ABSTRACT:
OBJECTIVE The present study aimed to assess the association between anxiety and depression symptoms
in patients with adhesive capsulitis.
METHODS This was a cross-sectional study carried out in a single center from a tertiary hospital
with patients presenting with secondary adhesive capsulitis. The control group did
not have shoulder disease, thyroid disease, anxiety, and/or depression. The instrument
used was the Hospital Anxiety and Depression Scale (HADS). An analysis of covariance
compared HADS scores between groups. The significance level was 5%.
RESULTS The final sample consisted of 17 patients (case group) and 27 (control group). The
Shapiro-Wilk test revealed normal distribution (p> 0.05). A HADS score > 0.70 (Cronbach alpha) was reliable and presented good internal
consistency. Patients with adhesive capsulitis reported “doubtful” (average/standard
deviation = 8.88/4.50) “anxious symptoms” (p= 0.019) but no “depressive symptoms” (average/standard deviation = 6.41/3.69), despite
p= 0.015.
CONCLUSION There is a “doubtful” positive association between anxiety symptoms and adhesive capsulitis
but a negative association with depressive symptoms.
Keywords: anxiety; adhesive capsulitis; depression; shoulder; signs and symptoms.
RESUMO:
OBJETIVO Avaliar a associação entre sintomas ansiosos e depressivos em pacientes com capsulite
adesiva.
MÉTODOS Trata-se de um estudo transversal realizado em centro único de um hospital terciário
com pacientes portadores de capsulite adesiva secundária. O grupo controle não apresentava
doença do ombro, tireoidopatias e nem ansiedade e/ou depressão. O instrumento utilizado
foi a escala hospitalar de ansiedade e depressão. A análise da covariância foi utilizada
para comparação dos escores do Hospital Anxiety and Depression Scale (HADS, na sigla
em inglês) entre os grupos. O nível de significância foi de 5%.
RESULTADOS A amostra final foi de 17 pacientes (caso) e 27 (controle). Os dados amostrais apresentaram
distribuição normal por meio do teste de Shapiro-Wilk (p> 0,05). A instrumento HADS com pontuação > 0,70 (alfa de Cronbach) se mostrou confiável
e com boa consistência interna. Os pacientes com capsulite adesiva, no quesito “sintomas
ansiosos” (p= 0,019), relataram sintomas no patamar de “duvidosos” (média/desvio padrão = 8,88/4,50).
Eles não apresentaram “sintomas depressivos” (média/desvio padrão = 6,41/3,69), apesar
do p= 0,015.
CONCLUSÃO Existe uma associação positiva “duvidosa” entre sintomas ansiosos e capsulite adesiva,
mas negativa para sintomas depressivos.
Palavras-chave: ansiedade; capsulite adesiva; depressão; ombro; sinais e sintomas.
FIGURES
| Citation: Fernandes MR, Borges RS, Ribeiro CJF, Sequeira MT. Association of Anxiety and Depressive Symptoms in Patients with Adhesive Capsulitis. 58(01):127. doi:10.1055/s-0042-1742693 |
| Financial Support: There was no financial support from public, commercial, or not-for-profit sources. |
|
Conflict of Interests: The authors have no conflict of interests to declare. |
|
Study developed at the Universidade Federal de Goiás, Goiânia, GO, Brazil. |
| Received: June 24 2021; Accepted: November 22 2021 |
INTRODUCTION
Adhesive capsulitis (AC) is a condition that causes pain and limited active or passive movement of the shoulder joint. Pain can be severe and hinder daily living activities. Its diagnosis is eminently clinical, with radiographs showing locoregional osteopenia. Magnetic resonance imaging (MRI) demonstrates a reduced axillary recess and thickening of the coracohumeral ligament, an important restrictor of external rotation.1,2
Although primary adhesive capsulitis develops spontaneously, its etiology remains uncertain. Some associated factors were reported, including diabetes mellitus, thyroid diseases, breast cancer, hemiplegia, and trauma; shoulder surgery followed by immobilization can result in secondary adhesive capsulitis.3
In addition, these patients supposedly have a characteristic constitutional profile with psychological abnormalities such as anxiety and depression, although the causal relationship between AC and these disorders is not well defined.4,5 The Hospital Anxiety and Depression Scale (HADS), a questionnaire used to detect the general status and severity of anxious and depressive symptoms, has been used to research psychological factors associated with several diseases.6-8
Several orthopedic diseases relate to anxiety and depression,9 but studies on the association between these psychological factors and AC are scarce and show divergent findings.4,10,11 Thus, the present study aimed to evaluate the association of anxiety and depression symptoms in patients with adhesive capsulitis. It is thought that patients with AC are positively associated with anxiety but not with depressive symptoms.
