Many Shoulder MRI Findings in Elite Professional Throwing Athletes Resolve After Retirement: A Clinical and Radiographic Study : Clinical Orthopaedics and Related Research®

Secondary Logo

Journal Logo

BASIC AND CLINICAL RESEARCH

Many Shoulder MRI Findings in Elite Professional Throwing Athletes Resolve After Retirement: A Clinical and Radiographic Study

Schär, Michael O. MD1; Dellenbach, Simone MD1; Pfirrmann, Christian W. MD MBA2; Raniga, Sumit MBChB, FRACS1; Jost, Bernhard MD3; Zumstein, Matthias A. MD1

Author Information
Clinical Orthopaedics and Related Research: March 2018 - Volume 476 - Issue 3 - p 620-631
doi: 10.1007/s11999.0000000000000042
  • Free

Abstract

Introduction

The high forces and the kinematic requirements of overhead throwing produce a unique pathophysiology that is most evident in the shoulders of elite throwing athletes. Throwing shoulders have been found to adapt by increasing external rotation, decreasing internal rotation, increasing humeral retrotorsion and glenoid retroversion, and developing anterior capsular laxity. The combination of the adaptations and the demands on the throwing shoulder means that it is vulnerable to common acute and chronic pathologic conditions.

Professional European handballers (team handball in the United States) perform at least 48,000 overhead throws per season with ball speeds reaching up to 100 miles per hour [15]. In an earlier study, Jost et al. [12] reported that structural abnormalities were present in 93% of the throwing shoulders of professional European handballers on MRI, with an average of seven findings per scan. Abnormalities identified included the presence of partial-thickness rotator cuff tears and tendinopathies, superolateral osteochondral defects of the humeral head, and changes according to so-called posterosuperior intraarticular impingement [23]. Interestingly, clinical symptoms correlated poorly with the clinical and MRI-observed findings. Only 37% of the players with evidence of an abnormality observed on MRI actually were symptomatic. The natural history of these findings once professional overhead athletes cease their playing career remains unclear. A better understanding of what happens with these radiographic and clinical findings after cessation of the career may have an influence on how we manage athletes with these shoulder disorders during and after their career.

We therefore asked: (1) What is the natural history of MRI-observed findings in the throwing and nonthrowing shoulders of professional European handball players after retirement from their sport? We specifically asked, what proportion of these individuals have diagnosable findings on MRI, and do these findings worsen or disappear after retirement? (2) Do clinical parameters like the Constant and Murley score and shoulder ROM change after retirement in these professional overhead athletes?

Methods

Study Population

We chose to evaluate the shoulders of professional European handballers owing to the very high repetitive stresses placed on their throwing shoulders. European handball is a team sport in which two teams of six field players each pass a 450 g ball (19 cm in diameter) with their hands, with the aim of throwing the ball into the goal of the opposing team. The goals are surrounded by a 6 meter zone where only a defending goalkeeper is allowed. Therefore, most of the throwing motions are performed in the overhead position resulting in high repetitive stresses on the throwing shoulder of a field player [15]. These stresses are at least comparable to those of a baseball pitcher, since a handballer’s throwing shoulder has to cope with the stress of sudden rapid deceleration as well, because opposition players may block these overhead throws with forceful contact.

Patient Selection

Our study received local ethics committee approval (KEK-ZH-2011-0199/4 and KEK-BE-112/11). In a previous study (initial followup) performed in 2001 [12], the symptomatic and asymptomatic shoulders of all the 30 full-time professional European handball players of the National Team except the goalkeepers were examined using a structured interview, detailed physical examination, and collection of the absolute (score of an individual obtained in points) and relative scores (score of an individual divided by the average score of an age- and sex-adjusted population), Constant and Murley score [6], and the subjective shoulder value (defined as a patient's subjective shoulder assessment expressed as a percentage of an entirely normal shoulder, which would score 100%) [9] of both shoulders. Each of these players also played in one of the top four teams of the Swiss Handball League.

Of the 30 subjects in the previous study, 10 were excluded (four continue to play professionally, four declined participation, and two had surgery), leaving 20 (66%) for analysis here at a mean of 6 years (SD, 3 years) after retirement. The two players who had surgery included one who underwent open refixation of the pectoralis major tendon in the throwing shoulder, and one who underwent arthroscopic acromioclavicular joint resection in the throwing shoulder. The mean duration of playing handball in this group was 18 years (SD, 4 years), and the mean duration of competition on a professional level up to retirement from professional sport was 11years (SD, 4 years), with an average practice per week including games of 19 hours (SD, 6 hours). The mean age of the players 6 years after retirement was 36 years (SD, 4 years).

