Youth sports injury is a public health concern, as it has detrimental effects on the health and well-being of young athletes.1 Increasingly competitive youth sports around the world require adolescents to undertake more prolonged and intensive training programmes2 and increased numbers of playing hours. Increased playing is mirrored by an increased risk of traumatic injury or overuse musculoskeletal injury by young sports participants.1

The knee joint is reported to be the most common joint injured by young sports participants.3 The complexity of the knee joint structure and the multidirectional forces imposed on the knee joint during sporting activities may explain why knee injuries are often more severe than injuries to other body regions.4 Knee injuries therefore often require surgery or extensive rehabilitation before the knee functions at a pre-injury level.5 The high economic cost of knee injuries to the individual and society is of particular concern in developing countries, where limited access to appropriate services and high healthcare costs are major barriers to appropriate management of youth sports knee injuries.3

Adolescent knee injuries are of further concern as the knee joint has been identified as having the highest rate of early development of osteoarthritis.6 Such degenerative joint changes require long-term management, and lead to disability that impedes future participation in sports, daily activities and occupational requirements.7 The potential loss of ability to participate in regular physical activity after injury is alarming, because physical inactivity is one of the major risk factors associated with systemic disease, disability and/or death worldwide.8

Prevention of sports injuries, to the knee in particular, is thus promoted as a cornerstone in the management of sports injuries. There is however a dearth of literature reporting on the prevalence of sports-related injuries to adolescents, or promoting successful strategies to prevent knee injuries among adolescents.9 Successful preventative strategies should stem from integrated basic science and epidemiological research, in order to understand the mechanisms associated with adolescent musculoskeletal sports injuries.9 A search of 10 databases (PubMed, CINAHL, SportsDiscus, Proquest Medical Library, Web of Science, PsycINFO, ProQuest Dissertations, ScienceDirect, PEDro and Cochrane Library) has identified that no systematic review of adolescent sports knee injury prevalence or causation has yet been published. This indicates a gap in the understanding of public health and global perspectives of adolescent knee injuries sustained during sporting activities. The aim of this review is thus to ascertain the extent of the prevalence of knee injuries, and factors associated with sports and recreational knee injuries among active adolescents.

The objectives of this review were to:

• Synthesise the epidemiological research reporting on the prevalence of knee injuries among active adolescents (aged 13–19 years).

• Determine the risk factors for knee injury identified by these epidemiological studies and identify modifiable risk factors associated with injury such as injury mechanism, playing conditions, etc.

• Critically appraise the methodological quality of the selected publications.

• Synthesise risk factor findings into recommendations to reduce adolescent knee sporting injuries.

METHODOLOGY

A systematic review, using transparent and robust principles, was conducted to ascertain a broad perspective of the knee injuries sustained by active adolescents around the world. Adolescents are defined as individuals aged between 13 and 19 years,10 and activity is defined as formal sports such as soccer and basketball, recreational activities such as roller-skating, and cultural activities such as ballet.11

Literature search strategy

A systematic search of 10 databases was conducted for eligible articles published between the years January 1980 and March 2006. These databases were PubMed, CINAHL, SportsDiscuss, Proquest Medical Library, Web of Science, PsycINFO, ProQuest Dissertations, ScienceDirect, PEDro and Cochrane Library. Each database has its own indexing terms and thus search strategies were developed for each database. The primary search terms included variations on the nomenclature of “adolescents”, “knee injury”, “sports” and “recreational activity” and all variations of these terms were searched. The medical subject headings used in PubMed included “knee injuries” and “recreation”.

The search strategy also included secondary searching (pearling) of the reference lists of included articles, as well as manual searching of journals such as the South African Sports Medicine Journal and African Journal of Sports Dance and Recreation, which are not indexed in any of the databases searched.

Inclusion criteria

The search approaches identified publications in the English language, epidemiological studies, and on adolescents (subjects aged 13–19 years). Cross-sectional, retrospective and prospective descriptive epidemiological studies were included. Studies were only included if they reported on both the prevalence of and risk factors associated with adolescent knee injuries. Studies not indicating the specific age of the sample, but which mentioned high school participants, were also included in the review, because the de facto age of high school students encompasses 13–19 years.

Exclusion criteria

Experimental studies, case studies and anecdotal reports were excluded because they did not report on the prevalence or risk factors for adolescent knee injuries. Studies not reporting on individuals aged 13–19 years, studies not published in English, and studies not available in full text from any library source were excluded.

