Prior mood disorder diagnoses do not relate to current mood disorder symptoms or patient-reported disease severity in rotator cuff patients

Abstract

Background

Surgery for rotator cuff syndrome does not always produce symptom improvement. Biological factors may explain some symptoms, but mood disorder symptoms may also contribute. The purpose of this study is to examine the interaction between disease severity, prevalence of mood disorder diagnoses, and current mood disorder symptoms in preoperative rotator cuff patients.

Methods

A prospective cohort of patients aged 35–75 years with unilateral rotator cuff disease awaiting surgery participated. Demographics, psychiatric history, the Hospital Anxiety & Depression Scale, and the Western Ontario Rotator Cuff index were collected. Descriptive and univariate statistical testing was performed.

Results

Of 140 participants (75M:65W) aged 55 ± 8 years, 34 reported a prior diagnosis of a mood disorder. There was a moderate positive relationship between disease severity and current depression and anxiety scores. Women were more likely to carry a diagnosis of a mood disorder, but there were no differences in current symptom levels between genders. No differences were found in patient-reported outcome measure scores between patients with and without a mood disorder diagnosis.

Discussion

Current mood disorder symptoms were associated with greater disease severity, whereas the presence of a past mood disorder diagnosis was not. Awareness of this relationship may reduce bias about past mood disorder diagnoses during decision-making.

Introduction

Anxiety and depression are common mood disorders, affecting 19.1% of Americans each year. ¹ Current understanding is that the presence of a mood disorder leads to increased sensitivity to pain, a greater overall pain experience, and reduced coping ability across several orthopedic populations.^2–6 Studies would suggest this relationship is complex, as it may be related to sex differences in sensitivity to pain, ³ as well as an interaction between psychological symptoms and one's pain experience. ⁵ For example, patients with chronic shoulder pain, depressive symptoms and/or a pain catastrophizing mindset were found to report greater disease severity. ⁷ This relationship may exert an important effect on patient-reported outcomes of musculoskeletal disease and warrants consideration alongside biological factors when interpreting outcomes. Pain and its negative effects on physical and emotional functioning is a prominent feature of rotator cuff syndrome.^8,9 In patients with rotator cuff syndrome specifically, poorer self-assessed pain and function have been reported at baseline among those also reporting mental health symptoms, including anxiety, depression, and/or catastrophizing.^10–12 Of concern, among shoulder patients presenting for surgical intervention, psychological distress before surgery was found to persist ¹³ and affect self-reported outcomes postoperatively.^13,14 While surgical management is a common treatment choice, it is not always successful at improving pain.^15–17 The purpose of this study is to examine the interaction between patient-reported disease severity, the prevalence of past anxiety and/or depression diagnoses, and current anxiety and depression symptom levels in preoperative rotator cuff patients. We hypothesize that patients with greater levels of current anxiety or depression symptoms will self-report greater rotator cuff symptom severity, and that prior diagnoses of anxiety or depression may not reflect current mood disorder symptom levels or self-reported rotator cuff disease severity.

Methods

This is a cross-sectional study of preoperative rotator cuff surgical patients. This study complied with the Strengthening the Reporting of Observational Studies in Epidemiology statement requirements for reporting of observational studies. ¹⁸

Participants

All eligible participants were aged 35–75 years undergoing elective surgery for unilateral partial- and/or full-thickness rotator cuff tendon tears confirmed by ultrasound or MRI and were recruited by one of two fellowship-trained shoulder surgeons at a single-center between February 2016 and June 2017. Exclusion criteria were bilateral symptomatic rotator cuff disease, previous surgery on the operative shoulder, rotator cuff arthropathy, significant alternate sources of pain such as cervical spine disease or a chronic pain disorder (such as complex regional pain syndrome, or other conditions often paired with chronic opioid use, which have been found to influence both treatment course and outcomes), and inability to complete questionnaires in English.^19,20 Participants with known psychiatric diagnoses such as anxiety, depression, or related conditions were not excluded, nor were patients with potential gain issues, such as Workers' Compensation board claims, litigation, or those injured in motor vehicle collisions. Informed consent was obtained from all participants prior to enrollment. The project was approved by the local Research Ethics Board (REB 15-1229).

Instruments

Demographic data collected included hand dominance, medical comorbidities, current medications, self-reported activity level, and smoking history. History of psychiatric diagnoses (major unipolar or bipolar depression, anxiety disorders) and their treatment with medication and/or therapy was documented.

