Upper Extremity Musculoskeletal Profiles in Tennis Players: A Systematic Review

Abstract

Context:

Tennis-specific musculoskeletal (MSK) screening can assess range of motion (ROM) and muscular imbalances. Identifying normative values before implementing a MSK screen is essential in contributing to athlete performance and injury risk profiles.

Objective:

To review upper extremity MSK data in healthy tennis players across age, sex, and level of play.

Data Source:

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines were followed for this review. A search was conducted in MEDLINE, SPORTDiscus, Embase, and CINAHL.

Study Selection:

This review included shoulder, elbow, and wrist ROM, isometric strength, or isokinetic strength in a tennis population. Each article was critically appraised to help identify the internal and external validity of each study.

Study Design:

Systematic review.

Level of Evidence:

Level 3.

Data Extraction:

A total of 41 studies met the search criteria. Each contributor organized the data elements of interest into data tables, with a second contributor assigned for review. Data elements of interest included player and study characteristics: ROM, isometric dynamometry, and isokinetic strength.

Results:

A total of 3174 players were included in the final studies. Most of the players included were competitive adolescents and young adults; 15 studies included ROM data. Male tennis players consistently had more external rotation (ER) gain (range, 1.8º to 8.8º) and internal rotation (IR) loss (range, -15.3º to -3.0º) when compared with their female counterparts (ER range, -2.5º to 5.8º; IR range, -10.4º to -3º). Shoulder IR and ER strength were measured in the majority of all the strength studies, with the external rotators generating at least two-thirds the strength of the internal rotators.

Conclusion:

Overall MSK data of tennis players indicate that shoulder strength values are often larger than nontennis players, but equal to or slightly lower than comparable athletes in other overhead sports. Adaptive changes of the glenohumeral joint and subsequent rotational motion are similar to those of other overhead athletes.

Tennis is enjoyed by about 83 million people around the world. ⁸ It is not uncommon for persons who play tennis to maintain active in the sport throughout life. ⁴⁷ Players who regularly engage in the sport of tennis have superior fitness levels (measured with different physiological tests such as VO₂max) and body composition when compared with matched controls. ⁴⁷ Recently, researchers also determined the presence of improved musculoskeletal (MSK) function in tennis players compared with nontennis players. ²⁹ In this study by Jackson et al, ²⁹ MSK function for both the upper and lower extremity was measured using anthropometry, dynamometry, and oxygen uptake. An MSK screen can also include range of motion (ROM)/flexibility measurements that focus specifically on key contributors to tennis-specific movement. The inclusion of all or some of these outcome measures in the MSK screen can be implemented by the healthcare team to identify normative values, symmetry between sides, or any existing flexibility or muscular imbalances.

The repetitive, tri-planar anatomic motions and multidirectional movements on court can make tennis players susceptible to injury.^1,36 Thus, implementing MSK screenings into practice may assist professionals in monitoring training and injury risk by identifying deviations from normal. In fact, a recent paper by Rice et al ⁵¹ highlights the importance of a tennis-specific MSK screen that assesses trunk stability, lower limb flexibility, and hip ROM in different ages and sex. Data collected from MSK screens can be stratified by sex, age, and level of competition to better identify deviations between different groups of tennis players. It has been reported that the information collected from the MSK screen can help improve athlete performance and/or alter the risk of future injury during sport.^7,33 In addition, the screening process helps players understand how they compare to players with similar characteristics, and helps healthcare professionals create evidence-based prevention programs for players that may need to work on improving MSK health.