Validity assessment

Two reviewers (QL, JM) independently assessed all possible studies at title and abstract stage against the inclusion criteria (age group, epidemiological study design, language). Uncertainty of study inclusion was discussed with a third reviewer (KG). The full text versions of all the relevant publications were then reviewed and inclusion/exclusion confirmed. The eligible studies were then included in the review.

Methodological appraisal

An assessment of methodological rigour was conducted primarily by one researcher (QL). A random sample of three studies was evaluated by a second reviewer (KG) to ensure that the methodological quality criteria were being reliably applied. There was no uncertainty in quality evaluation between these reviewers. The sports epidemiological appraisal tool developed specifically for this study was used to critically appraise the methodological soundness of the included studies.12 This appraisal tool was developed using elements of existing as well as recognised critical appraisal instruments (Liddle13 as adapted by the NHMRC, Ellwood14 and Crombie15). This instrument contained 12 key criteria under the headings of study design, analysis, and overall assessment of the injury event. This instrument was worded specifically for sports epidemiology and included questions about sports exposure, confounders, and outcome measures. Literature that reported different types of information on sports exposure, confounders and outcome measures was sourced and used as references, to ensure that the questions in the critical appraisal instrument comprehensively covered the range of reporting approaches/styles. The total score of the tool is 37 points. A score of 0 indicates very poor methodological quality whereas a score of 37 indicates the highest methodological quality possible. Table 1 provides a summary of the critical appraisal instrument.

Study classification

Details of included articles were collated using a MS Excel database with the key headings outlined in table 2. These headings were used to clarify and describe key elements of each study for synthesis and comparison purposes.

RESULTS

A summary of the results of the search is provided in figure 1, which outlines the number of “hits” from the initial search, the loss to exclusion criteria, and the number that was retained in the search. Overall, a total of 19 articles were considered appropriate for this review.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Search results.

Basic description of the studies included in the review

The 19 eligible studies for this review were mostly (90%) conducted in developed countries. Nearly 50% of the studies were conducted in the USA. The sample sizes ranged from 19 to 19 728. More than 60% of the studies reviewed to ascertain injury rate included male and female subjects. Schools provided the setting for 52% of the studies included in this review. Sports clubs provided the study setting in 20% of the studies included in the review, and prospective designs reflected 42% of studies reviewed. Details of the studies included in this review are tabulated in table 3.

Injury definitions

Six different injury definitions were used in the 19 included studies (see table 4). Factors that were differently applied in injury definitions included the loss of time of play as a result of the injury, medical care for the injury sustained, location of the injury (head or face), and whether the definition included both traumatic and overuse injuries.

Knee injury prevalence

The way in which injury prevalence was reported varied considerably between the included studies (see table 4). However, data for knee injuries that could be extracted from the included studies facilitated comparison of the percentage of knee injuries with the total number of injuries sustained. This information is presented in figure 2. Actual comparison of injury percentage across studies is complex, as the total number of injuries varies between the studies, as well as the time period over which knee injuries were captured (see table 5). Knee injury prevalence ranges between 10% and 25%, with more recent studies reporting higher percentages. This suggests either better injury definitions in more recent studies, or higher rates of injury.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Knee injury prevalence according to sport.

Three of the studies included in the review provided information on the risk of knee injuries per hours of play. Insufficient information was provided in the other studies to calculate the risk of knee injury per playing hours. The findings from the three relevant studies are summarised in table 5.

Relation between knee injuries and gender

Ten studies provided information on knee injury prevalence (knee injuries as a percentage of all injuries) for males and females. Females appeared to be more likely to sustain a knee injury compared with males (see fig 3). Figure 3 demonstrates that only three studies1 3 24 provided sufficient data to suggest that the boys sustained more injuries than girls.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3 Knee injury prevalence.

Knee injury prevalence related to type of sport

Considering knee injury prevalence according to the type of sport, the review facilitates identification of four papers in which sports-specific knee injury data were provided. This allowed extraction of knee injury prevalence data in soccer and basketball. Figure 4 demonstrates that both basketball and soccer pose a risk to knee injuries and young soccer players appear to have a higher percentage of knee injuries compared with all other injuries sustained. However, the differences in injury definition, total number of injuries and time period of data collection constrained generalisation of this data to all young people.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4 Methodological appraisal of findings.