Three patient-reported outcome measures (PROMs) were administered immediately following surgical consent by a research assistant in the clinic: the Western Ontario Rotator Cuff (WORC) Index, the Hospital Anxiety and Depression Score (HADS), and the Short Form 36 (SF-36). The WORC score is a validated, self-reported measure of rotator cuff disease severity. ²¹ It consists of 21 questions in 5 domains, each with a 100 mm visual analogue scale (VAS). Higher total scores on the WORC indicate increased pain and functional disability. The HADS is a validated tool used to assess current depression or anxiety symptom levels using a 14-item questionnaire and a 4-point Likert scale from 0 (“not at all”) to 3 (“most of the time”).^22,23 The HADS contains seven questions for symptoms of anxiety (HADS-A) and seven for depression symptoms (HADS-D), and these scores are not intended to be combined or to diagnose anxiety and/or depression. The SF-36 is a validated measure of general health status, and it has been used in previous rotator cuff and orthopedic outcome research. ²⁴ It contains two major domains: the physical component score and the mental component score.

Rotator cuff disease characteristics were grouped for the purposes of this study as follows: (1) any single partial-thickness tear in a single tendon, (2) partial-thickness tears in two or more tendons, (3) any full-thickness tear in a single tendon, (4) any full-thickness tear in a single tendon and any partial-thickness tear in a second tendon, and (5) full-thickness tears in two or more tendons.

Statistical analysis

Descriptive statistics were performed for demographics and patient-reported data. T-tests and Pearson product correlations were used for univariate testing. Power analysis to detect an 11.7% change in WORC scores is known to require a minimum of 35 participants per group for a primary binary comparison, as outlined by Kirkley et al. ²⁵ Statistical analysis was completed using Minitab Express version 1.5.1 (State College, Pennsylvania), Microsoft Excel 2011 (Redmond, Washington), and GraphPad Prism (La Jolla, California).

Results

Of the 217 patients screened for eligibility, 140 (75 men and 65 women) participated. The most common reasons for non-participation were unwillingness to participate (n = 35) and presence of bilateral rotator cuff disease (n = 32). Demographic data and radiographic disease severity are shown in Tables 1 and 2, respectively.

Table 1.

Participant demographics.

Criterion	No. of participants (140 total)
Gender	75M:65F
Mean age	55 ± 8 (range: 35–74)
Injury mechanism
Acute/Acute-on-chronic	95 (68%)
Atraumatic	45 (32%)
Dominant arm operative	87 (62%)
Activity level
Active	56 (40%)
Sedentary	69 (49%)
Retired	16 (11%)
On disability pension	1 (0.07%)
Unemployed	3 (2%)
Smoking status
Current	17 (12%)
Former	24 (17%)
Never	98 (70%)
Diabetes mellitus	8 (6%)
Workers compensation (WCB)	5 (4%)
Motor vehicle collisions	13 (9%)
Litigation	6 (4%)

Table 2.

Rotator cuff syndrome disease characteristics and concomitant pathologies as determined by imaging.

Feature	Number of patients (140)
Any single partial-thickness tear	16 (11%)
Partial-thickness tears in any two or more tendons	5 (4%)
Any full-thickness tear in a single tendon	64 (46%)
Any full-thickness tear in a single tendon and any partial-thickness tear in a second tendon	19 (14%)
Full-thickness tears in two or more tendons	36 (26%)
Biceps pathology	40 (29%)
Acromioclavicular joint arthritis	11 (8%)

There was a moderate positive relationship between reported disease severity and current depression (R = 0.53, p < 0.0001; Figure 1) and anxiety scores (R = 0.40, p < 0.0001, Figure 1). Statistical and clinical differences in disease severity were found between participants with current moderate-to-severe (HADS > 10) anxiety (1641 (95% CI 1470–1810) versus 1235 (95% CI 1170–1300), p < 0.001; (Figure 2)) or depression symptoms (1711 (95% CI 1580–1840) versus 1235 (95% CI 1170–1300), p < 0.001; (Figure 2)).

Figure 1.

WORC Index versus the HADS—Depression (a) or Anxiety Subscale (b). Higher scores on the HADS indicate a greater depression or anxiety symptoms, and higher scores on the WORC indicate a greater degree of disease severity. HADS: Hospital Anxiety and Depression Score; WORC: Western Ontario Rotator Cuff.

Figure 2.

Participants with moderate to severe mood disorder symptoms (HADS-A (n = 15) or HADS-D (n = 14) > 10) and those without symptoms (HADS-A (n = 95) or HADS-D (n = 99) < 7) on the WORC Index. ** indicates p < 0.001. Hospital Anxiety and Depression Score; WORC: Western Ontario Rotator Cuff.

Thirty-four participants (13M:21W) reported a prior or current diagnosis of anxiety and/or depression. All but one of these 34 participants had received treatment with medication or therapy in the past. Seventy-seven percent and 74% of participants with past diagnoses of anxiety or depression displayed normal levels of anxiety or depression symptoms (HADS-A or HADS-D ≤ 7). No differences on the HADS-A, HADS-D, SF-36, or WORC were found between those with a prior mood disorder diagnosis and those without (Table 3).