To understand the long-term evolution of shoulder findings associated with professional European handball, we also selected 18 former professional handball players, of whom all also had played for the Swiss National Team. We selected this group by inviting 20 former professional players. Of those, 18 (90%) were able to participate, and two (10%) refused participation in the study. Similar to the group that was evaluated initially and 6 years after retirement, asymptomatic and symptomatic former players were included in this group. No one was excluded because none had received previous shoulder surgery. These players had terminated their careers at a mean of 15 years ago (SD, 3 years) but were not part of the previous study and therefore received only a bilateral shoulder MRI, structured interview, and clinical examination 15 years after termination of their career but not during the career. The mean duration of playing handball was 15 years (SD, 3 years), with 11 years (SD, 3 years) of playing on a professional level, and 16 hours (SD, 3 hours) per week of average practice time including games. The mean age of the players 15 years after retirement was 50 years (SD, 5 years). Even though this second group does not represent the exact natural history of the group examined initially and 6 years after retirement, its findings may still provide important information with respect to the long-term evolution of the examined shoulder findings (Table 1).

T1
Table 1.:
Professional and retired handball players included in this study divided in three groups for evaluation

Clinical Evaluation

A standardized clinical evaluation of each former handball player was performed by two examiners (MOS and SD). The methodology was identical to that used in the initial study [12], namely a structured interview including patient history, pain in general using a VAS (0 = worst imaginable pain to 15 = no pain), loss of function, and intensity and duration of training. Furthermore every shoulder was tested for active and passive ROM, for instability using the anterior and posterior apprehension and hyperabduction test, for laxity using the sulcus sign, anterior and posterior drawer test, for rotator cuff lesions using the Jobe test, external rotation strength, lift-off test, and lag signs, for subacromial impingement using the Neer and modified Hawkins test, for posterosuperior glenoid impingement using the Walch test, and for biceps abnormalities using O’Brien’s test. The absolute and relative Constant and Murley score [6] and the subjective shoulder value [9] of both shoulders also were obtained.

A change of 10 points for the absolute Constant score [14] and 3 points in the 0 to 15 VAS pain score [18] were considered the minimum clinically important difference (MCID).

Pain at initial followup and 6 years after retirement was correlated with radiologic and clinical findings. This information is important because it might help to better understand which MRI-observed findings and clinical signs cause pain and therefore may influence the choice of treatment.

Radiologic Evaluation

Identical to the initial study, a bilateral shoulder MRI scan without intraarticular contrast of the throwing and nonthrowing shoulders of each former player was performed. The MRI methodology was identical to that used in the initial study [12]. Every MRI scan was assessed by an experienced and blinded musculoskeletal radiologist (CWP). MRI findings were reported as radiographic diagnoses, independent of the potential that these findings could be considered normal variations in people in this age group.

Rotator cuff abnormalities were divided into tendinopathies, and partial or complete tears of the supraspinatus, infraspinatus, and subscapularis tendons. Osseous abnormalities in the glenoid and the proximal humeral head were divided into edema, cyst formation, and osteochondral defects. Glenohumeral ganglion cysts were classified as large (> 15 mm), medium (5-15 mm), and small (< 5 mm). Posterosuperior glenoid impingment [23] was diagnosed when an articular-sided partial tear in the posterior aspect of the supraspinatus tendon was visible. Anterior glenoid rim impingement was graded as not existing, mild, or severe according to Weishaupt et al. [24].

Degenerative changes in the glenoid and humeral head cartilage also were graded as either subtle or marked. Signal alterations or an irregular surface of the cartilage and superficial cartilage defects were classified as subtle. Defects greater than 50% of the cartilage thickness and defects reaching the subchondral bone were graded as marked.

The radiographic evaluation of the acromioclavicular joint and the long head of the biceps tendon was identical to that in the initial study [12].

Statistical Evaluation

The Mann-Whitney test was used for unpaired groups and the Wilcoxon test was used for paired groups. For the frequency between two groups, the McNemar test was used for paired groups and the chi-square test with Yates correction or Fisher’s exact test where appropriate was used for unpaired groups. Odds ratios (OR) were calculated with 95% CI. In case of cells equal to zero, Haldane’s correction was performed. Spearman’s correlation coefficient was used to test relationships between variables. The level of significance was set at a probability less than 0.05.