Factors associated with knee injuries

Table 7 illustrates that 11 of the 19 included studies (58%) reported on the type of knee injuries sustained. The classification and nomenclature of knee injury type varied between studies. The range of injury types included patellofemoral joint disorders, tibiofemoral joint disorders, meniscal disorders, ligamentous injuries, strains, sprains, aches, bursitis and muscular injuries. It was difficult to ascertain the most common injury type because of the lack of definitions provided for the range of injury types reported in the literature. Only three studies reported knee injury mechanisms (16%). The most common injury mechanisms were landing from a jump, shooting, guarding, and collision with another player.3 18 25 The type of treatment provided was reported in seven of the studies reviewed, although four of these studies provided information only on surgical interventions.17 20 22 23

Injury severity varied in description between studies. The classification of injury severity varied between studies, and hampered comparison of knee injury severity across studies and sports (see table 6).

Knee surgery

Knee surgery is often used as a measure of injury severity, but it is also an indication of the economic costs of knee injuries and the capacity of the individual to pay for treatment. Table 7 illustrates that from four studies, which provided adequate data, at least 17% of all knee injuries sustained by adolescents in this research synthesis required surgical intervention.

Study variables

Table 8 provides a summary of the variables studied reported in the eligible studies. The most common outcomes reported include the level of play, type of sports, game conditions (that is, practice or game) and hours of play.

Methodological appraisal of studies reviewed

The average methodological appraisal score of the 19 studies was 56% and the scores ranged from 43% to 81%. Only about 20% of the studies reviewed reported the validity and reliability of measurement tools and data collection procedures.1 3 24 25 About 80% of the reviewed studies failed to adequately describe the sample, and often basic information such as gender and age was lacking.17 24 The findings of about 87% of the studies could also not be generalised beyond the study sample due to, for example, the sample recruitment method employed to select the study sample. The exposure measures by the some studies were also limited and thus resulted in a relatively lower quality score.

The methodological quality of the studies appears to have improved progressively from 1980 to 2005, as indicated by the trend line in figure 5. The findings of the methodological appraisal indicated that only two of the studies1 32 reviewed scored more than 70%, the recommended benchmark for methodological quality of epidemiological studies incorporated in systematic reviews.34 Since only two studies would have qualified if a 70% methodological score was deemed to be acceptable, it was therefore not possible to only include the high quality studies.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5 Relation between knee injuries and gender.

DISCUSSION

This paper provides the first known systematic review of adolescent sports epidemiology regarding knee injury prevalence and risk factors. This review consequently fills a gap in the current understanding of the prevalence of adolescent sports injuries. Synthesis of epidemiological information is important to provide a composite picture of injury prevalence and predisposing factors to injury and this is required to plan effective injury prevention strategies.35 Findings from this review indicate that adolescent knee injury should be of concern in a public health sense, as well as an individual player sense. This paper reports on a robust approach to study identification, critical appraisal, synthesis of study findings and provides a benchmark for systematic reviewing of sports epidemiology studies.

Critical appraisal of methodology is often lacking in systematic reviews of sports epidemiological studies.36 This review illustrates that there is opportunity to improve on the methodology of current youth sports epidemiological investigations by placing greater focus on injury definitions, reporting, and risk factor identification. For instance, four (21%) of the included studies scored below 50% on the proposed methodological scoring system. This provides the reader with an insight into what is required to improve the validity of epidemiological study findings. Validity and reliability of measurement tools were the most common shortcomings of the studies reviewed. Validity of measurement tools is an important element to be considered in any research study to ensure that the measurement tool is accurately evaluating the variable under study. A failure to determine the validity of measurement tools, such as questionnaires in injury surveillance studies, may lead to inconclusive study findings that lack reporting on important factors associated with injury. This may explain exclusion of important knee injury associated variables (table 9) noted in the studies reviewed. The poor attention to reliability of measurement tools is of further concern as the users of research in injury prevalence, such as prevention policy makers and sports administrators, may not have confidence in the findings of a prevalence study where reliability has not been established. Consequently, appropriate actions to prevent sports injuries may not be proposed by injury prevention policy makers, as the true scope of the problem may be masked by methodological errors. Future research in the field of knee injury research should thus ensure that valid and reliable measurements are used to collect the data.