While the sport of tennis requires contributions from the entire MSK system, much attention and research has focused on forces and injuries in the upper extremity. Each stroke results in large impulsive and smaller vibration forces transmitted to the player’s hand by the impact of the ball on the strings and racket.^34,35 Biomechanical studies have also estimated net forces and moments in the upper extremity during strokes (not specifically impact) to study coordination and the safety of net repetitive loads relative to mechanical tissue strength. More specifically, the trunk and upper extremity of skilled tennis players undergo relatively high net forces and moments during all the tennis strokes.^3,24,31 The combination of the repetitive nature of tennis strokes and high anatomic loads has resulted in both acute and overuse injuries of the upper extremity. ⁴⁶ In fact, 20% to 49% of all tennis-related injuries occur in the upper extremity, ¹ further demonstrating the importance of screening and monitoring the development of injury in tennis athletes. Consequently, sports medicine professionals hypothesize that playing tennis at a high level requires excellent muscular strength, flexibility, and neuromuscular control. ³⁶

Given the rigorous requirements of tennis play, it is imperative that, to make informed recommendations, sport medicine professionals have access to data related to upper extremity MSK function for all levels of players readily available. Therefore, the purpose of this paper is to review tennis-specific upper extremity MSK data in healthy tennis players across age, sex, and level of play. Specifically, we focused on the MSK variables of muscular strength and ROM in both dominant and nondominant arms.

Methods

The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines were followed during the creation of this review.

Literature Search

An online literature search was performed in several relevant databases (MEDLINE, SPORTDiscus, Embase, and CINAHL Complete) from November 1 to 30, 2021. The search was updated in the fall of 2022 to ensure no new articles would be missed. A Population (P), Intervention (I), and Outcome (O) (PIO) search strategy was implemented in each individual database to retrieve the best set of results possible. A combination of different words were used to describe the same thing for population, intervention, and outcome. The MeSH and Emtree functionality in MEDLINE and Embase were used to help develop our search terminology. This strategy helps to capture all the results relevant to the research question. The “population” included all levels of tennis players, and several search terms were used to describe the population, such as junior OR elite OR high school OR recreational OR college OR professional. The “intervention” included the specific measurement devices used to collect the main outcome measures of the study, such as handheld dynamometer OR isokinetic OR goniometer OR inclinometer. The “outcome” included the MSK variables of interest, such as: ROM OR grip strength OR flexibility OR strength for the scapular/shoulder region, elbow, and wrist/hand. Each of the search terms under P, I, and O were first combined using OR. Consequently, a set of results were generated for each part of the PIO model. The authors then used AND to bring all the P, I, and O results together. In addition, manual searches were conducted using the reference lists of the identified articles and author’s research files.

Selection Criteria

For a study to be included in this review, it needed to identify upper limb (shoulder, elbow, wrist/hand) ROM and strength profiles among healthy tennis players. The study designs associated with these types of investigations are Level 3 evidence according to the Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence. Tennis players could be defined as elite, junior, college, recreational, or adult players. Since the goal was to review MSK profiles in healthy tennis athletes, the review allowed for all kinds of competition levels. ROM needed to be reported in degrees, isometric strength reported in peak force, and isokinetic strength measures needed to be reported as peak torque or peak torque/body weight. In addition, the research team needed to have access to the full article in the English language. Studies were excluded if they investigated the effect of an intervention on ROM or strength (Level 1 and 2 evidence), did not report standard deviations, included wheelchair or adaptive tennis players, or were published in conference proceedings. A date range was not executed in this particular study as the authors wanted to identify all datapoints that had been collected on this topic.

Study Selection and Critical Appraisal

Following the literature search, the lead author screened the titles and abstract of all articles that were returned through the PIO search in each of the 4 databases. Full-text eligibility was confirmed by the lead author. The final articles to be included in the study underwent a risk of bias assessment using the Appraisal Tool for Cross-sectional Studies (AXIS). ¹⁵ The AXIS tool is used to appraise observational cross-sectional studies. The AXIS has 20 questions, and each question is accompanied by explanatory help text that gives background knowledge and explanation as to what the questions are asking. Questions are categorized into Introduction, Methods, Results, Discussion, and Other. Two authors independently rated each article that met the specified inclusion criteria. A third author was introduced to the appraisal process when any disagreements between the authors could not be resolved (<8% of articles needed a third review).

