Psychosocial factors associated with postoperative outcomes after arthroscopic rotator cuff repair

Abstract

Background

Arthroscopic rotator cuff repair (ARCR) is a common treatment for rotator cuff tears with postoperative outcomes being influenced by multiple factors. Patient characteristics and surgical techniques are well-documented associated with postoperative outcomes; however, the role of psychosocial factors, including pain catastrophizing, kinesiophobia, and pain self-efficacy, in influencing postoperative functional outcomes remains unclear. The present study explores whether psychosocial factors are associated with postoperative outcomes following ARCR.

Methods

This retrospective study included patients who underwent ARCR for full-thickness supraspinatus tears between January 2022 and May 2024. Comprehensive data were collected before and after surgery, including patient demographics, characteristics of the tendon lesion, surgical techniques, and psychosocial factors evaluated using the Tampa Scale for Kinesiophobia (TSK-11), Pain Catastrophizing Scale (PCS), and Pain Self-Efficacy Questionnaire. The primary outcome measure was the American Shoulder and Elbow Surgeons (ASES) score. Multiple linear regression analysis was conducted to evaluate the association between patient demographics, characteristics of the tendon lesion, surgical techniques, psychosocial factors, preoperative ASES scores, and the postoperative ASES scores.

Results

Significant improvements were observed postoperatively in ASES, TSK-11, PCS, and Pain Self-Efficacy Questionnaire scores (P < .01, all). Multiple linear regression analysis identified male sex, absence of diabetes, repair technique, higher preoperative ASES scores, and lower TSK-11 and PCS scores as variables independently associated with higher postoperative ASES scores. The model accounted for approximately 10% of the variance in postoperative ASES scores (adjusted R² = 0.10, P < .01).

Conclusion

This study suggests a potential association between psychosocial factors—such as kinesiophobia and pain catastrophizing—and functional recovery after ARCR. Although causality cannot be inferred, these associations highlight the potential clinical relevance of psychosocial factors in postoperative recovery. Poorer psychological scores, including high TSK-11 and PCS scores, were associated with lower postoperative outcomes. Further investigation is needed to explore how interventions targeting these psychosocial factors can enhance recovery and patient satisfaction.

Rotator cuff tears are a prevalent shoulder joint disorder in middle-aged and older adults.^22,36 The symptoms of rotator cuff tears include pain, sleep disturbance, and functional decline.²³ Arthroscopic rotator cuff repair (ARCR) is a commonly conducted surgical treatment for rotator cuff tears, with evidence supporting favorable clinical outcomes.^5,6 However, postoperative complications such as retears, persistent pain, shoulder stiffness, and insufficient functional recovery have been reported in some cases.^11,16,30 Consequently, a comprehensive investigation into the factors potentially associated with postoperative outcomes is essential to enhance patient satisfaction.

Prognostic factors influencing postoperative clinical outcomes after ARCR using multivariate analyses have been investigated and include patient characteristics such as age, sex, the body mass index, diabetes, rheumatologic disease, smoking status, and workers' compensation status.^{3,8,9,13,20,24,27} Tear tendon characteristics such as tear size, number of tendon tears, fatty infiltration, and retraction have been identified as prognostic factors.^3,8,18,20,24 Surgical procedures, including acromioplasty, treatment of the long head of the biceps (LHB), and repair techniques, have also been associated with postoperative clinical outcomes following ARCR.^3,13,20,24 To enhance patient satisfaction, recognizing these associations and addressing modifiable factors during the preoperative and perioperative phases may be beneficial in clinical practice.

