According to the Korean Gold Research Report in 2014, there are an estimated 5,310,000 Korean golfers. The size of this population ranks fifth worldwide and has seen a rapid increase recently. With such a large number of people playing golf, cases of golf-related pain are also increasing. Both professional and amateur golfers are reporting large numbers of golf-related injuries. The estimated prevalence of golf-related injuries ranges from as low as 25% to as high as 62% [1,2]. The most commonly injured body part in amateur golfers is the upper limbs, followed by the vertebrae and then the lower limbs [2,3]. Although different studies suggest different results, the shoulder is a commonly injured body part in golfers, after the elbow and the wrist. The most common cause of shoulder injury, including rotator cuff tear, is known to be impingement syndrome [4]. A previous study [5] reported that of 412 patients with golf-related injuries, shoulder injury was observed in 85 patients, 79 of whom had rotator cuff tears or impingement. Although shoulder injuries comprise a large portion of golf-related injuries, there are few studies on golf-related shoulder injuries, whereas golf-related elbow, wrist, and back pain are being actively researched. In particular, there are few studies on golf-related shoulder injuries in amateur golfers, who account for most of the golfing population in Korea.

Since its introduction as a method for assessing rotator cuff tears, ultrasound has been used as an important diagnostic tool. Ultrasound, a non-invasive procedure that requires a short examination time, is relatively inexpensive and does not produce radiation [6]. Because of many improvements in this technology, ultrasound currently has a high sensitivity and specificity, comparable to those of magnetic resonance imaging. Ultrasound allows for dynamic examinations and simultaneous comparison of both sides of the body, making it a highly valuable diagnostic tool [7].

In this study, golf-related surveys were administered to, and ultrasounds were conducted on, amateur golfers with golf-related shoulder pain to investigate the causes and characteristics of shoulder injuries.

With the increase in the population of golfers, interest in golf-related injuries is also increasing. Golf-related injuries are usually caused by insufficient warm-up, excessive nervousness, and lack of concentration [9]. Injuries in amateur golfers are usually caused by a lack of skills, whereas injuries in experts are mostly caused by overuse of muscles [10]. Although the prevalence of golf-related shoulder injuries is not high, their onset can greatly affect a golfer's performance.

Owing to the large proportion of right-handed persons in the general population, 94.8% of the participants had a right swing direction. In participants with a right swing direction, the right shoulder was considered the dominant shoulder and the left shoulder was considered the non-dominant shoulder. Previous studies [11] reported that non-dominant shoulders, which lead the swing, are at a relatively higher risk of injury during a radical swing. It has also been reported [12,13] that golf-related shoulder pain occurs mostly in the non-dominant shoulder. The results of the present study also suggest that the proportion of participants with pain in the non-dominant shoulder was significantly higher than the proportion of those with pain in the dominant shoulder. In addition, the frequency of abnormal ultrasonographic findings was significantly higher in the non-dominant shoulder. However, 5 of the 67 cases with abnormal ultrasonographic findings (7.5%) presented with no shoulder pain, which indicates that ultrasonographic confirmation is necessary even when no pain is present.

Of the total of 67 cases, abnormal ultrasonographic findings indicated that rotator cuff tear was present in 60 cases. Supraspinatus muscle tear was especially prevalent, accounting for 47 cases. Rotator cuff tear usually occurs when the supraspinatus muscle moves below the acromion and causes impingement. Small tears in the supraspinatus muscle extend from the anterior distal joint to the entire supraspinatus and expand to other rotator cuff muscles, including the infraspinatus [14]. Impingement caused by shoulder overuse is the major cause of supraspinatus muscle tear. Impingement-prone conditions are produced when an overhead motion is forced in the dominant shoulder during address and back swing and in the non-dominant shoulder during follow-through. Small tears are produced with repetition of rapid swings, which then progress to supraspinatus muscle tears. Supraspinatus muscle full thickness tears were observed only in cases with isolated supraspinatus findings in the non-dominant shoulder. It can be inferred that supraspinatus full thickness tears were produced by the golfers' radical swing movements during follow-through.

Thirteen cases presented with no supraspinatus muscle tears. Subscapularis muscle tear was the most prevalent of these injuries, and full thickness tears were also observed. The subscapularis muscle is usually the most active rotator cuff muscle during golf swings [13]. The subscapularis muscle is known to produce a concentric force on the dominant shoulder during down swing and an eccentric force on the non-dominant shoulder during follow-through, which are important for swing breaking [14]. Therefore, excessive swing movement and the shock caused by the contact of the club with the ground during down swing and follow-through could cause damage to the subscapularis muscle. Infraspinatus full thickness tears and a combination of subscapularis and infraspinatus partial thickness tears were also observed in some cases. These injuries could have been caused by trauma without excessive use of the rotator cuff. Very few studies have been conducted on whether muscle overuse or trauma is a major cause of golf-related injuries. It has been reported [15] that overuse accounts for 54.5% of injuries, whereas trauma accounts for 45.5% of injuries. Thus, the likelihood that many injuries are caused by trauma cannot be disregarded.

