‘dangerous’. Owing to the highly heterogenic approaches to measuring exposure, use of varied definitions of injury/illness and different contexts of surveillance (training versus competition), it was not possible to pool results.
Injury risk in combat sports
This review demonstrates that injuries in combat sport are minor compared with injuries in team sports. Studies reporting on concussion in competition show that rates were between 4.1 and 4.5/1000 hours in rugby union33 35 and 6.5/1000 hours in ice hockey.36 In comparison, concussion rates are low in boxing, at 0.53/1000 hours.25 Currently, the rate of catastrophic injury across combat sport is unable to be compared and, as such, should be an area of focus for future research. Like every sport, combat sports have a set of inherent risks and common injuries, and provided these are understood when a participant takes up a combat sport, practitioners can assist mitigating these risks with effective planning and monitoring.
Practical recommendations to help practitioners incorporate this evidence in an integrated performance system
Systems thinking frameworks are beginning to be overlaid on to injury/illness and performance models.34 Systems thinking, as applied to athletic performance, recognises that performance cannot be considered separate to health, and health status directly impacts performance. By focusing on the early warning signs of certain health problems, clinicians will be well placed to improve athlete care.
Clinicians have an opportunity to collect high-quality data
Quality injury and illness surveillance is a cornerstone of injury and illness prevention. Our systematic review highlights that methodological quality is lacking in much of the published literature on combat sports and makes suggestions to improve future data collection. Within a contemporary model of athlete care, it is acknowledged that clinicians play an important role in the collection of high-quality surveillance data. Clinicians have closer, more regular contact with athletes and are often the first point of contact when a health or injury problem arises. When clinicians use methods that promote the capture of high-quality data, a strong foundation is set for the development and implementation of effective injury and illness prevention programmes. Currently, international sporting federations favour prevention ahead of surveillance37; however, prevention cannot be targeted and effective without high-quality surveillance data, generated from the ground up by clinicians and researchers in partnership.
Suggestions for clinician-led surveillance data collection
The most common study design used the direct collection of injuries by medical team members.24–26 31 Other studies used a mix of interviews and questionnaires, all requiring some kind of recall on the athlete’s part and potentially introducing recall bias.27–30 32 Two studies administered questionnaires at competition,28 29 introducing survival bias, as the athletes who were already experiencing recent or severe injury or illness would have been unlikely to attend competition due to medical limitations. To overcome recall and survival bias, epidemiological studies should span more than 1 year using consistent and regular prospective surveillance across training and competition. Ideally, a study on Olympic combat sports would cover a full 4-year Olympic cycle to evaluate the impact of injury and illness within the competitive cycle and lead-up to competition. To report injuries/illnesses at the time they occur, a validated questionnaire14 or a training log alongside medical team reports could be used. This would allow for tracking and management of smaller injuries/illnesses that affect the quality of training, alongside the larger injuries/illnesses associated with time loss.
In general, most injury definitions focused on sports incapacity (time loss from training/competition), giving rise to a number of issues. First, such injury definitions discount any injury or illness where an athlete did not miss any training or competition, a limitation well acknowledged in the literature.14 38 39 Second, an athlete may self-manage their injuries/illnesses, which could impact the quality of training but not necessarily reduce its amount. In cases of self-managed injuries/illness, it is plausible that the supporting medical team is not aware of such events, and subsequently no record is created. To improve the capture of injuries/illnesses, a broader definition could be used relating to ‘physical complaints’14 rather than musculoskeletal conditions that prevent the athlete from training or competition. These data may provide a system of early detection of health problems and in doing so, assist in both surveillance records and practical elements of their role within the integrated performance team.
Categories of severity have been included in injury/illness consensus statements across a number of sports.14 38 39 The categorisation of severity can be a simple way to compare the time–loss effects of health problems across sports; however, most studies did not describe a severity measure,25–27 29–31 and there was no common method where severity was described.24 28 32 It is recommended that future research leverage current literature34 38 39 and routinely report injury severity measures.
Exposures used to calculate incidence and prevalence are number of matches/training sessions and hours of training and/or competition. Unfortunately due to the disparity in the measures of exposure, incidence and prevalence cannot be directly compared across the combat sports on the basis of the currently existing literature. This problem could be addressed in future research by using time as a measure of exposure (per 1000 hours) in both training and competition.
Strengths and limitations of this study
The strength of this study was the sensitive search strategy (>45 000 articles assessed), strict inclusion criteria and the assessment of risk of bias and level of evidence. This allowed the comparisons of studies to include risk of bias assessment and level of evidence for their results. One low-quality study with high risk of bias study was found, which published data after rule changes took effect.32 The comparison between the rates prior to and following these changes is beyond the scope of this review.
A limitation is that only English articles were included. Many European and Asian countries have long combat sports and consistent Olympic success; therefore, future reviews should attempt to include studies from these regions. A single investigator performed the search strategy and applied the selection criteria. However, given the extensive list of identified articles that were screened, the risk of missing a relevant study is low, particularly given that we undertook citation tracking and that expert advice was sought. The heterogeneity of the articles meant that it was not possible to perform a meta-analysis; however, qualitative review is still a valuable and novel contribution to the literature and highlights that a more uniform approach to data collection and analysis is required before cross-sport comparisons can be made.
Conclusion
Musculoskeletal injury is a common issue across boxing, judo, wrestling and taekwondo. Investigating the factors associated with musculoskeletal injury and developing prevention programmes targeting the common areas would have far-reaching implications across these combat sports. In the interim, sport-specific data that identifies body areas with high injury frequency can provide direction to clinicians, enabling them to focus their attention on any developing pathologies in these areas.
Clinician-led data collection can assist in surveillance and the early detection of health problems. In doing so, the practical elements of the clinician’s role within an integrated performance team are modernised.
What is already known on this subject?
Musculoskeletal injury is a concern across judo, taekwondo, wrestling and boxing.
What are the new findings?
In boxing, the injuries with the highest incidence were head/face, wrist and lower back.
In judo, injuries mostly occurred in the lower back, shoulder and knee.
In taekwondo, the thigh and fingers were the most frequently injured sites.
In wrestling, knee, shoulder and head injuries and skin infections have the highest incidence.
The lack of consensus in the collection of injury/illness data limits the development of prevention programme for combat sport.
References
Footnotes
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.