Golf is a popular worldwide sport with rounds typically lasting between 3.5–6 hours [1,2]. While played by individuals of all ages, most golfers are unusually frequently adults of middle or older age [3]. Whilst golf has well-established benefits, including positive impacts on wellness, cardiorespiratory health and metabolic health profiles [4], the golf swing is a repetitive and strenuous motion, and injuries can occur [5]. A systematic review has reported that the elbow (24.9%), shoulder (18.6%) and lumbar spine (15.2%) are the sites most commonly injured in amateur golfers [6]. A range of shoulder pathologies have been reported in golfers [7-9], primarily affecting their lead shoulders during their swing [10].

Persons who play golf may also sustain injuries in everyday life that can impact their golf. Rotator cuff (RC) tears are a common shoulder pathology in middle-aged and older adults [11], and the RC in health has been shown in electromyographic analyses to be actively engaged in both lead (non-dominant) and trail (dominant) shoulders throughout the golf swing [12,13]. The involvement of the RC musculature in the golf swing is asymmetric, with the supraspinatus and infraspinatus muscles mostly engaged in the lead shoulder during acceleration and follow-through of the swing and the subscapularis primarily involved in the trail shoulder from the start of the downswing [12,13]. Whilst golfers can sustain non-golfing shoulder injuries in a similar manner to the wider population, it has been suggested that overuse syndromes account for 92% of RC tears in golfers [5].

Given the significant involvement of the RC in the golf swing, symptoms from RC tears can restrict patients’ ability to play golf at their usual premorbid level or even participate in golf at all [9]. RC repair is a procedure typically employed to address tears unsuitable for nonoperative management or for which nonoperative management has failed [14], and systematic reviews of the literature have reported that most patients are able to return to sport postoperatively [15]. However, lower rates of return have been reported in baseball and softball pitchers, which are sports with significant engagement of the shoulder, and rates of return to the same level of play for these patients have been reported to be as low as 38% [16]. However, there is a paucity of literature exploring the return to golf following RC repair despite the shoulder-specific demands of golf, and, as such, surgeons and physicians are unable to reliably counsel patients about this preoperatively.

The primary aim of this study was to determine whether golfers undergoing RC repair were able to return to golf postoperatively and to what level they returned to. Secondary aims included revealing the barriers to returning to golf and the factors associated with failing to return to golf.

Approval was obtained from the Research Ethics Committee, South East Scotland Research Ethics Service, Scotland (16/SS/0026) for analysis and publication of the presented data. Data collection was carried out in accordance with the general medical council guidelines for good clinical practice and the Declaration of Helsinki. Informed consent was obtained from all patients.

Patients were identified from a prospectively compiled elective shoulder surgery database held at the study centre. Patient-reported outcome measure (PROM) completion Quick Disabilities of the Arm, Shoulder and Hand (QuickDASH) was used to determine which patients were golfers (by them answering the sports module component). All golfers who underwent RC repair between 2012 and 2020 for whom outcomes were available over a period of ≥12 months postoperatively were included. Patients not consenting to follow-up, individuals aged <18 years and patients with associated ipsilateral bony injury were excluded. The decision to undergo concomitant subacromial decompression (SAD) was made on a case-by-case basis by the fellowship-trained specialist shoulder surgeon performing the RC repair according to the case history, clinical examination and intraoperative findings. For biceps pathology, tenodesis was performed where possible, but where the tendon was found to be of extremely poor quality intraoperatively, tenotomy was performed instead. Demographic data were collected preoperatively. Patients were retrospectively followed up with to collect PROMs in order to assess whether they had returned to golf following surgery and to determine their degree of involvement in golf and their postoperative performance level. RC tear size was determined using preoperative magnetic resonance imaging scans or intraoperative measurements. Tear size was then grouped according to the DeOrio and Cofield classification [17].

The most important findings from this study were as follows: (1) the rate of return to golf following RC repair was 76.3%, (2) most patients returned to golf with similar handicaps and frequency of play as their premorbid state, (3) patients who returned had greater quality of life outcome measures compared to those who did not return, and (4) persistent subjective shoulder dysfunction (78%) was the most common barrier preventing patients from returning to golf.

To our knowledge, this is the largest published series addressing golf outcomes for patients following RC repair, reporting a return to golf rate of 76.3%. Published rates of returning to golf following RC repair range from 60% to 100% [24,25], with smaller series and subgroups within published papers on return to sports documenting rates of >90% [25-29]. Whilst Vives et al. [9] reported an 89.7% return to golf rate among 29 patients undergoing RC repair, a larger series documented in conference proceedings calculated a 75.6% rate of return postoperatively, which is in keeping with the findings reported here [30]. The rate of return reported in our study is less than the 84.7% rate reported in a systematic review analysing involvement in a range of sports postoperatively [15], but concurs with the findings from a review showing that 79% of baseball and softball players returned to their sport following RC repair [16]. It has been highlighted that reaching extremes of the overhead range of motion is not as necessary for the golf swing compared to in throwing sports [9], but golf does involve significant engagement of the shoulder musculature throughout the swing and a throwing-type motion [13].