Table 3.

Comparison of patient-reported outcome measures (PROMs) for those with and those without a previous mood disorder diagnosis.

	Any previous mood disorder diagnosis (n = 34)	No previous mood disorder diagnosis (n = 109)	p-value
HADS-A	5.2 ± 3.3	4.9 ± 3.7	0.60
HADS-D	6.7 ± 3.7	5.3 ± 3.7	0.15
SF-36 PCS	41.8 ± 6.6	41.3 ± 6.6	0.70
SF-36 MCS	47.4 ± 11.6	51.6 ± 11.1	0.059
WORC	1303 ± 343	1330 ± 355	0.60

HADS: Hospital Anxiety and Depression Score; MCS: mental component score; PCS: physical component score; SF-36: Short Form 36; WORC: Western Ontario Rotator Cuff.

Participants represented the full spectrum of structural rotator cuff disease as well as many different life circumstances (Table 1). With respect to tear size/severity, there was no difference between mean WORC scores in the different tear size groups (p = 0.324). Overall, no relationship between gender and any PROMs scores was noted (Table 4).

Table 4.

Comparison of rotator cuff, anxiety, depression, and general health outcome measures for men and women.

	Men (SD)	Women (SD)	p-value
WORC	1291 ± 335	1342 ± 360	0.39
HADS-D	5.3 ± 3.7	5.4 ± 3.9	0.29
HADS-A	4.7 ± 3.5	5.8 ± 3.8	0.53
SF-36 PCS	41.3 ± 6.8	41.5 ± 6.3	0.86
SF-36 MCS	49.3 ± 11.6	52.2 ± 10.8	0.13

HADS: Hospital Anxiety and Depression Score; MCS: mental component score; PCS: physical component score; SF-36: Short Form 36; WORC: Western Ontario Rotator Cuff.

Discussion

This cross-sectional study examined 140 participants presenting with symptomatic rotator cuff disease. The primary finding was that current mood disorder symptoms correlated with patient-reported disease severity. Secondary findings were that WORC scores did not correspond to psychiatric history nor tear severity, and that no differences in mood disorder scores or WORC scores were found between men and women. A prior diagnosis of mood disorder did not predict current mood disorder symptoms or patient-reported rotator cuff disease severity.

Mood disorder symptoms and disease severity

The link between mood disorders and rotator cuff syndrome has been previously explored. Cho et al. ²⁶ examined 107 preoperative rotator cuff patients using the Korean Shoulder Scale, the American Shoulder and Elbow Score (ASES), the World Health Organization Quality of Life Scale (WHOQOL-BREF), and a VAS. The results of our study did align with some of their findings, revealing that anxiety and depression symptoms had a negative impact on self-reported outcome measures in preoperative rotator cuff patients. ²⁶ That study also showed that depression symptoms were a stronger predictor of patient-reported disease severity. ²⁶ Similarities between the cohorts for prevalence of both anxiety and depression symptoms exist, however. Cho et al. ²⁶ found that 23% of participants displayed anxiety symptoms and 26% displayed depression symptoms (versus 24 and 23% for anxiety and depression in this cohort, respectively). It is important to note that although this study did not exclude any patients with prior mood disorder diagnosis, Cho et al. ²⁶ did exclude eligible participants with a history of a psychiatric diagnosis. Nor is this study isolated in its approach to psychiatric comorbidities, which has affected the applicability of outcome-based findings in other shoulder populations. This study even included those with active workers compensation board claims, as studies have shown they experience a greater incidence of depression.^27–30 Other evidence presented by Potter et al. aligns with the findings from our study. Their study demonstrated an association between disease severity and symptoms of mood disorder among rotator cuff surgical patients, but rather used the ASES, the Simple Shoulder Test, and a VAS. ¹⁰ Additionally, Thorpe et al. ¹⁴ also investigated rotator cuff surgical patients and found that poorer psychological measures accounted for differences in ASES compared to those who did not experience psychological distress. Interestingly, such differences still existed following surgical intervention at 3 and 12 months postoperatively. Continued follow-up from our study is needed therefore to corroborate such findings.

Mood disorder history and disease severity

Our study demonstrates that a previous history of mood disorder diagnoses did not predict current mood disorder symptom levels (as measured on the HADS) or their WORC scores. All but one of the patients in our cohort had received prior treatment via medication or therapy for their condition following a mental health diagnosis. This suggests that having a prior diagnosis of a mood disorder may not reflect a patient's current situation accurately. This challenges ideas that a prior mental health diagnosis may have a negative impact on rotator cuff patients. To our knowledge, this topic has yet to be investigated among a comparable population of shoulder patients.