Psychosocial factors, including depression, anxiety, pain catastrophizing, kinesiophobia, and pain self-efficacy are reportedly associated with clinical outcomes in various musculoskeletal disorders.^{10,21,27,30,35} Furthermore, preoperative anxiety and depression are also associated with postoperative clinical outcomes after ARCR.^14,17,19 Thorpe et al³¹ investigated whether psychosocial functioning affects functional outcomes following ARCR or arthroscopic subacromial decompression using latent class analysis. They reported that patients with poor psychological scores (depression, anxiety, pain catastrophizing, kinesiophobia, and pain self-efficacy) had lower American Shoulder and Elbow Surgeons (ASES) scores preoperatively and postoperatively.³¹ In contrast, Wang et al³⁴ reported no difference in ASES scores at the final follow-up of patients who underwent ARCR having preoperative high and low Tampa Scale of Kinesiophobia (TSK) scores, which is a tool used to assess kinesiophobia. Furthermore, Schwank et al²⁸ found that the Pain Catastrophizing Scale (PCS) score did not affect the postoperative Western Ontario Rotator Cuff Index after ARCR. Nonetheless, it remains uncertain whether preoperative psychosocial factors other than depression and anxiety are associated with or may act as prognostic factors for postoperative functional outcomes in patients undergoing ARCR. Understanding the association between clinical outcomes and modifiable factors such as psychosocial factors may be crucial for enhancing clinical decision-making in preoperative and perioperative patient management.²⁸

This study aimed to comprehensively identify the preoperative and perioperative factors associated with postoperative outcomes, including psychosocial factors that influence postoperative clinical outcomes following ARCR using a multivariate analysis model. This study hypothesized a potential association between psychosocial factors and postoperative ASES scores following ARCR.

Materials and methods

Study design

This retrospective study included patients subjected to ARCR at a single institution between January 2022 and May 2024. At the end of the formal, medically prescribed rehabilitation and follow-up for ARCR, the ASES score was evaluated. Rehabilitation was generally continued for at least 4 months postoperatively. Except in cases where patients were unable to attend outpatient sessions due to personal circumstances, all patients completed a minimum of 4 months of rehabilitation. Preoperative and perioperative data, including patient demographics, tendon lesion characteristics, surgical techniques, and psychosocial factors (TSK-11, PCS, and Pain Self-Efficacy Questionnaire (PSEQ) were collected to investigate their potential associations with the postoperative ASES score. The TSK-11, PCS, PSEQ, and ASES scores were evaluated at the final follow-up, and changes from preoperative to postoperative were also determined.

Patients

This study included patients who underwent primary ARCR for full-thickness supraspinatus tears and underwent medical assessment both preoperatively and at the final follow-up. Supraspinatus tears were diagnosed using magnetic resonance imaging (MRI) and were confirmed arthroscopically. A total of 429 eligible patients met the inclusion criteria. The study excluded patients who met any of the following criteria¹: previous surgical interventions on the ipsilateral shoulder joints²; appearance of osteoarthritic changes in the shoulder joint³; presence of neurological ailments or rheumatoid arthritis⁴; partial-thickness supraspinatus tears⁵; patients subject to ARCR combined with Bankart repair; and⁶ incomplete medical records. In the final analysis, the study cohort comprised 368 patients (Table I) (Fig. 1). The study was conducted in accordance with the Declaration of Helsinki and approved by our institutional review board (approval number: 0079). Written consent was obtained from all patients.

Table I.

Patient characteristics.

Age, yr	68.2 ± 9.5 (40-93)
Sex
Female	194 (53.0)
Male	172 (47.0)
BMI (kg/cm2)	24.3 ± 3.6 (16.4-36.2)
Dominant side affected
Yes	247 (67.5)
No	119 (32.5)
Diabetes
Yes	39 (10.7)
No	327 (89.3)
Smoking
Yes	36 (9.8)
No	330 (90.2)
Workers' compensation
Yes	15 (3.6)
No	353 (96.4)
Follow-up time	5.9 ± 1.3 (4.0-12.0)

BMI, body mass index.

Continuous data are presented as mean ± SD (range) and categorical data as n (%).

Figure 1.

Flowchart for participant enrollment.