Data Extraction

All authors collaborated to extract the appropriate data. Once the data were extracted, a research assistant reviewed all the data to ensure accuracy. The data extracted included player demographics, inclusion/exclusion criteria, ROM measures, handheld dynamometry (HHD) measures, isokinetic measures, grip strength, the unit of measure, and any reported reliability metrics. The means and standard deviations were extracted from all the studies. One study reported median and interquartile ranges. ³⁰ Using the data from the included studies, tables were developed for each outcome to report data for age, sex, and level of play for all the dependent variables when available.

Results

Upon completion of the literature search, a total of 41 articles met the specified inclusion criteria and were categorized as Level 3 evidence according to the Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence (Figure 1). All articles were cross-sectional in nature, investigating a cohort of tennis players. The critical appraisal analysis of all 41 articles can be found in Appendix A (all appendices available in the online version of this article). AXIS scores ranged from 10 to 18, with 20 being the highest possible score (Appendix A). Higher total scores represented better quality studies. Answers that met the criteria for each question were labeled as “yes” and answers that did not meet the criteria were labeled as “no” or “don’t know.” An answer of “yes” for all questions on the AXIS (except for questions 13 and 19) were preferred as this response indicated improved quality of each study. Appendix A includes cell shading so the readers can quickly see the majority of answers that met the appraisal criteria. Only 1 study reported a power analysis that justified sample size, and no study addressed categorization of nonresponders.

Figure 1.

Flow diagram for article selection.

This review presents upper extremity ROM and strength data (isometric HHD and isokinetic dynamometry) on 3174 (male: 1912 [60%]; female: 1196 [38%]) healthy tennis players (not all articles specified the sex of tennis players). The majority of the players included in this review were high-level competitors at a variety of age ranges, with few studies including recreational level players. When available, all data were partitioned out between age and/or sex. Detailed study characteristics, along with patient position and instrumentation, are presented in Appendix B Tables A1 and A2, respectively. ROM data were reported using degrees. The studies were not specific with joint convention used and whether zero was referenced in the horizontal or vertical plane. Strength data are reported as either N, N/kg, kg, or Nm.

All ROM studies (n = 15) included both dominant and nondominant glenohumeral internal rotation (IR) and external rotation (ER) measurements (Appendix C, Table A1). Only 1 article reported elbow and wrist ROM. ¹² Elbow extension was 0.5 ± 4.9° and -0.5 ± 4.7° for the dominant and nondominant arms, respectively. Elbow flexion was 142 ± 4.1° and 143 ± 3.8° for the dominant and nondominant arms, respectively. Pronation was 74.7 ± 6.6° and 72.8 ± 7.5° for the dominant and nondominant arms, respectively. Supination was 82.9 ± 8.7° and 83.5 ± 8.4° for the dominant and nondominant arms, respectively. Wrist extension was 67.5 ± 10.3° and 68.1 ± 10.3° for the dominant and nondominant arms, respectively. Wrist flexion was 78.7 ± 6.8° and 80 ± 6.4° for the dominant and nondominant arms, respectively.

Isometric strength using HHD was reported in 7 studies (Appendix C, Table A2). Of these studies, 4 included both glenohumeral IR and ER, and 3 included scapular muscle strength. Grip strength was reported in 16 articles (Appendix C, Table A3). Normative values for dominant arm grip strength were aggregated based on level of play, sex, and age (Appendix C, Table A4). Isokinetic shoulder (Appendix C, Tables A5 and A6), elbow and wrist (Appendix C, Table A7) strength were reported in 15 and 6 studies, respectively. One study reported on shoulder horizontal abduction and adduction isokinetic strength (Appendix C, Table A6). ⁵⁹ Articles included in the systematic review that are not referenced within the full text manuscript can be found in Appendix D.