Results

Natural History of Findings on MRI

The number of abnormal findings on MRI evaluation per subject was not different between the three times, with the numbers available, in the throwing shoulder (initial followup: 7 [SD, 3], 6 years after retirement: 6 [SD, 4], and 15 years after retirement: 7 [SD, 3]; p = 0.717) (Table 2).

T2
Table 2.:
Findings on MRI evaluation

None of the tendinopathies and partial-thickness tears of the rotator cuff progressed to full-thickness tears. A greater proportion of throwing shoulders compared with nonthrowing shoulders showed supraspinatus tendon disorders at the initial followup (17 of 20 versus eight of 20; OR, 8.5; 95% CI, 1.9-38.8; p = 0.004) (Table 3). However, with the numbers available no difference was observed at a mean of 6 years after retirement (17 of 20 versus 14 of 20; OR, 2.4; 95% CI, 0.5-11.5; p = 0.375), nor in the group studied 15 years after retirement (17 of 18 versus 16 of 18; OR, 2.1; 95% CI, 0.2-25.8; p = 0.999). There was a greater number of throwing shoulders compared with nonthrowing shoulders with infraspinatus disorders in the initial evaluation (12 of 20 versus 0 of 20; OR, 60.3; 95% CI, 3.2-1137; p = 0.002) and at a mean of 6 years after retirement (10 of 20 versus 0 of 20; OR, 4.1; 95% CI, 2.2-770; p = 0.004), but with the numbers available, there was no such difference in the group assessed 15 years after retirement (eight of 18 versus seven of 18; OR, 1.6; 95% CI, 0.3-4.7; p = 0.999) (Table 3).

T3
Table 3.:
Frequency of MRI-observed findings in throwing shoulders versus nonthrowing shoulders

From the initial followup to the mean followup 6 years after retirement, four of eight subjects had the infraspinatus and subscapularis tendinopathies disappear (Table 4). Furthermore, two of six of the partial infraspinatus tears and one of the partial subscapularis tears disappeared. In contrast, three of 14 subjects who did not show signs of infraspinatus tears at initial followup had partial-thickness tears of the infraspinatus tendon develop 6 years after retirement while two of 19 subjects had partial-thickness tears develop in the subcapularis tendon.

T4
Table 4.:
Frequency of MRI-observed findings in the throwing shoulders of subjects during and after career

In the throwing shoulders, more posterosuperior glenoid impingement lesions were found at initial followup versus 6 years after retirement (eight of 20 versus one of 20; OR, 12.7; 95% CI, 1.4-114.4; p = 0.016) and 15 years after retirement (initial followup: (eight of 20 versus 15 years after retirement: 0 of 20; OR, 25.2; 95% CI, 1.3-476; p = 0.003) (Table 4). The throwing shoulders at initial followup showed more posterosuperior glenoid impingement lesions than the nonthrowing shoulders (eight of 20 versus 0 of 20; OR, 27.9; 95% CI, 1.5-526; p = 0.008), while there were no differences with the numbers available at a mean of 6 years after retirement (one of 20 versus 0 of 20; OR, 3.2; 95% CI, 0.1-82.2; p = 0.999), and 15 years after retirement (0 of 18 versus 0 of 18; p = 0.999) (Table 3).

Edema in the superolateral aspect of the humeral head of the throwing shoulders slightly decreased from seven of 20 at initial followup to three of 20 at a mean 6 years after retirement (OR, 3.1; 95% CI, 0.7-14.1; p = 0.289) and to 0 of 18, 15 years after retirement (OR, 20.6; 95% CI, 1.1-392; p = 0.009) (Fig. 1; Table 4).

F1
Fig. 1 A-B:
An axial proton density fat-saturated MRI scan of the left throwing shoulder obtained (A) at initial followup shows edema in the humeral head and greater tuberosity (arrows). (B) Six years after retirement, the edema is almost entirely gone.