MATERIALS AND METHODS
Study Design
This is a cross-sectional study carried out in a single center from a private tertiary hospital located in a capital city in the central region of Brazil between February 2019 and November 2020.
Participants
The study sample was nonprobabilistic, of the consecutive type, consisting of patients with secondary AC. Adhesive capsulitis was defined by constant shoulder pain for at least 4 weeks with anterior elevation of up to 130° and external rotation of up to 50°;12 in addition, it was secondary to systemic diseases and extrinsic or intrinsic to the glenohumeral joint. Imaging showed osteopenia from local disuse on radiographs and volume reduction, as well as thickening of capsuloligamentous structures from the axillary recess on nuclear magnetic resonance imaging.
The inclusion criteria were subjects > 18 years old with AC secondary to the following factors, according to Zuckerman et al.:13 previous surgery, trauma to the affected shoulder, prolonged immobilization, cervical radiculopathy, rotator cuff tear, calcareous tendinitis, diabetes mellitus, neuropathies, and hypothyroidism or hyperthyroidism during treatment.
Patients with glenohumeral arthrosis, blocked shoulder dislocation, humeral head necrosis, vicious consolidation of the proximal humerus, and primary AC were excluded from the study.
Data Collection Procedure
Data collection was carried out prospectively by three researchers in a reserved environment and at a single moment; for the case group, data collection occurred before the beginning of treatment with serial suprascapular nerve blocks. The control group consisted of healthy people, not from the hospital environment, aged between 40 and 70 years old, who did not have AC or thyroid diseases and were not under treatment for anxiety and/or depression. The study subjects were invited to participate and signed an Informed Consent Form.
The self-administered instrument, HADS, contains 14 multiple-choice questions divided into two subscales, one for anxiety and one for depression, with seven items each. It is a short and easy to fill instrument, and answers refer to the last week. The global score ranges from 0 to 21 for both anxiety and depression, with ≤ 7 for negative cases, 8 to 10 for doubtful cases, and ≥ 11 for positive cases.6,8
Outcomes/Independent Variables
The primary outcomes included symptoms of anxiety and depression surveyed using HADS. The independent variables were the following: age (in years); gender (male/female); race (white/non-white); education (in formal years); monthly income (in minimum wages); pain duration (months); illness duration (months); side affected (right/left); and dominant side (right/left).
Statistical Analysis
Data analysis used IBM SPSS Statistics for Windows, version 26.0 (IBM Corp., Armonk, NY, USA). Sample characterization employed absolute (n) and relative (%) frequency for categorical variables and mean and standard deviation/median and interquartile range for continuous variables. The Shapiro-Wilk test assessed data normality. The Student t-test and the Pearson chi-squared test determined the homogeneity of sociodemographic features between case and control groups. Internal consistency, measured as Cronbach alpha, established the reliability of the HADS questionnaire.
A post-hoc chi-squared test, with Bonferroni correction, assessed the distribution of anxiety and depression symptoms from the HADS instrument between the case and control groups. If there were significant differences in the sociodemographic variables between groups, an analysis of covariance (ANCOVA) was performed to control its confounding effect when comparing HADS scores. The probability of rejecting the null hypothesis was 5%.
The Research Ethics Committee of the institution approved the present study.
RESULTS
From a total of 46 patients with AC admitted to the single-center outpatient clinic where the present study was performed, we excluded 9 for presenting the primary form of the disease, 10 for not accepting treatment, 8 for being unable to answer the questionnaire, and 2 because the disease was not characterized as primary or secondary. Thus, the final sample consisted of 17 patients in the case group and 27 subjects in the control group.