Surgical procedure and tendon lesion characteristics

A rotator cuff tear was diagnosed by an experienced physician based on the MRI findings and the arthroscopic findings were confirmed. Surgery was performed under general anesthesia and brachial plexus block with the patient in the side-lying position with 4 kg-arm traction. Initially, a manual examination under anesthesia was conducted to evaluate the range of motion (ROM) of the shoulder joint, to examine impingement signs, and to assess instability. If shoulder joint stiffness was observed, joint manipulation was performed at the discretion of the surgeon. Tenotomy of the biceps long LHB was performed in patients with the following conditions¹: partial lesions >50%²; loss of extensibility; and³ degeneration of the intertrochanteric groove. A superior labrum from anterior to posterior tear was also treated with débridement when it was diagnosed as type 1 and repaired when considered types 2, 3, or 4. The surgical procedures routinely included bursectomy, acromioplasty, and excision of the coracoacromial ligament for all patients. Tear size was classified based on the DeOrio and Cofield classification as small (<1 cm), medium (1-3 cm), large (3-5 cm), or massive (>5 cm).⁷ ARCR was conducted using the triple-row technique for small to large tear size. A medial row anchor was inserted medial to the humeral footprint and the rotator cuff tear. The middle anchors were placed at the lateral edge of the footprint, and the suture limbs were passed through the torn rotator cuff. The knot tying of the middle anchors was performed using a sliding knot technique. Finally, the suture limbs from the medial anchors were retrieved and secured using the suture-bridge technique with lateral row anchors. The double-row or single-row technique was performed for large tear sizes that could not be addressed using the triple-row technique, particularly if excessive tension was anticipated to be applied to the repaired tendon when pulling the torn tendon to its original footprint during arthroscopic surgery. Lastly, a manual examination under anesthesia was undertaken to verify no restriction in the ROM of the shoulder joint. Tendon lesion characteristics, tear size, and number of tears were determined.

Postoperative rehabilitation

Patients with small-to medium-sized rotator cuff tears were instructed to wear an abduction brace (Global Sling; Cosmos, Sapporo, Japan) for 1 week, whereas those with large-sized tears were instructed to wear the same brace for 3 weeks. All patients underwent supervised postoperative rehabilitation guided by physical therapists following a standardized protocol in our institution. On the first postoperative day, all patients were permitted to perform passive shoulder flexion ROM exercises to the full range as tolerated, as well as active ROM exercises for the elbows and hands. ROM exercises were performed within the patient's tolerance and adjusted based on pain levels. One week after surgery, patients began passive external rotation ROM exercises for the shoulder and muscle strength training for the scapular muscles. At 3 weeks postoperatively, patients with small- to medium-sized tears initiated passive shoulder ROM exercises in all directions—including internal rotation, adduction, and extension—along with active shoulder motion training. Patients with large tears began these exercises at 4 weeks postoperatively. The patients were advised to gradually resume light work activities 2 months after surgery. At approximately 3 months after surgery, resistance exercises targeting shoulder abduction were generally initiated, depending on patient recovery and symptom status. Psychosocial factors were not specifically addressed as part of the rehabilitation process.

Psychosocial factors

Among psychosocial factors, kinesiophobia, pain catastrophizing, and self-efficacy, are potential targets for intervention during the preoperative and perioperative periods. The PCS comprises 13 items aimed at evaluating thoughts and emotions in the context of pain, yielding scores between 0 and 52, wherein higher scores indicate heightened pain catastrophizing tendencies.¹⁰ The Tampa Scale of Kinesiophobia-11 (TSK-11) is a shortened version of the original TSK. This 11-item questionnaire devised to measure fear of movement or reinjury, yielding scores from 11 to 44, with elevated scores indicative of greater fear associated with pain.¹⁵ The PSEQ is a 10-item assessment tool designed to gauge confidence in managing pain across various domains, with scores ranging from 0 to 60, where lower scores indicate lower belief in self-efficacy.³³ The PCS, TSK-11, and PSEQ are valid and reliable questionnaires used to assess cognitive-psychosocial factors.^26,32,33