There were more superolateral osteochondral humeral head defects in the throwing shoulders compared with the nonthrowing shoulders at initial followup (12 of 20 versus two of 20; OR, 13.5; 95% CI, 2.4-74.9; p = 0.006) and at 6 years after retirement (nine of 20 versus one of 20; OR, 15.5; 95% CI, 1.7-139.7; p = 0.008) (Table 3). With the numbers available, no difference between throwing and nonthrowing shoulders was seen 15 years after retirement (five of 18 versus one of 18; OR, 6.5; 95% CI, 0.7-63.0; p = 0.219). Osteochondral humeral head defects were found in 12 of 20 at initial followup and nine of 20, 6 years after retirement (OR, 1.8; 95% CI, 0.5-6.4; p = 0.375), and to five of 18, 15 years after retirement (OR, 0.6; 95% CI, 0.1-2.3; p = 0.095) (Table 4).

With the numbers available, no differences of large superolateral humeral head bony cysts were found at initial followup compared with 6 years after retirement (six of 20 versus one of 20; OR, 8.1; 95% CI, 0.9-75.5; p = 0.125) or 15 years after retirement (six of 20 versus three of 18; OR, 1.5; 95% CI, 0.3-6.5; p = 0.587) (Table 4). Nine of 13 subjects who showed bony cysts during their career showed a decrease in size 6 years after retirement (Fig. 2).

F2
Fig. 2 A-B:
A coronal oblique proton density fat-saturated MRI scan obtained (A) at initial followup of a subject’s right throwing shoulder shows a large superolateral osteochondral defect (arrows) on the humeral head close to the supraspinatus insertion. (B) Six years after retirement the same player was asymptomatic and the same defect (arrows) was reduced in size.

There were more ganglion cysts larger than 5 mm in diameter in athletes’ throwing shoulders during the career than 6 years after retirement (six of 20 versus 0 of 20; OR, 14.5; 95% CI, 0.7-283; p = 0.044), with no difference between the 6 years after retirement group and 15 years after retirement group (0 of 20 versus one of 18; OR, 0.3; 95% CI, 0.01-7.4; p = 0.957), with the numbers available (Table 4).

Clinical Outcomes Scores and ROM

With the numbers available, we found no difference in the percentage of subjects who were symptomatic at any of the three times. Specifically, eight of 20 of the throwing shoulders were symptomatic at initial followup, six of 20 were symptomatic 6 years after retirement (OR, 1.6; 95% CI, 0.4-5.8; p = 0.688), and five of 18 were symptomatic 15 years after retirement (OR, 1.7; 95% CI, 0.4-6.8; p = 0.652) (Table 5).

T5
Table 5.:
Comparisons of clinical outcomes in throwing shoulders at different times

The absolute and relative Constant and Murley scores were lower in the throwing shoulders assessed during the career compared with the throwing shoulders examined 6 years after retirement (absolute Constant and Murley score: mean, 93 points [SD, 6 points] versus 98 points [SD, 3 points]; mean difference, 5 points [SD, 5 points]; 95% CI of the mean difference, 2.5-7.4; p < 0.001 and relative Constant and Murley score, mean, 96 points [SD, 6 points] versus 105 points [SD, 4 points]; mean difference, 10 points [SD, 6 points]; 95% CI of the mean difference, 7-12; p < 0.001) (Table 6). However all the differences presented here are below the typically reported MCID for the absolute Constant and Murley score, and so are unlikely to be clinically important.

T6
Table 6.:
Comparisons of clinical outcomes in throwing shoulders versus nonthrowing shoulders at each time

With the numbers available, we did not find a difference with respect to the subjective shoulder value between the three times. More shoulder pain in general was reported in the throwing shoulders compared with nonthrowing shoulders (13 [SD, 3] versus 15 [SD, 2]; mean difference, 1 [SD, 3]; 95% CI of the mean difference, 0-3; p = 0.047) during the active career of the players (Table 6). However, this difference was below the MCID for pain of 3 points for the VAS pain score, 0 to 15. After cessation of a professional handball career, pain levels were not different, with the numbers available, in the throwing and the nonthrowing shoulders 6 years after retirement (14 [SD, 2] versus 15 [SD, 1]; mean difference, 0 [SD, 1]; 95% CI of the mean difference, -0.3 to 1; p = 0.248) and 15 years after retirement (15 [SD, 1] versus 14 [SD, 2]; mean difference, 2 [SD. 2]; 95% CI of the mean difference, -1 to 1; p = 0.916).