According to our findings, subacromial-subdeltoid bursitis, acromioclavicular osteoarthritis, and biceps tenosynovitis are observed as accompanying findings in patients with rotator cuff tears. Subacromial-subdeltoid bursitis was only observed in cases with supraspinatus muscle tears. This result indicates that the cause of subacromial-subdeltoid bursitis could also be repetitive overuse and could be a secondary effect of impingement. Acromioclavicular osteoarthritis was the sole finding in 3 cases without rotator cuff tears. Similar to subscapularis and infraspinatus tear, acromioclavicular osteoarthritis may be caused by trauma. Biceps tenosynovitis was observed in 2 cases in the group with supraspinatus tears in the non-dominant shoulder and 1 case in the group without supraspinatus tears in the non-dominant shoulder. It can be deduced that biceps tenosynovitis is caused by excessive flexion of the non-dominant elbow during follow-through.

A study [15] that investigated golf-related habits between amateur golfers with and without golf-related injuries reported that, although there were no differences in age and length of golfing experience, the injured group underwent more golf rounds or practices per week. In addition, the injured group experienced more fatigue after a golf round or practice session. In this study, there was no control group composed of uninjured amateur golfers. Therefore, an accurate comparison was not possible. The results of this study, which was confined to injured participants, showed that age, golf experience, and number of golf practices per week were not correlated with shoulder injuries. However, participants who spent on average 4 hours or more per week practicing golf showed a significant association with the supraspinatus muscle tear group, which concurs with the results of previous studies [15]. Taking into account that supraspinatus tears are usually caused by repetitive muscle overuse, the amount of time spent practicing golf could reasonably be a more influential factor in shoulder injuries than the number of golf practice sessions.

Each golfer's personal preferences and habits cause force exertions on different body parts during the swing. Therefore, the location of pain and painful swing phases are different in each golfer. The muscles that are responsible for each swing phase have been identified in previous studies. The locations of injuries and abnormal findings are different in each swing phase [15]. However, the results of this study showed no characteristic abnormal ultrasonographic findings related to address and back swing, forward swing and impact, or follow-through. These results may be attributable to the method of evaluation of pain in different swings, which was only done through subjective surveys and did not involve meticulous clinical evaluations such as motion analysis. Accordingly, a future study using meticulous evaluation tools such as motion analyzers is necessary.

No abnormal ultrasonographic findings were observed in 20 of 87 cases (23.0%), even though these participants complained of pain. The pain in these cases could have been localized to the soft tissues of the trapezius and the levator scapulae [16], or caused by shoulder instability, such as from damage due to a superior labral tear from the anterior to the posterior direction [17]; however, such damage is uncommon. Further examinations would be necessary to confirm such cases.

No differences were observed in the proportion of rotator cuff tears between the group with additional upper limb pain in the elbow or the wrist and the group without additional upper limb pain. Rotator cuff tears probably occur independently of other upper limb injuries. However, the fact that 50 participants (64.9%) had additional upper limb pain indicates that the possibility of additional upper limb injuries should be considered in patients with golf-related shoulder pain. The group with a history of clinic visits had a relatively low proportion of abnormal ultrasonographic findings. It can be concluded that early treatment of golf-related shoulder injuries may delay aggravation and progression of damage. A previous study [18] reported that 39% of the participants had previously been treated for golf-related shoulder pain at healthcare institutions. Only 41.6% of the participants in this study reported a history of clinic visits owing to golf-related injuries, which may indicate that golfers are not aware that golf-related injuries require clinic visits. Therefore, healthcare visits at the time of onset of shoulder pain should be promoted among golfers.

This study has some limitations. First, it was conducted without a control group composed of amateur golfers without shoulder pain. Second, although the participants were divided according to sex before the investigation, the small number of female participants did not allow for a statistical comparison between the sexes. Third, although ultrasonographic examination is a very accurate screening tool, lack of other imaging techniques such as arthrography did not allow confirmation of the findings. Fourth, golf-related factors such as type of golf lesson, accuracy, instructor level, and time of lesson were not addressed. Lastly, the small number of participants does not reflect the entirety of the amateur golfer population and the recruitment notice may have only attracted participants with severe shoulder pain. The limitations described above should be addressed in future large-scale studies to accurately determine the characteristics of golf-related shoulder injuries.

Golf is a non-contact sport that involves striking an immobile ball with a club. Compared with other sports, golf is a low-intensity activity; therefore, golf-related injuries and their prevention are often neglected. However, golfers are at a moderate risk for injuries [19] and the number of golf-related injuries, including shoulder injuries, is expected to increase with the increasing number of amateur golfers. The results of this study revealed that golf-related shoulder injuries most frequently occur in the non-dominant shoulder and multiple abnormal findings are usually observed. The supraspinatus muscle was much more frequently damaged than were other muscles. A correlation was observed between the duration of golf practice and supraspinatus tears. On the basis of these results, patients evaluated for golf-related shoulder pain should be advised on the appropriate amount of golf practice and should undergo ultrasonographic evaluation.