Statistical analysis

Power analysis was conducted using G∗Power v.3.1 software (Heinrich Heine University, Düsseldorf, Germany). The sample of 238 patients yielded 95% statistical power with an α of 0.05 for a medium effect size (f² of 0.15) for linear multiple regression with 24 independent variables. Statistical analyses were performed using JMP Pro v.17 software (SAS Institute Inc., Cary, NC, USA). The normality of continuous variables was assessed using the Shapiro–Wilk test. Preoperative and postoperative ASES scores and psychosocial factors were compared using the paired t-test or Wilcoxon rank test, depending on the normality distribution of the data. Effect sizes (Cohen's d) were calculated to assess the magnitude of change between baseline and final follow-up for the ASES score and psychosocial measures (TSK-11, PCS, and PSEQ). Multiple linear regression analysis was adopted to examine the potential associations between independent variables and the dependent variable (postoperative ASES score). Preliminary analyses were conducted using univariate linear regression to assess the association between postoperative ASES scores and the following variables: demographic data (sex, age, body mass index, smoking status, dominant side affected, diabetes, workers' compensation), preoperative psychosocial factors (TSK-11, PSEQ, and PCS), preoperative ASES scores, and perioperative factors (LHB procedure, superior labral anterior-posterior procedure, repair technique, and glenohumeral joint manipulation). The preoperative and perioperative variables included in the initial analysis were selected based on previously published predictors of functional outcomes after ARCR.^12,20 Because only 2 patients underwent single-row repair, they were excluded from the multiple linear regression analysis to avoid statistical instability. Finally, multivariate linear regression analysis was conducted with the postoperative ASES score as the dependent variable. All independent variables that were significantly associated (P < .05) with the final ASES score in the univariate analyses were evaluated for inclusion in the multivariable model. Only variables with P < .10 were retained in the final model. In the multiple linear regression analysis, β coefficients were calculated to assess the independent variables associated with postoperative ASES scores. The variance inflation factor was calculated to evaluate multicollinearity. Statistical significance level was set at P < .05.

Results

Patient and tear tendon characteristics

Patient and tear tendon characteristics and perioperative factors are presented in Table I, Table II, and III, respectively. The average follow-up period 5.9 ± 1.3 months. Table IV shows the preoperative and postoperative ASES, TSK-11, PCS, and PSEQ scores. Postoperatively, significant improvements were observed for all measures. The mean ASES score increased from 54.6 ± 19.4 (range 3-100) preoperatively to 85.5 ± 12.0 (range 35-100) postoperatively (P < .01). Similarly, the TSK-11 score decreased from 25.0 ± 4.5 (range 11-37) to 20.8 ± 5.5 (range 4-36) (P < .01), the PCS score decreased from 20.6 ± 10.1 (range 0-52) to 12.7 ± 9.4 (range 0-52) (P < .01), and the PSEQ score increased from 37.9 ± 11.8 (range 0-60) to 46.6 ± 11.6 (range 0-60) (P < .01).

Table II.

Tear tendon characteristics.

Cofield classification
Small	45 (12.3)
Moderate	243 (66.4)
Large	78 (21.3)
Tear tendon
SSP	319 (87.2)
SSP + ISP	34 (9.3)
SSP + SSC	2 (0.5)
SSP + ISP + SSC	11 (3.0)
Long head of the biceps
Intact	242 (66.1)
Partial tear	84 (22.3)
Complete tear	40 (11.0)
Superior labral anterior-posterior
Intact	145 (39.6)
Type Ⅰ	151 (41.3)
Type Ⅱ	68 (18.6)
Type Ⅲ	2 (0.5)

SSP, suprascapularis; ISP, infraspinatus; SSC, subscapularis.

Continuous data are presented as mean and categorical data as n (%).

Table III.

Perioperative variables.

Long head of the biceps procedure
None or débridement	306 (83.6)
Débridement + tenotomy	20 (5.5)
Not applicable (complete tear)	40 (10.9)
Superior labral anterior-posterior procedure
None or débridement	310 (84.7)
Débridement + repair	56 (15.3)
Repair technique
Triple row	355 (96.5)
Double row	22 (6.0)
Glenohumeral joint manipulation
Yes	78 (21.3)
No	288 (78.7)

Continuous data are presented as mean and categorical data as n (%).

Table IV.

Preoperative and postoperative ASES sore and TSK-11, PCS, and PSEQ.