Discussion

The MSK screening data presented in this review can be used to monitor and inform health and performance training programs of junior and adult tennis players. Integration of these variables into a screening allows for clinicians to monitor for tennis-specific upper extremity strength and ROM imbalances and compare athletes with similar and aspirant populations. The MSK measures summarized in this review can be tested easily by sports medicine professionals and, if needed, can be altered using appropriate interventions that target soft tissue flexibility and strength. Results for tennis players were interpreted in the discussion relative to extensive normative data previously reported for the tests in untrained children and adults, as well as similar athletes from sports with repetitive upper extremity motions.

Range of Motion

Adequate shoulder rotation ROM is necessary to help mitigate the forces generated across the shoulder complex during different tennis strokes. The repetitive nature of tennis results in adaptions of the available ROM of the dominant shoulder complex. ³² This review identified 15 articles reporting tennis players’ rotational ROM adaptions across varying sex, age groups, and level of skill to provide clinicians and coaches with normative rotational ROM profiles. It is important to note, however, that the results of these data should be interpreted with caution. Four studies did not differentiate player sex while investigating ROM changes,^9,10,13,44 7 studies did not report reliability for ROM measures,^{9,11,12,20,22,39,49} and a wide variety of measurement type (active versus passive ROM), methodology (stabilized versus nonstabilized scapula), and instrumentation (goniometer versus digital inclinometer) was used in the included studies.

Qualitative trends were noted as they pertain to glenohumeral rotational ROM. With the exception of 2 studies,^11,22 all athletes measured had a gain of ER ROM on their dominant limb as compared with the nondominant limb. In addition, all athletes included had a loss of dominant limb IR ROM. When total rotational ROM was available, only 1 study found a gain in total rotational ROM on the dominant limb compared with the nondominant limb ⁴⁴ ; all other studies that reported total rotational ROM reported a loss of total rotational ROM of the dominant limb compared with the nondominant limb. Male tennis athletes had larger reported values of both glenohumeral ER ROM gain (male range, 1.8 º to 8.8º; female range, -2.5 º to 5.8º) and IR ROM loss (male range, -15.3º to -3.0º; female range, -10.4º to -3º). Based on the included studies, trends in rotational ROM based on playing age or level of play were mixed. Of the 2 available studies based on age,^13,30 there was a slight trend toward greater ER ROM gain and IR ROM loss in older athletes as compared with younger athletes, with no trends in total rotational ROM noted. In regard to competitive playing level, less ER ROM gain was noted in men compared with women. ³⁰

Adaptive changes of the glenohumeral joint and subsequent rotational ROM are not unique to tennis athletes. Athletes participating in other overhand throwing sports such as baseball, softball, swimming, and volleyball display similar differences in glenohumeral rotation in the dominant shoulder. ⁴⁸ While the results vary from study to study, it is generally well accepted that baseball and softball athletes demonstrate similar changes to dominant limb rotational ROM. Baseball and softball athletes typically demonstrate a pattern of glenohumeral ER ROM gain and IR ROM loss.^20,60 These phenomena may vary based on playing age, level of play, and position played.^4,56,57 Like the tennis athletes represented in this review, volleyball and swimming athletes have also been shown to have dominant limb gain of ER ROM and loss of IR ROM.^26,28,52

Glenohumeral IR loss has been previously associated with shoulder and elbow injury in overhead sport.^6,64 However, it is important to consider the adaptive changes of osseus and soft tissue structures and their effects on the interpretation of rotational ROM. The development of humeral retrotorsion of the dominant limb, ⁴³ changes to the posterior glenohumeral joint capsule,^62,63 and changes to the soft tissue morphology of the rotator cuff have all been associated with changes in rotational ROM in baseball athletes. ⁶¹ The recognition of these adaptations in overhead athletes allows clinicians to better understand normative changes to rotational ROM, typically presenting as ER ROM gain and IR ROM loss of the dominant limb. Due to these noted changes of the glenohumeral joint, IR ROM loss is expected to a certain extent. More recently, the interpretation of total rotational ROM loss of the dominant limb has been recommended for use to differentiate those at risk for injury. ⁵⁸ Utilizing total rotational ROM loss allows clinicians to consider the effects of both ER ROM gain and IR ROM loss while assessing rotational ROM. While most studies investigating morphological changes of the glenohumeral joint have occurred in the baseball population, the authors hypothesize that similar anatomic changes would be found in tennis athletes given the similarities in the repetitive use of the dominant limb during participation. Further research is recommended to confirm this hypothesis.

