Return to yoga following anatomic or reverse total shoulder arthroplasty

Abstract

Background

Returning to sport is an important consideration for joint replacement. Currently, there is a lack of knowledge regarding returning to yoga following shoulder arthroplasty. The purpose of this study was to examine patients’ ability to return to yoga after primary total shoulder arthroplasty. Secondary outcome measures were analyzed including range of motion (ROM) and patient-reported outcomes (PROs). Additionally, any differences between anatomic total shoulder arthroplasty (aTSA) and reverse total shoulder arthroplasty (rTSA) were investigated.

Methods

A retrospective study was performed on patients who underwent primary aTSA or rTSA between 2012 and 2022. Inclusion criterion was a minimum follow-up of 2 years, and yoga was reported as their primary sporting activity prior to arthroplasty. Yoga participation was recorded before and after shoulder arthroplasty, as well as time to return to previous activity level and satisfaction. Range of motion including forward flexion, external rotation, and internal rotation, as well as PROs including visual analog scale for pain, American Shoulder and Elbow Surgeons score, and subjective shoulder value were collected preoperatively and postoperatively. Postoperative results were compared based on patients’ ability to return to participation in yoga.

Results

Fifteen patients (6 aTSA and 9 rTSA) with a mean age of 69 ± 6 years met the study criteria and were evaluated at a mean of 44 ± 22 months postoperative. At 6 months postoperative, 60% (9/15) of patients reported being able to return to their activity, whereas 100% (15/15) had returned by the 12-month postoperative point. At the final follow-up, 80% (12/15) of patients noted that their ability to return to yoga following shoulder arthroplasty improved compared to baseline, 13% (12/15) reported that this ability was unchanged, and 7% (1/15) reported decreased ability.

There was no statistically significant difference in PROs between rTSA and aTSA (visual analog scale P = .08, American Shoulder and Elbow Surgeons score P = .79, subjective shoulder value P = .29). Likewise, there was no statistically significant difference in postoperative ROM when comparing rTSA and aTSA groups (forward flexion P = .47, external rotation P = .22, internal rotation P = .29).

Conclusion

Patients who engage in yoga have a high return to sport rate following either aTSA or rTSA, although full return may take up to 1 year postoperatively. The vast majority of patients also report improvement in their ability to perform yoga following shoulder arthroplasty. Among yoga participants, postoperative ROM, including internal rotation, appears to be similar between aTSA and rTSA.

Yoga originates from ancient India and encompasses several spiritual disciplines, including Hinduism, Buddhism, and Jainism.^2,21,24 The main goal of the practice of yoga is to center the body and mind, detaching from mundane suffering. While some of the earliest mentions of yoga are found between 1200 and 900 BCE,²⁷ modern yoga is practiced worldwide. Specifically in the Western regions, “yoga” describes a form of the practice that is focused on posture-based physical fitness, stress relief, and relaxation.¹ The practice of yoga is so common that typical yoga poses like downward dog, warrior pose, and tree pose are within the realm of familiar vocabulary of Western societies.

Many yoga poses and transitions place stress on the shoulder joint. For example, downward dog is characterized by one placing their weight equally across the palms of their hands and their feet, with their head down and hips bent in the air as high as possible.¹² Fishman et al noted that the shoulder is one of the most stressed joints in yoga practice.⁶ Cramer et al reported that the most common reported yoga practices in Germany were associated with acute injury resulting from hand, shoulder, and headstands in their cross-sectional survey of 1702 yoga practitioners.³ With such stress placed on the shoulder joint, individuals with shoulder pain who practice yoga may be required to modify many yoga poses.