	Pre	Post	Effect size (Cohen's d)	P value
ASES	54.6 ± 19.4 (3-100)	85.5 ± 12.0 (35-100)#	1.92	<.01
TSK-11	25.0 ± 4.5 (11-37)	20.8 ± 5.5 (4-36)#	0.84	<.01
PCS	20.6 ± 10.1 (0-52)	12.7 ± 9.4 (0-52)#	0.81	<.01
PSEQ	37.9 ± 11.8 (0-60)	46.6 ± 11.6 (0-60)#	0.74	<.01

ASES, American Shoulder and Elbow Surgeons Standardized Shoulder Assessment Form; PCS, Pain Catastrophizing Scale; PSEQ, Pain Self-Efficacy Questionnaire; TSK-11, Tampa Scale of Kinesiophobia-11; SD, standard deviation.

Continuous data are presented as mean ± SD (range) and categorical data as n (%).

^#P < .05, comparison between preoperative and postoperative scores.

Multiple linear regression analysis

In the multiple linear regression analysis, the independent variables associated with higher postoperative ASES scores were male sex (β = 0.12, P = .02), absence of diabetes (β = 0.12, P = .01), higher preoperative ASES scores (β = 0.12, P = .03), lower TSK-11 scores (β = 0.13, P = .02), and lower PCS scores (β = 0.11, P = .048). The multivariable regression model demonstrated an R² value of 0.12 and an adjusted R² of 0.10, indicating that the model accounted for approximately 10% of the variance in postoperative ASES scores (P < .01) (Table V).

Table V.

Multivariable linear regression analysis with postoperative ASES score as the dependent variable.

Prognostic factors	Standardized coefficient β	VIF	P value
Age	−0.02	1.15	.75
Sex: male	0.12	1.09	.02
Absence of diabetes	0.12	1.02	.01
Number of tear tendons: 1	0.04	1.19	.47
Number of tear tendons: 2	0.03	1.05	.60
Repair technique: Double row	−0.08	1.14	.13
Preoperative ASES score	0.12	1.11	.03
TSK-11	−0.13	1.25	.02
PCS	−0.11	1.33	.048

ASES, American Shoulder and Elbow Surgeons Standardized Shoulder Assessment Form; TSK-11, Tampa Scale of Kinesiophobia-11; PCS, Pain Catastrophizing Scale; VIF, variance inflation factor.

Bold text indicates P < .05.

Discussion

The present study demonstrated that preoperative and perioperative factors, including psychosocial variables, were associated with postoperative clinical outcomes following ARCR. The findings, based on multivariate regression analysis, showed that male sex, absence of diabetes, higher preoperative ASES scores, lower TSK-11 scores, and lower PCS scores were independently associated with better ASES scores after ARCR. These results suggest that poor psychological scores, particularly those reflecting high levels of kinesiophobia and pain catastrophizing, were associated with unfavorable postoperative clinical outcomes after ARCR.

The primary finding of this study was that the psychosocial factors TSK-11 and PCS were independent variables for the postoperative ASES score in a multiple linear regression analysis. Thorpe et al³¹ observed that the lower the physiological score, which encompasses depression, anxiety, pain catastrophizing, kinesiophobia, and pain self-efficacy, the lower the ASES score. However, how individual psychological factors reduce ASES scores remains unclear. Previous studies have not evaluated whether TSK or PCS scores influence postoperative clinical outcomes following ARCR.^29,34 Using a large sample size and multiple regression analysis incorporating patient characteristics, torn tendon characteristics, and intraoperative information, this study found that TSK-11 and PCS scores were independently associated with postoperative ASES scores following ARCR. Thus, the management of kinesiophobia and pain catastrophizing in the preoperative and perioperative periods may be important considerations in the clinical management of patients undergoing ARCR.

