Strength testing, especially of the abducted shoulder, aids in diagnosis of rotator cuff tears: a study of senior athletes participating in upper extremity dominant sports.

BACKGROUND: Small supraspinatus tears can be difficult to diagnose on history and physical examination alone. Normalized strength assessment may aid in diagnosis. We examined shoulder strength normalized to body weight and the nondominant shoulder as well as abduction (AB)/external rotation (ER) and ER/internal rotation strength ratios in senior athletes to determine if they aid in the diagnosis of supraspinatus tears.

METHODS: Ultrasound imaging of the dominant shoulder's rotator cuff was performed in 79 athletes over the age of 60 who were participating in an upper extremity dominant sport. The supraspinatus tendon condition was categorized into 3 groups: 1) intact (normal/tendinosis), 2) partial-thickness tear, and 3) full-thickness tear. Shoulder strength was assessed with a handheld dynamometer and normalized to body weight, the nondominant shoulder or as AB/ER and ER/internal rotation strength ratios. Receiver-operating characteristic (ROC) curves assessed the accuracy of cut points in the strength data for discriminating full tears from partial or no tears. We hypothesized there would be diminished shoulder strength for glenohumeral AB and ER, when normalized to body weight, in athletes with full-thickness supraspinatus tendon tears compared to those with partial-thickness supraspinatus tears or an intact tendon.

RESULTS: No differences among the 3 supraspinatus classifications were found for shoulder strength normalized to the nondominant shoulder or for shoulder strength ratios. However, senior athletes with full-thickness supraspinatus tendon tears had significantly less strength as a percentage of body weight for shoulder AB at 90° and shoulder ER at 90° of AB compared to athletes with intact tendons. ROC curve analysis revealed the best strength test for discriminating a full-thickness supraspinatus tendon tear from an intact tendon was strength of ER at 90° of AB. The ROC curve analyses found a full-thickness supraspinatus tear could be correctly detected 76% of the time when a strength cut point of 12% of bodyweight for ER strength at 90° of AB was used to categorize subjects as having an intact or torn tendon.

CONCLUSION: No differences among the 3 supraspinatus classifications were found for shoulder strength normalized to the nondominant shoulder or for shoulder strength ratios. However, senior athletes with full-thickness supraspinatus tendon tears had significantly less strength as a percentage of body weight for shoulder AB at 90° and shoulder ER at 90° of AB compared to athletes with intact tendons. ROC curve analysis revealed the best strength test for discriminating a full-thickness supraspinatus tendon tear from an intact tendon was strength of ER at 90° of AB. The ROC curve analyses found a full-thickness supraspinatus tear could be correctly detected 76% of the time when a strength cut point of 12% of bodyweight for ER strength at 90° of AB was used to categorize subjects as having an intact or torn tendon.