Pain at initial followup positively correlated with disorders of the long head of the biceps seen at initial follow up (r = 0.57; p = 0.009), tendinopathies of the infraspinatus seen at initial followup (r = 0.48; p = 0.032), disorders of the subscapularis tendon at initial followup (r = 0.69; p < 0.001), greater tubercle edema at initial followup (r = 0.52; p = 0.018), and the Jobe relocation test at initial follow up (r = 0.52; p = 0.018) (Table 7).

T7
Table 7.:
Pain correlated with several clinical and radiographic findings

Pain 6 years after retirement positively correlated with tendinopathies of the infraspinatus (r = 0.51; p = 0.021), greater tubercle cysts (r = 0.5; p = 0.025), cartilaginous changes at the glenoid (r = 0.64; p = 0.002), and the Jobe relocation test (r = 0.50; p = 0.025) assessed at initial followup. It also correlated with greater tubercle edema (r = 0.5; p = 0.025).

External rotation in 90° abduction was increased in the throwing shoulder compared with the nonthrowing shoulder at all the times (initial followup: 102° [SD, 15°] versus 95° [SD, 10°]; mean difference, 8°; 95% CI, 2o-14o; p = 0.014; 6 years after retirement: 110° [SD, 24°] versus 97° [SD, 9°], mean difference, 14° [SD, 23°], [95% CI, 3°-25°], p < 0.001; 15 years after retirement: 99° [SD, 18°] versus 95° [SD, 13°], mean difference, 4° [SD, 15°], [95% CI, -3° to 12°], p = 0.026). Internal rotation was decreased in the throwing compared with the nonthrowing shoulders (during the career: 63° [SD, 24°] versus 68° [SD, 20°], mean difference, 5° [SD10°], [95% CI, 0°-10°], p = 0.036; 6 years after retirement: 71° [SD, 6°] versus 73° [SD, 6°], mean difference, 2° [SD, 3°], [95% CI, 1°-4°]; p = 0.007; and 15 years after retirement: 73° [SD, 4°] versus 75° [SD, 4°], mean difference, 3° [SD, 4°]°, [95% CI, 1°-5°], p = 0.021) .

Discussion

Elite overhead throwing athletes have a high prevalence of shoulder disorders during their careers. The natural history of these findings after cessation of the professional career remains unclear. This lack of knowledge and because pathologic changes poorly correlate with symptoms make it difficult to create a reliable treatment algorithm for shoulder disorders in these athletes. We therefore investigated the natural history of MRI observations and clinical findings in the shoulders of professional European handball players after retirement from the sport and sought to evaluate whether these findings get worse, remain unchanged, or disappear after cessation of the career. We found that a large proportion of findings identified in the shoulders of elite professional European handball players resolved after retirement from the sport. Shoulder scores in these players were rather high during and after finishing their career: ROM differences between throwing and nonthrowing shoulders persisted. The Constant and Murley score and pain improved slightly after retirement, although the improvements observed were below the MCID for the instruments we used, and thus the subjects may not have considered these improvements important.

Limitations

This study has several limitations. First, the number of investigated subjects is rather small and this was limited by the first study in which 30 professional players were examined. Furthermore, the loss to followup (four of 30 patients, 13%) and exclusion of two of 30 patients (7%) owing to previous shoulder surgery could result in overestimation of the function of the study group relative to what is going on with the group of 26 overall. Second, the athletes were evaluated at an average of 5 years before termination of their professional careers and not at the end of their career. Reevaluation at the time of retirement may have shown further differences between initial followup and 6 years after retirement. Third, to investigate the long-term effects of overhead throwing, we included a second cohort of players who were retired from professional European handball for an average of 15 years. For this group, no clinical or radiologic data from the time during their professional careers were available. Therefore, this limited our ability to compare the group that was assessed initially and 6 years after retirement with the 15 years after retirement group, thus limiting our conclusions regarding the evolution of disorders up to 15 years after a professional career. Fourth, only one observer (CWP) evaluated the MRI scans and no intraobserver reliability testing was performed. However, this observer was an experienced senior consultant in musculoskeletal radiology, having worked as a fellowship-trained radiologist with a focus on orthopaedic disorders for more than 18 years.

Not all MRI-observed findings are clinically important. Such findings may be present in the shoulders of asymptomatic volunteers. Furthermore, using outcome scores (Constant and Murley score and subjective shoulder value) which are validated for evaluation of a general population rather than for professional throwing athletes may not detect accurately differences in outcome in this cohort of throwing athletes. However, there are no outcome scores specifically validated for this group of athletes. Finally, our investigational review board did not allow MR arthrography. The MR without arthrography did not allow us to fully evaluate the possible intraarticular disorders.