Similar to the findings of Cho et al., ²⁶ the presence of significant levels of undiagnosed mood disorder symptoms present in our cohort was salient. Symptoms of anxiety and depression are a potentially important part of the illness experience in this patient population and merits further attention. Use of a PROM such as the HADS questionnaire in rotator cuff surgical patients as a means to screen for anxiety or depression symptoms may be useful to care providers for preoperative planning in efforts to optimize care plans by allowing surgeons to more readily identify patients who may be experiencing increased levels of psychological distress. Surgeons may consider incorporating evaluation and treatment of mood disorders as part of a multidisciplinary preoperative optimization.

The presence of a formal mental health diagnosis moderating disease severity warrants further investigation at this time. Our study does, however, suggest that continued efforts to better understand the relationship between symptom severity and symptoms of mental distress among rotator cuff patients prior to their surgical intervention may be useful to inform and streamline care pathways. As previous studies have demonstrated a link between preoperative disease severity and symptoms of mental distress in orthopedic populations (including lower extremity joint replacement surgery and adult spine surgery), this phenomenon may also relate to rotator cuff patients.^6,31–33 In these studies, surgical intervention was associated with enhanced function and lower measures of anxiety or depressive symptoms, postoperatively. While treatment of mood disorder among orthopedic patients has not previously represented a research priority, two studies have demonstrated improved Health Related Quality of Life (HRQoL) and lower self-reported pain in hip or knee osteoarthritis and those with chronic low back pain.^19,34 Since treatment of mood disorders with medication and/or therapy is effective for those with chronic mood disorder symptoms,^35,36 research into treatment of existing mood disorders symptoms among rotator cuff surgical patients as part of preoperative optimization may be useful. Considering that this study found mood disorder symptoms were related to the patient-reported disease severity in pre-surgical patients, it is possible that such treatment may even alleviate the need for surgery completely in some patients or enhance final postoperative outcomes.

Gender and mood disorders

It is also notable that no differences existed in this cohort between men and women when evaluating current mood disorder symptoms. The fact that more women carried a previous mood disorder diagnosis, despite similar symptom levels to men, suggests a possible difference in diagnosis making in our health care system. Potential explanations may include men being less engaged with primary care or differences in how men are screened by their primary care providers. Our study therefore challenges prior research findings about mood disorder symptoms being more prevalent in women presenting with rotator cuff disease.^15,37,38 This is an addition to the literature from our study, as previous studies that have discussed sex or gender outcomes in the rotator cuff literature have found differences between men and women. Specifically, a longitudinal study by Cho et al. found that women reported higher scores on the both HADS-A and HADS-D, as well as reporting poorer HRQoL scores on the WHOQOL-BREF than men following rotator cuff repair. ³⁷ Further work needs to be done, but the question of whether gender or mood disorder symptoms are the dominant factor versus a confounding factor is not settled.

Strengths and limitations

The main strength of this study was that we included groups typically excluded in other research, such as those with mood disorder-related psychiatric histories and compensation cases. This serves to improve the generalizability of the results as these patients have been frequently excluded in similar studies. Another strength was that patients with bilateral disease and those over the age of 75 years were excluded from participating. With respect to bilateral disease, this improves the generalizability of the results by preventing interference by pain from the contralateral shoulder in the PROMs. Exclusion of patients over the age of 75 years is also necessary to improve the generalizability, in that these patients are more likely to be considered for non-operative treatment, or may be better suited for a reverse total shoulder arthroplasty.

Limitations of this study were the use of heterogeneous imaging modalities (as they were predominantly completed by ultrasound) that limited the ability to precisely measure tear size. While dominant use of ultrasound imaging is highly sensitive for identification of rotator cuff tears (at least relative to MRI), this lack of specificity and precision inherent to ultrasound imaging could account for how tear size was not related to WORC scores. This does, however, represent the real world situation of our health care system and the preoperative information likely available to surgeons, as homogenous MRI imaging is neither feasible nor timely. We were also limited by the exclusion of (1) non-English speakers, who may have illness experiences different than English speaking patients, and (2) the low sample size of smokers and patients with diabetes mellitus, as these factors may physiologically alter a patient's experience with rotator cuff syndrome and could not be comprehensively evaluated.

Conclusion

Current symptoms of anxiety and/or depression were associated with greater patient-reported disease severity, whereas the presence of a past diagnosis of mood disorder was not. There were significant levels of undiagnosed anxiety and depression in this preoperative patient cohort. No gender differences existed for measures of anxiety, depression, or rotator cuff disease severity. This study serves to emphasize that surgeons should not make assumptions about current mental health status based on past diagnoses or lack thereof. Further research is needed to determine if current mood symptom severity, rather than psychiatric history, exerts an effect on postoperative outcomes.