With regard to demographic data, male sex and absence of diabetes were identified as factors independently associated with higher ASES scores in our analysis. Previous large-sample-size studies have not reached a consensus on whether sex differences affect postoperative clinical outcomes.^9,13,20 Frangiamore et al⁹ suggested that such differences may be related to variations in postoperative pain management strategies, as well as intrinsic tendon- or bone-related characteristics. Although this study did not investigate these factors, future research should consider aspects such as bone density, nutritional status, hormonal influences, and individualized pain management approaches. As for diabetes, previous large-sample-size studies have generally reported no significant association with postoperative outcomes.^13,20 Contrary to those findings, the present study showed that the absence of diabetes was associated with better postoperative outcomes. However, the proportion of patients with diabetes was small in both our study and the previous studies,^13,20 and thus, larger studies are warranted to draw more definitive conclusions.

The multiple regression analysis conducted in this study revealed that the adjusted coefficient of determination was 0.10, a value similar to that reported by Malavolta et al,²⁰ who recently examined predictors in a large sample size study. Nonetheless, the results differ from those reported by Jenssen et al,¹³ likely because of variations in clinical outcomes. As highlighted by Malavolta et al,²⁰ rotator cuff tears are complex musculoskeletal disorders influenced by multiple factors. Interestingly, the present study found that psychological factors were associated with postoperative clinical outcomes, highlighting the importance of further investigating other potentially relevant factors.

This study suggests that lower levels of kinesiophobia and pain catastrophizing during the preoperative period may be associated with better postoperative outcomes. In other musculoskeletal disorders, kinesiophobia causes compensatory patterns in joint movement and changes in proprioception and muscular activity.^1,2,25,26 Patients with shoulder impingement syndrome who exhibit high level of pain catastrophizing reportedly struggle with completing progressive rehabilitation programs.⁴ These impairments related to kinesiophobia and pain catastrophizing may hinder functional recovery during standard rehabilitation protocols. Therefore, it is crucial to investigate the biomechanical changes and compensatory patterns of the shoulder joint associated with kinesiophobia. In addition, the impact of pain catastrophizing on the rehabilitation process should be carefully considered in future studies.

This study had some limitations that should be acknowledged. First, MRI findings regarding the characteristics of the teared or repaired tendon were not obtained. At our institution, T1-weighted sagittal view imaging to preoperatively assess fatty infiltration of the rotator cuff muscle was not conducted routinely. Similarly, postoperative MRI images were not consistently obtained; thus, the assessment of tendon integrity or retears could not be addressed. Second, most patients included in the study presented isolated supraspinatus muscle tears; therefore, the results should be interpreted with caution in cases of multiple tendon tears. Additionally, because ARCR was mostly performed using the triple-row technique at our institution, different results may have been obtained depending on the tendon repair technique employed. Third, the final follow-up period for this study was limited to approximately 6 months, and long-term outcomes remain unavailable. Although previous studies with large-scale cohorts examining factors associated with postoperative outcomes employed longer follow-up durations,^9,13,20 the emphasis of this study aligns with patient concerns regarding the early postoperative period. Patients undergoing ARCR are particularly interested in understanding the timeline for pain relief and the resumption of normal functional activities following surgery. Therefore, the identification of factors associated with early postoperative clinical outcomes may have a direct impact on patient satisfaction. Furthermore, the ASES score at the time of the final follow-up in this study demonstrated a notable improvement, from 54.6 to 85.3. This finding suggests that the overall shoulder function of patients generally recovered.

Conclusion

The present study demonstrated that preoperative and perioperative factors, including psychosocial variables, were associated with postoperative clinical outcomes following ARCR, as determined using multiple linear regression analysis. The main findings were that male sex, absence of diabetes, higher preoperative ASES scores, lower TSK-11 scores, and lower PCS scores were independently associated with better postoperative ASES scores following ARCR. Overall, the study found that better psychological scores, particularly those related to kinesiophobia and pain catastrophizing, were associated with favorable outcomes after ARCR. Although this study does not establish causality, the observed associations indicate that psychosocial factors could serve as meaningful targets for future interventions to enhance postoperative recovery and patient satisfaction.

Acknowledgment

The authors would like to thank all the participants in this study.

Disclaimers:

Funding: No funding was disclosed by the authors.

Conflicts of interest: The 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.

Data availability

The datasets are available upon reasonable request from the corresponding author.