As patients remain more active even with advancing age, the desire to return to strenuous activities following shoulder arthroplasty is increasing.^9,10 Although traditionally a young person’s sport, yoga has become increasingly popular among older populations. An important consideration for deciding on surgical intervention is a patient’s quality of life and their ability, or lack thereof, to resume the activities or sports that they engaged in preoperatively.^14,22 The activities or sports that patients engage in vary widely, and while studies have been conducted relating yoga and improvements in adhesive capsulitis,¹⁹ there is a lack of investigation into patients’ ability to return to yoga following shoulder arthroplasty. The purpose of this study was to examine return to yoga after primary shoulder arthroplasty at a minimum 2-year follow-up. Secondary outcome measures were also analyzed including range of motion (ROM) and patient-reported outcomes (PROs). Additionally, any differences between anatomical total shoulder arthroplasty (aTSA) and reverse total shoulder arthroplasty (rTSA) were investigated. The hypothesis was that return to yoga would be high following shoulder arthroplasty.

Methods

Study design

A retrospective review of a prospectively maintained database was queried for all patients who underwent primary rTSA or aTSA between January 2012 and January 2022 at a single institution. Inclusion criteria were any primary shoulder arthroplasty patients who reported yoga as their primary sporting activity prior to arthroplasty and a minimum follow-up of 2 years. Exclusion criteria were revision arthroplasty, proximal humerus fracture sequela, history of fracture or nerve injuries, incomplete follow-up, other sports reported as their primary activity, and those who did not agree to participate.

Surgical technique and postoperative rehabilitation

All procedures were performed by the senior author (P.J.D.). A standard deltopectoral approach was used in all cases. Subscapularis management differed depending on the type of arthroplasty. A lesser tuberosity osteotomy was performed for aTSA while a subscapularis peel was utilized for rTSA. For aTSA, an all-polyethylene pegged glenoid and a short stem (Univers Apex; Arthrex, Inc., Naples, FL, USA) or stemless humeral component (Eclipse; Arthrex, Inc.). For rTSA, a 135° inlay component (Univers Revers; Arthrex, Inc., Naples, FL, USA) was placed with a lateralized glenoid. The subscapularis was repaired in all cases. Postoperatively, patients wore a shoulder immobilizer for 4 weeks, after which slow progression of activities with minimal weight bearing was encouraged. Strengthening was allowed at 8 weeks.

Yoga questionnaire

During the study period, a total of 1204 primary shoulder arthroplasties were performed. All patients were contacted via email to assess participation in yoga. If a patient did not fill out the survey after 3 contact attempts, they were considered lost to follow-up. A questionnaire was designed to explore patients’ preoperative and postoperative athletic activity and ability to participate in yoga (Supplementary Appendix SA). Questions were close-ended multiple-choice questions that required each subject to respond before proceeding to the next item. Incorporated into the survey, patients were asked to specify: (1) their participation status both preoperatively and postoperatively; (2) their capacity to resume yoga with the same level of intensity, frequency, and duration postoperatively; and (3) the timeline of their postoperative return to yoga. In addition, there were questions on sport-specific considerations, such as yoga type and specific poses practiced by each patient.

Study variables

Preoperative diagnoses, implant type, age, sex, body mass index (BMI), type 2 diabetes mellitus, tobacco use, and length of follow-up were collected. Active ROM and PROs were documented at baseline and postoperatively. ROM measurements were recorded by the treating surgeon (P.J.D.) or a physician assistant at the initial and a 2-year clinic visit. These included forward flexion, external rotation at the side, and internal rotation (IR), which was numerically scaled based on the nearest spinal level achieved with the thumb (ie, T10 = 10, T12 = 12, L2 = 14, L4 = 16, S1 = 18, hip = 20). Preoperative and postoperative PROs, including American Shoulder and Elbow Surgeons score, visual analog scale for pain, and subjective shoulder value, were collected from electronic medical records or the emailed questionnaire if not available from the 2-year clinic visit.

Statistical analysis

Categorical and continuous variables were identified for analysis. Categorical variables were reported as fractions and percentages, whereas continuous variables were reported as means and standard deviations. Questionnaire responses were displayed descriptively for all patients and per sport. A paired t-test was used to assess for clinical improvement for patients overall as well as for aTSA and rTSA independently. A P value threshold of <.05 was used to denote statistical significance.