►Table 1 shows the sociodemographic and clinical data from both groups.
| Groups | Totaln= 44 | p-value | ||
|---|---|---|---|---|
| Case17 (38.6%) | Control27 (61.4%) | |||
| Mean±standard deviation | ||||
| Agea | 56.12 ± 10.37 | 48.44 ± 7.80 | 51.41 ± 9.55 | 0.01* |
| Educational levela | 12.13 ± 2.41 | 15.67 ± 5.09 | 14.40 ± 4.62 | 0.02* |
| Monthly incomeb | 8.12 ± 13.60 | 6.48 ± 4.84 | 7.03 ± 8.63 | 0.58* |
| Pain durationc | 5.85 ± 4.58 | na | ||
| Disease durationc | 5.41 ± 4.93 | na | ||
| n (%) | ||||
| Gender | ||||
| Female | 8 (47.1) | 15 (55.6) | 23 (52.3) | 0.58** |
| Male | 9 (52.9) | 12 (44.4) | 21 (47.7) | |
| Ethnicity | ||||
| White | 8 (47.1) | 21 (77.8) | 29 (65.9) | 0.07** |
| Non-white | 9 (52.9) | 6 (22.2) | 15 (34.1) | |
| Affected side | ||||
| Right | 7 (41.2) | na | ||
| Left | 10 (58.8) | |||
| Dominant side | ||||
| Left | 0 (0.0) | na | ||
| Right | 17 (100.0) | |||
*Student t test; **Pearson chi-squared test; ayears; bminimum wage; cmonths.
%, relative frequency; n, absolute frequency; na, not applicable.
Direct or indirect causes of secondary AC included the following: previous shoulder surgery (11.76%), shoulder mobility restriction after breast surgery (5.88%), trauma (fracture and instability; 11.76%), cervical radiculopathy (5.88%), rotator cuff tear (5.88%), calcareous tendinitis (5.88%), diabetes mellitus (29.41%), neuropathies (11.76%), and thyroid diseases (11.76%) (►Fig. 1).
The Shapiro-Wilk test sample revealed normal data distribution (p> 0.05). A HADS score > 0.70 was reliable and presented good internal consistency (►Table 2).
| Number of items | Groups | ||
|---|---|---|---|
| Case | Control | ||
| Anxiety | 7 | 0.76 | 0.71 |
| Depression | 7 | 0.72 | 0.74 |
| Total | 14 | 0.81 | 0.77 |
Cronbach's alpha.
►Table 3 shows the comparative distribution of anxiety and depression symptoms according to the HADS in both groups. Anxiety symptoms (> 11 points) with no depressive symptoms (≤ to 7 points) were predominant in the case group, with p= 0.001 and 0.037, respectively, when compared with the control group.
| Groups | Total | p-value* | ||
|---|---|---|---|---|
| Casen (%) | Controln (%) | |||
| Anxiety | ||||
| No symptoms | 6 (35.3) | 18 (66.7) | 24 (54.5) | 0.025 |
| Doubtful | 4 (23.5) | 8 (29.6) | 12 (27.3) | 0.074 |
| Anxiety symptoms | 7 (41.2) | 1 (3.7) | 8 (18.2) | 0.001 |
| Depression | ||||
| No symptoms | 12 (70.6) | 25 (92.6) | 37 (84.1) | 0.037 |
| Doubtful | 2 (11.8) | 2 (7.4) | 4 (9.1) | 0.200 |
| Depression symptoms | 3 (17.6) | 0 (0.0) | 3 (6.8) | 0.024 |
*Post-hoc chi-squared test; %, relative frequency; n, absolute frequency.
►Fig. 2 shows mean and confidence interval values for anxiety and depression symptom scores in the case and control groups.
The homogeneity analysis between the two groups revealed a significant difference in two variables, namely age and education (p< 0.05). Thus, when comparing the HADS scores of the two groups, an ANCOVA was performed, and this significant difference did not change the final analysis of our findings (►Table 4).
| Groups | Total | p-value* | ||
|---|---|---|---|---|
| CaseM ± SD | ControlM ± SD | |||
| Anxiety | 8.88 ± 4.50 | 5.67 ± 2.95 | 6.91 ± 3.91 | 0.019 |
| Depression | 6.41 ± 3.69 | 2.89 ± 2.89 | 4.25 ± 3.62 | 0.015 |
*ANCOVA (r2 = 0.14); covariables: age, educational level.
The group with AC reported “doubtful” “anxious symptoms” (score = 8.88) despite the difference compared with the control (p= 0.019). This group presented no “depressive symptoms” (score = 6.41), despite p= 0.015.
