The design of the study and recruitment of the study population
This study used a cross-sectional, descriptive, correlational design. Participants were recruited from three major communities in Beijing using convenience sampling. In the three communities, elderly patients with OA of the knee were asked to join the study. The inclusion criteria were: (a) age 60 years and older and (b) diagnosed with knee OA. The exclusion criteria were: (a) under antidepressant treatment, (b) unable to take part in the tests due to symptoms of severe disability, visual impairment, or nerve dysfunction, or (c) combined with a severe disease, such as Parkinson disease, cancer, or stroke. The sample size was determined using power analysis and multiple regression with the PASS2021 software (NCSS LLC, Kaysville, UT, USA). We tested 22 predictive variables and the regression model (R2) was set at 0.2 to explain the amount of variation. To achieve a power of 0.8 (alpha = 0.05), 91 participants (minimum) were required.
All the participants completed questionnaires, including the Short-Form Geriatric Depression Scale (GDS), the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), and the society dimension of Short-Form Arthritis Impact Measurement Scales 2 (AIMS2), which included providing personal information. In addition, the participants performed a timed up and go test (TUG) and the stair-climb test (SCT).
This study developed a demographic sheet to collect personal information, including age, sex, body mass index (BMI), marital status, ethnicity, residence, educational level, average monthly household income, pre-retirement occupation, medical expense, comorbidities, duration of the disease, number of diseased knees, a number of falls in the past year, use of walking aids, and consumption of pain relief or cartilage protectors.
The depression of the elderly was measured by using the Chinese version of the GDS. This scale comprises 15 self-reported measures about the patients’ feelings regarding their daily lives. The only acceptable answers were ‘yes’ (1) or ‘no’ (0). ‘Yes’ represented a positive score for depression in 10 of the questions; whereas, ‘no’ represented a positive score for depression in the other five questions, e.g., ‘Are you basically satisfied with your life?’ ‘Do you think it is wonderful to be alive now?’). The total score could range from 0 to 15, in which more depressive symptoms were indicated by higher scores. A score ³ 5 defined mild depression, and a score ³ 11 defined moderate/severe depression. Sensitivity and specificity of the GDS have been assessed in many studies and have been proven to have good validity and reliability.[23-25] When used in community-dwelling older adults, the Chinese eversion of the GDS had a split-half reliability of 0.84 and a Cronbach’s alpha score of 0.90.
Disease extent was assessed using the WOMAC index, which comprises a questionnaire with 24 items on three subscales: physical function (17 items, score range 0–68), stiffness (two items, score range 0–8), and pain (five items, score range 0–20). Higher scores indicated increased stiffness and pain, and worse physical function. The Chinese version of WOMAC has strong internal consistency (Cronbach’s alpha = 0.84–0.96) and acceptable test-retest reliability, as indicated by the intraclass correlation coefficients (ICC = 0.76–0.85) for all domains. For the WOMAC index, the present study achieved a Cronbach’s alpha score of 0.95, and a split-half coefficient of 0.80.
To measure the respondents’ social support, the society dimension of AIMS2 was used. AIMS2 measures arthritis patients’ quality of life using a self-assessment scale, comprising 26 items measured on a 5-point Likert scale. The society dimension contains four items, which are ‘Often get together with friends or relatives’, ‘Often make phone calls with friends or relatives’, ‘Often visit friends or relatives’, and ‘Family or friends are willing to help me to solve problems’. The range of possible scores was 0–20. Better family/friends social support was indicated by a higher score. An ICC of 0.90 and a Cronbach’s alpha score of 0.89 supported the reliability of society dimension of the Chinese version. In this study, the reliability of the social support assessment had a Cronbach’s alpha score of 0.77.
The TUG test is used to determine the basic mobility skill of patients with knee OA. The TUG test comprises determining the time taken to get up from a standard chair, walk 3 m, turnaround, walk back, and sit down again. Poorer mobility was indicated by a longer TUG test time. Participants were allowed one practice trial before a timed TUG test. The TUG test was proven to have an excellent test-retest reliability and discriminant validity in community-dwelling older people. The ICC of the TUG test was 0.97 in this study.
The stair-climb test (SCT) assesses the lower extremity strength of elderly people. The SCT comprised climbing five stairs to a platform, turning around, and then climbing down. The whole test process was timed until the participant walked down the last step. Longer times indicated weaker muscle strength and poorer physical coordination. Participants were allowed one practice trial before a timed performance. The test-retest reliability and concurrent validity of SCT were determined as excellent in a study of knee OA. In the present study, as assessed using the ICC, the test–retest reliability had a score of 0.96.
Analysis of the data
SPSS software version 23 (IBM Corp., Armonk, NY, USA) was used for all data analyses. Variables such as demographics, depression, WOMAC, social support, TUG, and SCT were assessed using descriptive statistics, i.e., the mean, standard deviation (SD), frequencies, and percentages. Among the study variables, correlations were determined using Pearson’s correlation coefficient. Predictors of depression were assessed using hierarchical multiple regression analysis. All statistical tests were two‑tailed and statistical significance was accepted at p £ 0.05.