Strength

A solid upper extremity strength foundation is required in tennis due to the repetitive stroke demands placed on the upper limb during tennis groundstrokes and overhead serves. This review combined the available evidence on tennis players so that clinicians and strength and conditioning professionals could reference normative strength profiles for different age and competition levels. The authors felt it important to report both isokinetic and HHD isometric values so clinical decision making could be made in the presence of both laboratory and clinically practical equipment. Regardless of isokinetic speed, trends in elbow torque are mixed across the studies included in this review. One study reported that isokinetic elbow extension and flexion torques were relatively equal when taking into account the standard deviations on the dominant arm. ⁶⁶ Another study observed the strength of dominant arm elbow extensors to range between 2% and 8% and 9% and 14% larger for male and female tennis players when compared with elbow flexion torque. ¹⁹ The torque differences between dominant forearm supination and pronation were highly variable and dependent on isokinetic speed, sex, and age. Trends in forearm torque demonstrated the strength of the pronators to be 30% to 40% stronger when compared with the supinators.^21,25 One study reported isokinetic supinator and pronator torques to be relatively equal when taking into account the standard deviations on the dominant arm. ¹⁶ The differences in findings may be attributed to the variety in player age and competition level. Wrist flexion torque was consistently larger (range, 17%- 34%) than wrist extension torque regardless of isokinetic speed.^16,21,25,66 Interestingly, these wrist flexor torques were about 17% smaller than similar levels of inactive adults <30 years of age. ²⁷ On the other hand, tennis players wrist extensor strength that likely hypertrophy from repetitive 1-handed backhands were equal to, and up to 40% larger than, norms for age-matched untrained adults.^27,50 These data confirm that regular tennis play tends to create strength imbalances, favoring primarily forearm pronation, wrist flexion, and elbow extension of the dominant arm.

The majority of articles reporting on upper extremity strength of tennis players were specific to glenohumeral ER and IR. IR and ER peak torque on the dominant arm at lower isokinetic speeds ranged from 40 to 70 N·m and 25 to 35 N·m, respectively. When compared with healthy adults (no tennis background) with an average age of 25 years, tennis players generate as much as 20% more IR torque and 25% more ER torque. ⁴⁵ When measuring rotational strength isometrically with an HHD, it appears that tennis players display lower levels of isometric ER strength when compared with the throwing arm of baseball (25%) and softball (14%) players. ⁴⁴ Tennis HHD isometric ER data, when compared with other sports like volleyball and handball, are also lower by 4% to 10% in male players. In female players, the data present differently, with handball players presenting with 25% stronger ER when compared with tennis players, and tennis and volleyball players exhibiting nearly similar ER strength profiles. ¹⁴ The isokinetic strength data show the same trends, and, regardless of how strength was measured (side, age, or level of play), IR strength was greater than ER strength.