Natural History of Findings on MRI

The number of abnormal findings per throwing shoulder did not decrease after termination of a professional career. While some of the pathologic changes decreased, degenerative changes of the cartilage, the acromioclavicular joint, and the long head of the biceps tendon were found more frequently after retirement. One of the most interesting MRI findings among the players while they were active and also after retirement were abnormal findings of the rotator cuff insertion. To our knowledge, this is the first study to show that tendinopathies and partial rotator cuff tears in professional overhead throwing athletes do not progress to full-thickness tears after retirement at mid-term followup. Matthews et al. [17] showed that the smaller the size of a rotator cuff tear, the better the regenerative potential. They observed that small rotator cuff tears show increased blood vessel proliferation and fibroblast cellularity when compared with larger tears. This might be a reason why our cohort, which only showed partial and no full-thickness tears, showed no progression to full-thickness tears after retirement. Maman et al. [16] obtained similar results with respect to the progression of partial rotator cuff tears in a young population. While full-thickness tears progressed in 52%, partial tears progressed in only 8% of shoulders. Furthermore, age was an important predictor for tear progression, with 54% of tears in patients older than 60 years progressing versus only 17% of tears progressing in patients younger than 60 years. These data are in agreement with data from our study, where all of our subjects were younger than 60 years. Investigating a nonathletic cohort with a mean age of 62 years (SD, 10 years), Keener et al. [13] reported tear progression in 44% of partial rotator cuff tears after a median followup of 5.1 years. However, their population included patients treated for degenerative tears, and they did not specifically investigate overhead activities; their findings may not apply as specifically as ours do to the throwing athlete.

Some authors found posterosuperior glenoid impingement to be a consistent finding in overhead throwing shoulders [10,11,19]. However, Burkhart et al. [2] reported that this form of impingement is physiologic since it also is found in nonthrowers. In our study, posterosuperior impingement was present only in the throwing shoulders and decreased after retirement, suggesting that this form of impingement is associated with overhead throwing and might not be as physiologic as proposed earlier.

Superolateral osteochondral abnormalities of the humeral head seen during the active professional career have been considered to be caused by posterosuperior impingement [12,20]. In our study, superolateral humeral head edema and cysts decreased after retirement in more than half of the subjects, and the throwing shoulders were affected more frequently. These defects were similar in nature to Hill-Sachs lesions but were situated more superiorly in the humeral head. Because none of our subjects reported anteroinferior shoulder dislocations, these changes may be attributed to the repetitive throwing motion combined with increased external rotation. The lack of power or endurance to make a scapular retraction and subsequently a scapular dyskinesis, which may lead to a conflict between the glenoid and the humeral cysts, has been suggested as a reason for superolateral humeral head edema and cysts [2,3]. This may explain why those defects partly decreased after retirement. Ganglion cysts greater than 5 mm and superolateral bony cysts of the greater tubercle also decreased from the initial followup to 6 years after retirement in many subjects. Similar to the other superolateral findings, the reason for this decrease could be found in the cessation of repetitive throwing after retirement. Resolution of these osseous changes may be another factor that might have contributed to the relatively good clinical outcome despite the total amount of findings not decreasing after retirement. Posterosuperior glenoid impingement and superolateral osteochondral abnormalities seem to represent typical throwing shoulder findings which resolve subsequent to the sports activity.

Clinical Outcomes Scores and ROM

Clinical outcome measured with the Constant and Murley score slightly improved from the initial followup to 6 years after retirement although some pathologic changes per shoulder did not decrease during this period. To define which of the many assessed clinical and radiologic factors are the main contributors to this improvement in the Constant and Murley score is challenging and will require a much greater sample size owing to the multiple radiologic and clinical factors that influence clinical outcome. To our knowledge, no studies exist investigating the correlation between the natural history of shoulder abnormalities and clinical outcome scores such as the Constant and Murley score. The correlation analysis of this study suggests that other than disorders of the long head of the biceps, rotator cuff tendinopathies of the infraspinatus, and disorders of the subscapularis, osseous and cartilaginous glenohumeral changes and posterosuperior intraarticular impingement may affect pain levels up to 6 years after retirement. One possible explanation why not more findings correlate with pain is that the physical activity could result in alteration of pain perception, suggesting that professional athletes are less sensitive to pain [22].