Despite the limitations in methodological quality, the studies reviewed signify a high prevalence of knee injuries sustained by adolescent sporting participants. This should be of concern to sports administrators, coaches, sports scientists and parents, as increasing injury prevalence means increasing health costs and time lost from physical activity. According to this review, knee injury appears to be a common occurrence among adolescent sports participants and, given the importance of maintaining physical activity in adolescence and adulthood, should flag concerns regarding maintaining community health. Knee injury is a common reason for temporary incapacity to play sports4 and can incur long-term disability.9

Scientifically supported preventative strategies are considered worldwide to be the first line of defence against knee injuries.9 This review illustrates that all except one of the epidemiological knee injury studies were conducted in the developed world (see table 3). The developed countries are characterised by relatively better injury surveillance systems, coaching and training, and health resources, than developing countries. Although it appears that knee injury prevalence from sports participation in developing or third world countries3 may mirror the developed countries,23^–25 developing countries are faced with limited access to appropriate health care, poor and restricted training opportunities for coaches, and limited financial resources to fund youth injured in sports.3 Prevention of knee injuries should be considered to be a critical factor in sports medicine worldwide, although primary predictors and effective prevention strategies of modifiable risk factors, such as improving coaching training and playing conditions, may vary between developed and developing countries.

The findings of this review suggest that knee injuries constitute a significant proportion of the injuries sustained by young sports participants (see fig 2). The studies included in this review are however not truly comparable due to differences in methodological quality, injury definitions and time frames of injury prevalence. The methodological shortcomings, particularly with respect to reliability and validity warrant study findings to be considered with caution. Variation in injury definition between the studies reviewed is another barrier to comparing injury prevalence between studies. For instance, this review highlights the shortage of knee injury risk information per playing hours (see table 8) for adolescents. Therefore, in order to gain true insight into the epidemiology of knee injuries among adolescents, globally acceptable standardised definitions of injury and exposure must be developed for each type of sport. The definition of injury should clearly indicate what an injury episode constitutes, how exposure should be calculated and how severity should be classified. It will be crucial that the development of these definitions must be through the consensus opinion of international researchers in the field of sports injury prevalence.

The apparent high proportion of knee injuries requiring surgical intervention is alarming (see table 10), and indicates that knee injuries sustained by young sports participants generally involve serious articular or ligamentous damage that cannot be repaired using non-invasive approaches. Adolescents would thus seem to be susceptible to many of the injuries sustained by their adult counterparts.37 The high economic costs of surgical intervention and postoperative rehabilitation to ensure that players are safely able to resume sporting activity are of particular concern to players and parents from relatively poorer countries in the developed world.3 However, it is also of concern that none of the 19 studies reviewed provided information on postoperative functional rehabilitation (its nature or its success). Functional rehabilitation is often neglected after surgery, and may be one of the reasons for the high reported recurrence rates of sports knee injuries.16

Females are reported have a higher risk of sustaining a knee injury than males, and this concurs with gender-specific reports on knee injuries.4 A myriad of reasons is provided in the literature to explain why females are more at risk of knee injuries.38 These risk factors include biomechanical factors such as increased dynamic valgus, high abduction loads on the knee during landing tasks, hormonal factors, and neuromuscular factors associated with puberty.38 Recent preventative strategies by Hewett et al indicate that these biomechanical risk factors among females be addressed by neuromuscular exercise programmes. The findings of a recent meta-analysis of injury prevention demonstrate that neuromuscular exercises can be effective in reducing knee injuries39 and therefore warrant wide-scale implementation.

As an organised non-contact sport, basketball arguably has the highest rate of knee injury (see fig 4).25 38 Repetitive jumping in basketball imposes recurring consistent vertical ground reaction forces of up to four times body weight on the weight-bearing knee joint.40 The maturing (adolescent) neuromuscular system may be unable to maintain knee stability and around-joint control, leading to forces above the physiological threshold, with inevitable injury to the knee joint structures.40 The level and intensity of play in high risk sports is one area which may be addressed by parents, coaches, sports administrators and injury prevention policy makers. The focus of sports activity by young people should be directed at technique development in order to train the neuromuscular system of young people to control the biomechanical strain imposed by sports such as soccer and basketball. Preventative knee injury exercise programmes focussing on neuromuscular control and technique development should be developed collaboratively by the key stakeholders such as physiotherapists, physical trainers and coaches.41 Practical considerations, such as the duration and frequency of the preventative exercise programme and the ability of the coach to deliver it, could be a potentially cost-effective solution in successful implementation of knee injury prevention programmes at club or school level.

CONCLUSIONS

The high prevalence of knee injuries sustained by adolescent sports participants as described in this review is of worldwide concern. The outcomes from this review are important in identifying future research and injury surveillance directions, as well as flagging advances in epidemiological research, which will appropriately inform effective knee injury preventative programmes for youth. It is important that researchers take note of the findings of this review, and focus on developing standard injury definitions as well as descriptions of injury causes. Only with worldwide standardised knowledge will it be possible to make progress in preventative strategies for adolescent knee injuries.