Results

Baseline demographics

Fifteen patients (6 aTSA and 9 rTSA) with a mean age of 69 ± 6 years met the study criteria and were evaluated at a mean of 44 ± 21 months postoperatively. Baseline demographics are represented in Table I. Most patients (87%, 13/15) were female. PROs and ROM improved significantly in all patients (Table II). There was no statistically significant difference among aTSA vs. rTSA when comparing postoperative PROs or ROM.

Table I.

Baseline demographics for all patients.

Demographic	Overall (15)	TSA (6)	rTSA (9)
Operated shoulder (right, %)	9 (60%)	5 (83%)	4 (44%)
Dominant arm (n, %)	9 (60%)	5 (83%)	4 (44%)
Age at surgery, y (mean, SD)	69 ± 6	69 ± 6	69 ± 6
Follow-up, mo (mean, SD)	44 ± 21	43 ± 21	44 ± 22
Male (n, %)	2 (13%)	2 (33%)	0 (0%)
BMI (mean, SD)	27 ± 7	27 ± 7	27 ± 7

BMI, body mass index; SD, standard deviation; TSA, total shoulder arthroplasty; rTSA, reverse total shoulder arthroplasty.

Table II.

Comparison of functional and patient-reported outcomes.

	Overall preoperative (mean, SD)	Overall postoperative (mean, SD)	Overall improvement (mean, 95% CL)	P value	TSA preoperative (mean, SD)	TSA postoperative (mean, SD)	TSA improvement (mean, 95% CL)	rTSA preoperative (mean, SD)	rTSA postoperative (mean, SD)	rTSA improvement (mean, 95% CL)
ROM
FF (degrees)	90 ± 21	168 ± 25	69 (55-82)	<.01	99 ± 21	168 ± 25	57 (40-74)	101 ± 22	171 ± 23	77 (57-96)
ER (degrees)	23 ± 23	82 ± 14	59 (46-73)	<.01	23 ± 23	82 ± 14	57 (28-85)	25 ± 24	81 ± 14	61 (43-79)
IR (spinal level)	S1 ± 2	L4 ± 3	2.4 (0.6-4.2)	.01	S1 ± 2	L4 ± 3	3 (0.2-5)	S1 ± 2	L4 ± 3	2.2 (0.5-5.0)
PROs
VAS	6.5 ± 1.2	0.6 ± 1.0	6.1 (5.1-7.1)	<.01	6.5 ± 1.2	0.5 ± 1.1	6.5 (5.1-8.0)	6.5 ± 1.2	0.5 ± 0.9	6.3 (5.2-7.2)
SSV	46 ± 23	89 ± 10	45 (32-58)	<.01	46 ± 23	89 ± 10	32 (8-57)	46 ± 24	90 ± 9	58 (39-76)
ASES	39 ± 12	89 ± 12	52 (42-62)	<.01	39 ± 12	89 ± 12	37 (19-55)	39 ± 13	90 ± 11	47 (22-71)

95% CL, confidence level of 95%; TSA, total shoulder arthroplasty; rTSA, reverse total shoulder arthroplasty; ROM, range of motion; FF, forward flexion; ER, external rotation; IR, internal rotation; PROs, patient-reported outcomes; VAS, visual analog scale; SSV, Subjective Shoulder Value; ASES, American Shoulder and Elbow Surgeons score; SD, standard deviation.

Internal rotation conversions: T10 = 10, T12 = 12, L2 = 14, L4 = 16, S1 = 18, hip = 20.

Yoga outcomes: all arthroplasties

Nine overall patients (60%, 9/15) reported being able to return to yoga within 6 months: 7 of these were rTSA patients (7/9, 78%), and 2 were aTSA patients (2/6, 33%). By the 12-month postoperative point, 100% (15/15) were able to return. There was a 93% (14/15) overall satisfaction rating from yoga participants, with 7% (1/15) of yoga participants reporting neutral satisfaction and 0% dissatisfaction.