Whilst postoperative outcomes for pain and function did not significantly differ between patients who did and did not return to golf, quality of life measures (as assessed by EQ-VAS and EQ-5D-5L) were significantly greater in those who returned to golf. This is an association and does not imply causation, but the beneficial impacts of golf participation on health and wellness have been well-documented [4]. Patients who returned to golf were satisfied with their involvement in the game in 91.1% of cases, which may be reflected in that those returning were found to be able to play at a similar frequency as that during their premorbid state with similar performance levels (as indicated by handicap data) despite them being older. A systematic review investigating golf outcomes following shoulder arthroplasty found similarly maintained performance outcomes (according to both handicap change and driving distance) [31]. Reduced postoperative performance levels in elite athletes following RC repair have been observed in professional tennis and baseball players [32,33], but the golfers in this population are primarily recreational rather than professional players. Given the lack of evidence in the literature addressing outcomes among elite golfers, it is difficult to determine whether the outcomes observed in this study would be similarly applicable to this patient group. Despite the high rates of satisfaction regarding postoperative golf involvement observed in this study, 37.9% of patients experienced at least mild pain during play. However, it has been reported that 47% of “healthy” athletes experience shoulder pain during full performance, which may account for these findings [34].

Whilst most reasons for failure to return to golf were related to shoulder pain, weakness and stiffness (Table 1), two patients cited time out from the sport as their reason for non-return. As 86.8% of patients were unable to play preoperatively due to shoulder disability, it remains unclear how long patients are spending away from play due to RC tear symptoms. Orthopaedic waiting lists have grown significantly in recent years, and streamlining this process could serve to reduce the duration of preoperative morbidity prior to shoulder surgery [35,36]. In addition, golf-specific rehabilitation protocols have been documented [37], which could serve to optimise return to golf timelines and performance, although adequately powered studies with control groups will be required to determine their benefit.

Preoperative function, demographics and golfing variables were not associated with rates of return to golf postoperatively. A meta-analysis of the RC repair literature reported that patient age, the involvement of multiple structures, tear size, diabetes mellitus, cuff fatty infiltration, preoperative strength and insurance status all affect outcomes following RC repair [38]. Biceps involvement has also been reported to be a predictor of poor outcomes following cuff repair [39]. Similarly, in this study, patients who underwent RC repair with biceps tenotomy were significantly less likely to return to golf compared to those not who did not undergo biceps tenotomy. There was no association found between RC tear size and return to golf, although this trend may be accounted for by the lack of a broad range of tear sizes evident in the study population, with 60.5% of tears measuring 1–3 cm in size. The golf swing is asymmetric in nature, and the lead and trail shoulders engage differing components of the RC to different extents throughout the swing [12,13]. Despite the differing strains across the lead and trail shoulders, shoulder dominance did not impact the rate of return to golf in this study, in keeping with the findings reported by Vives et al. [9] However, electromyography has shown that the supraspinatus and infraspinatus are more engaged in the lead shoulder, while the subscapularis is more engaged in the trail shoulder [12], and study size prevented us from investigating the interplay between shoulder dominance, the individual muscles that were torn and golf performance postoperatively.

This study should be interpreted considering its limitations, including those commonly seen in retrospective studies, such as inconsistent reporting of information, loss to follow-up and lack of a control group. However, loss to follow-up was limited to 19.9% of the population, which is less than the benchmark used for quality assessments in retrospective series [40]. Additionally, patients lost to follow-up had similar baseline demographics and underwent similar procedures compared to those not lost to follow-up. The patient population consisted predominantly of men, although this is reflective of the overall golfing demographic. The limited sample size predisposes the findings to type 2 error. Patients with clinical examination findings suggestive of an intact RC repair did not undergo any direct radiological assessment of RC integrity at final follow-up. Potentially, some of these patients may have experienced a re-tear of their repair, although this has been shown to not correlate with functional outcomes [41,42]. In addition, various additional procedures were performed, and other markers of cuff tear severity out with the number of structures requiring repair and the indication for additional procedures were not measured. A further limitation existed in that handicap was used as a proxy measure for performance, but further metrics—including shot distances, club speed and accuracy—were not recorded. Furthermore, it has been proposed that up to 92% of RC tears in golfers are secondary to overuse syndromes [5]. However, this study reports outcomes for a high proportion of traumatic RC tears. This may be because this study focuses solely on RC tears requiring surgical repair, discounting those for which nonoperative management is most suitable. Finally, this study did not explore the impact of length of golf career or time away from the game on patient outcomes, nor did it assess the timing of return to golf postoperatively.

Golfers were likely (76%) to return to golf following RC repair, which was associated with a greater health-related quality of life, to their premorbid performance level with little residual symptomatology. However, persistent subjective shoulder dysfunction (78%) was the most common barrier preventing patients from returning to golf.