Connor et al. [5] examined the prevalence of rotator cuff tears in a similar cohort of asymptomatic professional overhead athletes during their careers. They observed that 40% of asymptomatic throwing shoulders showed partial- or full-thickness rotator cuff tears while the nonthrowing shoulders showed none. At a 5-year followup interview, none of these patients had any subjective symptoms such as pain or had undergone surgical treatment thus suggesting that there is no clear correlation between pain and rotator cuff tears in athletes. Dunn et al. [8] did not find a correlation between rotator cuff tear severity and pain severity. In contrast, Cadogan et al. [4] found that pain can accurately predict the presence of rotator cuff tears in a primary care cohort.

To our knowledge, the current study is the first to show that external rotation in 90° abduction does not decrease after cessation of repetitive external rotation stress up to 6 years after retirement. One possible reason could be osseous alterations such as increased retrotorsion of the humerus and retroversion of the glenoid. Data from the study by Reagan et al. [21] support this hypothesis. They found a correlation between external rotation in 90° abduction and retroversion of the humeral head. Furthermore, Borsa et al. [1] reported increased external rotation with decreased internal rotation, but no differences in anterior joint stiffness in professional baseball pitchers. Both groups concluded that the difference in motion between throwing and nonthrowing shoulders may be related more to the osseous anatomy than to soft tissue adaptations. This is also in line with findings of several studies that showed an increased retrotorsion of the humerus and increased retroversion of the glenoid in throwing shoulders of professional throwers [7,25,26].

Conclusions

The results of this study suggest that partial rotator cuff tears of overhead throwing athletes do not progress to full-thickness tears after retirement from sports and that pain poorly correlates with those tears. Superolateral osseous changes also resolve after termination of a professional European handball career. These former professional athletes show good to excellent clinical scores at long-term followup even though the number of findings per shoulder remains high and unchanged. Because many of these pathologic findings do not seem to progress after retirement, indications for surgical treatment should be evaluated with great caution. Further studies with a larger sample size are needed to confirm our findings.

Acknowledgement

We thank Lilianna Bolliger MSc ETH (Department of Orthopaedic Surgery and Traumatology, University of Bern, Inselspital, Bern, Switzerland) for performing the statistical analysis.