Over half of all patients (53%, 8/15) reported returning to practicing yoga 2 to 4 times per week following their shoulder replacement, whereas 27% (4/15) reported participation at 1 to 2 times per week, and 20% (3/15) reported participation less than once a week. Eighty percent (12/15) of patients noted that their ability to engage in yoga following their shoulder arthroplasty improved compared to baseline, 13% (2/15) reported that this ability was unchanged, and 7% (1/15) reported decreased ability. Return rates, satisfaction, and improvement for both groups are shown in Table III.

Table III.

Rates of return, satisfaction, and improvement reported by TSA and rTSA patients.

	TSA (%) n = 6	rTSA (%) n = 9	P value
Rate of return
0-6 mo	33	78	.36
6-12 mo	66	22	.36
Satisfaction
Very dissatisfied	0	0	N/A
Dissatisfied	0	0	N/A
Neutral	17	0	.36
Satisfied	50	0	.08
Very satisfied	33	100	.03
Improvement
Worsened	17	0	.36
Stayed the same	33	0	.17
Improved	50	100	.17

TSA, total shoulder arthroplasty; rTSA, reverse total shoulder arthroplasty.

P values reported as N/A indicate no differences between operative methods.

Yoga outcomes: rTSA

One hundred percent (9/9) of rTSA patients reported that their ability to practice yoga improved following surgery and reported overall satisfaction. Frequency of yoga practice increased postoperatively, with 89% (8/9) engaging 2 or less times per week prior to their operation and 100% (9/9) engaging in 2 or more times per week postoperatively. The type of yoga rTSA patients engaged in did not change from preoperative to postoperative (Hatha: 11%, 1/9, restorative: 89%, 8/9). Prevalence of each yoga pose increased postoperatively compared to baseline following rTSA, with standing poses having increased 22%, forward bends increasing 44%, backbends increasing 23%, twists increasing 56%, inversions increasing 56%, and arm balances increasing 56%.

Yoga outcomes: aTSA

Of the aTSA patients, 3 reported that their ability to practice yoga following their operation improved (50%), while 33% (2/6) reported that their ability was unchanged and 17% (1/6) reported a decreased ability to practice yoga. Overall satisfaction in the aTSA group was 83% (5/6), with 1 patient reporting neutral satisfaction. Frequency of yoga decreased following aTSA, with 50% (3/6) practicing less than once a week postoperatively. Compared to preoperative, the incidence of standing poses decreased by 33%, forward bends increased by 34%, backbends remained the same at 33%, twists increased by 16%, and arm balances decreased by 16%. The type of yoga performed postoperatively was mostly restorative (83%, 5/6), which was the same yoga type mostly performed prior to aTSA (67%, 4/6), though the practice of Hatha style yoga decreased by 16% following aTSA.

Discussion

The most important finding of the current study was the rate of return to yoga following shoulder arthroplasty. In our series, 100% of patients returned to yoga by 1-year follow-up after shoulder arthroplasty, with 93% being overall satisfied with their postoperative performance. All patients experienced significant improvements in ROM and PROs compared to baseline.

Returning to sports is generally high following shoulder arthroplasty. A review including 944 patients of recreational and competitive sports performed by Liu et al reported an 85.1% return to sport rate following shoulder arthroplasty at an average follow-up of 5.1 years, with those undergoing aTSA returning at an increased rate compared to those who underwent rTSA.²⁰ In the current study, we evaluated yoga specifically and noted a 100% return. The difference in return rate reported by this review and our study may be due to the differences in each study’s demand on shoulder use.