Glenohumeral rotational muscular imbalance is a common occurrence across tennis along with other overhead sports. ¹⁴ As such, it is not uncommon for researchers to calculate the ER/IR ratio from HHD to help determine the extent to which IRs outproduce ERs. Isometric ER/IR ratios vary among healthy athletes between different overhead sports (0.59-1.02). ¹⁴ The studies included in this systematic review reported ER/IR isometric strength ratios ranging from 0.66 to 0.90. It is important for practitioners and coaches to evaluate strength profiles of their athletes as agonist/antagonist ratio values provide muscular balance information. Typically, before imbalances develop, shoulder ratios are between 66% and 75%, meaning the external rotators are as least two-thirds the strength of the internal rotators. ¹⁷ Isokinetic glenohumeral strength profiles take on similar trends to isometric profiles. If equipment is available, it may be beneficial to test torque using isokinetic equipment. In fact, some researchers have started investigating a functional ratio of eccentric ER to concentric IR, and argue that eccentric ER should be greater than concentric IR torque.^41,42 The studies included in this systematic review report functional EccER/ConIR isometric strength ratios ranging from 0.95 to 1.19.^{5, 12} Functional EccER/ConIR ratios vary among healthy athletes between different overhead sports (0.84-1.17).^40,42,53 In summary, muscular imbalances do exist and can be a normative finding within a particular range. The authors of this review suggest not only interpreting the ratio differences between the internal and external differences but using the raw data in addition to the ratios to determine whether your tennis athlete falls into a normative range for isokinetic strength for his or her particular age and competition level.

When available, the grip strength data presented in this review have been presented by sex, level of play, and age to better assist in the clinical decision-making process. Elliott et al ²³ reported grip strength data on young male and female players and found male high performance players display as high as an 8% stronger dominant arm grip strength when compared with age similar recreational players. ²³ High performance female players display a similar grip strength discrepancy that can go as much as 15% higher than recreational players. ²³ When comparing the grip strength in this review with that of adults of a similar age range (18-24 years), most data reported fall below the 50% percentile. ⁶⁵ Only 1 study on professional male players reported grip strength in the 50% percentile when compared with similarly aged adults.^55,65 This comparison study did have a low sample size, and reports with isometric grip strength from larger samples of tennis players indicate similar and within the standard deviations of age-matched male and female junior players (15 years). ⁵⁴ In general, there are inconsistent results of regular tennis play resulting in greater grip strength than age-matched persons. It is possible that regular tennis play does not dramatically increase maximal grip strength, which is consistent with studies of pre-impact grip forces in strokes being moderate (32%-62% MVC), ³⁵ and considerable grip muscular endurance of tennis players. ³⁷ Nevertheless, the grip strength metrics presented in this review (Appendix C, Table A3 and A4) may be of value in overall screening of player strength and used to exercise programming. ¹⁸ Player grip strength can also be interpreted relative to extensive normative data on general upper extremity strength. ²

Limitations

Research synthesis of measures of MSK variables of tennis players is a difficult task given the diversity of the sport (surfaces, playing styles) and players (age, skill level, injury history). It is unclear how history of injury, outside exercise, or other sports activity contribute to the MSK fitness values in this study. Inconsistent use of terms and definitions of skill level is a major limitation of all sports and exercise science research. Other limitations of the current data are variability across measuring instruments, testers, and testing protocols. ³⁸ For example, differences in body positions and stabilization for HHD strength measures make it hard to make direct comparisons across studies. While the authors of this review made a conscious effort to diminish the risk of bias through dual author critical appraisal, only 1 author ran the initial searches for article inclusion. In addition, the data extracted were not compared between each of the authors; however, all data were reviewed rigorously to ensure accuracy. The current MSK data emphasize high level to elite tennis players, with less data on recreational players. Lastly, the game of tennis has evolved over time in terms of equipment, strategy, physical demands, and stroke technique; readers must take this into consideration when comparing an early (1980) publications with later (2000) publications.

Conclusion

The repetitive motions of regular tennis play likely result in upper extremity strength and ROM imbalances. Overall MSK data of tennis players indicate that strength and ROM values are sometimes larger than those of similar untrained adults, but equal to or slightly lower than comparable athletes in other repetitive overhead motion sports like baseball and volleyball. These data may be useful for MSK screening for tennis players; however, longitudinal and training studies are needed to make evidence-based recommendations for performance or reduction of injury risk.