DISCUSSION
There was a “doubtful” positive association between anxiety symptoms and AC, with a mean HADS score of 8.88 (range: 8 to 10), despite the difference from the control group (p= 0.019). Regarding “depressive symptoms”, although there was a significant difference (p= 0.015) between mean values from the case (6.41) and control (2.89) groups, an association with AC was not determined because these figures are < 7.6-8 Our hypothesis was partially correct, because we expected to have confirmed, rather than doubtful anxiety symptoms.
A pioneering study by Fleming et al.14 evaluated the association between AC and personality traits using the Middlesex Hospital Questionnaire. The authors found evidence of personality differences from normal inpatients with adhesive capsulitis. Furthermore, they showed an association for somatic anxiety, especially in females.
Ding et al.4 carried out a similar study using the HADS scale and found a prevalence of depression and anxiety of 28.2 and 24.2%, respectively. Two mental parameters presented statistical differences compared with the control group; however, the mean values were < 7, indicating “no symptoms.” We believe there was an interpretation bias when analyzing these findings. Anyway, this corroborates our findings from the point of view of depression, but not of anxiety.
Bagheri et al.10 also evaluated patients with AC but used the Hamilton questionnaire for depression and anxiety. These authors detected only mild depressive symptoms, with no anxiety. Ebrahimzadeh et al.,5 coauthors of the Bagheri study, used the same questionnaire and concluded that depressive symptoms (77%) were more present than anxious symptoms (27%). Both studies are inconsistent with our findings, probably because of the score of the instrument, in which values > 8 indicate “depression” and those > 14 mean “anxiety.” Thus, he was more likely to have depressive symptoms with the use of the Hamilton questionnaire.
Beleckas et al.9 evaluated patients with upper limb disease using a Patient-Reported Outcomes Measurement Information System (PROMIS), including anxiety and depression. They concluded that some shoulder conditions, including AC, were associated with high anxiety scores, with a greater prevalence than depression in this population, maybe because this shoulder dysfunction is extremely painful. The illnesses that progressed with no pain had the lowest scores for anxiety and depression.
Debeer et al.15 conducted a study to determine personality differences between patients with AC using the NEO Five-Factor Inventory scale. They concluded that patients with AC did not have personality differences compared with the control group. However, this scale did not assess anxiety and depression but other personality traits.
Toprak et al.11 analyzed the relationship between AC and anxiety/depression using the Beck Anxiety Inventory and Beck Depression Inventory in addition to other outcomes. They showed a high prevalence and a close relationship between anxiety and AC, but not depression, which is in line with our findings.
Although the term “psychological distress” is used in studies of musculoskeletal conditions when referring to mental health,16-18 we prefer to refer separately to “anxiety” and “depression”, as several of the previously mentioned authors.4,5,11 This allows for a direct and individual comparison of each term with science that would not be possible otherwise.
The cross-sectional design of the present study limits our ability to determine any cause-effect relationship between AC and anxiety/depression, since some diseases may also have such outcomes. A nonprobabilistic, consecutive sampling may have imposed a selection bias, not allowing all patients to participate in the research. Moreover, not classifying the severity of AC as mild, moderate, and severe did not allow us to verify whether more rigid patients had more anxiety and/or depressive symptoms.13
Another issue would refer to anxiety, which showed clinical and statistical differences favoring the case group; however, the mean score was 8.88, characterized as “doubtful symptoms” according to the HADS. This was probably due to the sample size, and a higher sample number could result in a higher mean score. We must remember that we are not facing a type I error with falsely rejecting the null hypothesis because there was a difference, and the issue is in interpreting the analysis performed.
On the other hand, we used a reliable HADS (Cronbach alpha > 0.70), validated for the language of the studied sample and which assessed mental symptoms in situations of physical, nonpsychiatric alterations.19 The presence of a control group with healthy subjects allowed data comparison to expand the scientific knowledge on the “psychological status” of patients with AC, as it is a disabling disease with significant functional impairment.
It is essential to emphasize that our study was not aimed at diagnosing anxiety and depression as diseases but to seek an association with these symptoms. In any case, we strengthen the idea that we must recognize these potential associative changes to better direct patients with AC to specialized treatment along with orthopedic follow-up.20-23
CONCLUSION
There is a “doubtful” positive association between anxiety symptoms and AC, despite the difference from the control group (p= 0.019). Regarding depression, although there is a higher prevalence in the case group, it is not possible to infer an association, since the mean score was below the HADS reference.