Sports are commonly categorized by the stresses and demands they place on the shoulder as low, medium, and high demands. Examples of low- to medium-demand sports are golf, Pilates, and yoga, as they do not include heavy lifting or repetitive stress, whereas high-demand sports are those such as baseball/softball, climbing, tennis, and volleyball, as they accompany heavier lifting and above-shoulder aactivity.^8,23 Sports that place increased demand on the shoulder joint are more difficult for patients to return to at the same level at which they practiced preoperatively^8,23 due to things such as pain, issues related to surgery, and loss of interest.¹³ Further, Geyer et al reported increased incidence of radiolucency around arthroplasty components in a population considered to engage in higher-demand activities.⁸ Promisingly, there is a high return-to-sport rate associated with sports that are considered to place lower demand on the shoulder.^7,17 Tramer et al investigated return to golf following rTSA and reported overall satisfaction and an ability to return to golf with a small risk of lowering driving distances due to altered mechanics.²⁵ Given yoga’s low-intensity nature, data on return to yoga are more limited in comparison.^5,8 Tangtiphaiboontana et al retrospectively reviewed data from 109 patients regarding their ability to return to sports after primary rTSA.²³ Though the sample size of individuals who engaged in yoga was small (n = 3), they reported a 100% rate of return. Similarly, our study supports this 100% return to yoga at 1 year with all 15 patients. Fourteen of 15 patients returned at the same or an improved level compared to preoperatively.

Although the number in the current study is small, subtle differences may be present in return to yoga following either aTSA or rTSA. Patients who underwent rTSA were able to engage in yoga 2 or more times per week more commonly compared to aTSA patients (P = .03). Additionally, many types of yoga poses were not practiced following aTSA, whereas all rTSA patients increased what types of poses they were able to perform following their operation. The type of yoga practices (restorative, Hatha, Vinyasa, etc) did not change in frequency for either operative method from preoperative to postoperative (67% aTSA, 89% rTSA), though more rTSA patients were able to return to yoga within the 6-month postoperative point (7/9, 78%) compared to aTSA (2/6, 33%). One explanation for the broader range of positions and earlier return to yoga following rTSA could be the semiconstrained nature of rTSA, which may provide additional stability during yoga poses.

Previous studies have reported limitations in IR following rTSA compared to aTSA^11,18,26 resulting in decreased patient satisfaction.¹⁵ However, in the current study, there was no significant difference between postoperative IR between the 2 groups. One possibility for this finding could be attributed to patient demographics. Prior studies that have reported loss in IR following rTSA when compared to aTSA have related this to variables including gender, dominant-sided surgery, older age, and a higher BMI.^4,11 Eichinger et al reported that a low BMI can be attributed to optimized IR following rTSA, reasoning that IR is impacted by BMI due to difficulty clearing torso soft tissue, arm weight, and deconditioning.⁴ These findings could explain the inverse correlation seen with IR improvement and BMI with the current study, as the average BMI here was 27, while the United States national average as of 2020 is 30.¹⁶ Another possible explanation is the nature of yoga practice, which involves various poses that require IR. It is possible that the return to yoga as a sport influenced increased stretching, encouraging restoration of IR following rTSA. Further study is required, but it is possible that yoga may even be considered as a form of rehab for regaining IR following rTSA.

Limitations

There are several limitations to the current study. Due to the small sample size, a multivariate analysis considering the effects of age, sex, laterality, hand dominance, or preoperative diagnosis was not performed. Second, due to the retrospective nature of the study, there is a risk of selection bias. Third, there is a risk for recall bias, which is inherent to survey use. Fourth, the generalizability of these findings may be limited due to the surgeries being performed by a single high-volume surgeon at a single institution. Lastly, there is an increased risk for type 2 error based on the small sample size used. Additional comparison studies with longer follow-up periods are needed to reach more definitive conclusions.

Conclusion

Patients who engage in yoga have a high return to sport rate following either aTSA or rTSA, although full return may take up to 1 year postoperatively. The vast majority of patients also report improvement in their ability to perform yoga following shoulder arthroplasty. Among yoga participants, postoperative ROM, including IR, appears to be similar between aTSA and rTSA.

Acknowledgment

The authors acknowledge Colin T. Donnelly for their assistance in data collection.

Disclaimers:

Funding: No funding was disclosed by the authors.

Conflicts of interest: Patrick J. Denard received royalties and consultant payments from Arthrex Inc., as well as speaking fees from Pacira, which are both related to the subject of this work. All the other authors, their immediate families, and any research foundations with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.