References

1. Borsa PA, Dover GC, Wilk KE, Reinold MM. Glenohumeral range of motion and stiffness in professional baseball pitchers. Med Sci Sports Exerc. 2006;38:21–26.
2. Burkhart SS, Morgan CD, Kibler WB. The disabled throwing shoulder: spectrum of pathology Part I: pathoanatomy and biomechanics. Arthroscopy. 2003;19:404–420.
3. Burkhart SS, Morgan CD, Kibler WB. The disabled throwing shoulder: spectrum of pathology Part III: The SICK scapula, scapular dyskinesis, the kinetic chain, and rehabilitation. Arthroscopy. 2003;19:641–661.
4. Cadogan A, McNair P, Laslett M, Hing W, Taylor S. Diagnostic accuracy of clinical examination features for identifying large rotator cuff tears in primary health care. J Man Manip Ther. 2013;21:148–159.
5. Connor PM, Banks DM, Tyson AB, Coumas JS, D'Alessandro DF. Magnetic resonance imaging of the asymptomatic shoulder of overhead athletes: a 5-year follow-up study. Am J Sports Med. 2003;31:724–727.
6. Constant CR, Murley AH. A clinical method of functional assessment of the shoulder. Clin Orthop Relat Res. 1987;214:160–164.
7. Crockett HC, Gross LB, Wilk KE, Schwartz ML, Reed J, O'Mara J, Reilly MT, Dugas JR, Meister K, Lyman S, Andrews JR. Osseous adaptation and range of motion at the glenohumeral joint in professional baseball pitchers. Am J Sports Med. 2002;30:20–26.
8. Dunn WR, Kuhn JE, Sanders R, An Q, Baumgarten KM, Bishop JY, Brophy RH, Carey JL, Holloway GB, Jones GL, Ma CB, Marx RG, McCarty EC, Poddar SK, Smith MV, Spencer EE, Vidal AF, Wolf BR, Wright RW. Symptoms of pain do not correlate with rotator cuff tear severity: a cross-sectional study of 393 patients with a symptomatic atraumatic full-thickness rotator cuff tear. J Bone Joint Surg Am. 2014;96:793–800.
9. Gilbart MK, Gerber C. Comparison of the subjective shoulder value and the Constant score. J Shoulder Elbow Surg. 2007;16:717–721.
10. Giombini A, Rossi F, Pettrone FA, Dragoni S. Posterosuperior glenoid rim impingement as a cause of shoulder pain in top level waterpolo players. J Sports Med Phys Fitness. 1997;37:273–278.
11. Halbrecht JL, Tirman P, Atkin D. Internal impingement of the shoulder: comparison of findings between the throwing and nonthrowing shoulders of college baseball players. Arthroscopy. 1999;15:253–258.
12. Jost B, Zumstein M, Pfirrmann CW, Zanetti M, Gerber C. MRI findings in throwing shoulders: abnormalities in professional handball players. Clin Orthop Relat Res. 2005;434:130–137.
13. Keener JD, Galatz LM, Teefey SA, Middleton WD, Steger-May K, Stobbs-Cucchi G, Patton R, Yamaguchi K. A prospective evaluation of survivorship of asymptomatic degenerative rotator cuff tears. J Bone Joint Surg Am. 2015;97:89–98.
14. Kukkonen J, Kauko T, Vahlberg T, Joukainen A, Aarimaa V. Investigating minimal clinically important difference for Constant score in patients undergoing rotator cuff surgery. J Shoulder Elbow Surg. 2013;22:1650–1655.
15. Langevoort G. Glenohumeral InstabilityIn Langevoort G, ed. Sports Medicine and Handball. Basel, Switzerland: Beckmann; 1996:39–44.
16. Maman E, Harris C, White L, Tomlinson G, Shashank M, Boynton E. Outcome of nonoperative treatment of symptomatic rotator cuff tears monitored by magnetic resonance imaging. J Bone Joint Surg Am. 2009;91:1898–1906.
17. Matthews TJ, Hand GC, Rees JL, Athanasou NA, Carr AJ. Pathology of the torn rotator cuff tendon. Reduction in potential for repair as tear size increases. J Bone Joint Surg Br. 2006;88:489–495.
18. Michener LA, Snyder AR, Leggin BG. Responsiveness of the numeric pain rating scale in patients with shoulder pain and the effect of surgical status. J Sport Rehabil. 2011;20:115–128.
19. Neumann CH, Holt RG, Steinbach LS, Jahnke AH Jr, Petersen SA. MR imaging of the shoulder: appearance of the supraspinatus tendon in asymptomatic volunteers. AJR Am J Roentgenol. 1992;158:1281–1287.
20. Paley KJ, Jobe FW, Pink MM, Kvitne RS, ElAttrache NS. Arthroscopic findings in the overhand throwing athlete: evidence for posterior internal impingement of the rotator cuff. Arthroscopy. 2000;16:35–40.
21. Reagan KM, Meister K, Horodyski MB, Werner DW, Carruthers C, Wilk K. Humeral retroversion and its relationship to glenohumeral rotation in the shoulder of college baseball players. Am J Sports Med. 2002;30:354–360.
22. Tesarz J, Schuster AK, Hartmann M, Gerhardt A, Eich W. Pain perception in athletes compared to normally active controls: a systematic review with meta-analysis. Pain. 2012;153:1253–1262.
23. Walch G, Boileau P, Noel E, Donell ST. Impingement of the deep surface of the supraspinatus tendon on the posterosuperior glenoid rim: an arthroscopic study. J Shoulder Elbow Surg. 1992;1:238–245.
24. Weishaupt D, Zanetti M, Tanner A, Gerber C, Hodler J. Lesions of the reflection pulley of the long biceps tendon. MR arthrographic findings. Invest Radiol. 1999;34:463–469.
25. Wyland DJ, Pill SG, Shanley E, Clark JC, Hawkins RJ, Noonan TJ, Kissenberth MJ, Thigpen CA. Bony adaptation of the proximal humerus and glenoid correlate within the throwing shoulder of professional baseball pitchers. Am J Sports Med. 2012;40:1858–1862.
26. Yamamoto N, Itoi E, Minagawa H, Urayama M, Saito H, Seki N, Iwase T, Kashiwaguchi S, Matsuura T. Why is the humeral retroversion of throwing athletes greater in dominant shoulders than in nondominant shoulders? J Shoulder Elbow Surg. 2006;15:571–575.
Copyright © 2018 by the Association of Bone and Joint Surgeons