Study design and sample
A cross-sectional study was analyzed in consecutive inpatients with PCa during January 2018 and August 2019. The study took place at the Department of Urology in our Hospital, which is the main provider of cancer services to a geographically defined area of 8.2 million people. The eligibility criteria for patient recruitment were (1) age 18 years or older, (2) being histologically diagnosed with PCa, (3) aware of the cancer diagnosis, (4) able to understand and read Chinese well enough to answer the questionnaires. Exclusion criteria were the following: (1) patients had a history of psychiatric problems before cancer diagnose, (2) patients had intellectual and/or cognitive impairments, (3) patients had other active cancers. All registered patients were all volunteers and anonymous. The volunteer patients’ attending physicians discussed eligibility on a case-by-case basis to avoid biased judgment and selection bias. After obtaining patient written consent, clinical data was collected from the medical record and a set of self-report questionnaires were distributed to patients at the time of hospitalization. Among a total of 667 registered patients, 10 patients refused to participate, and 5 patients had other active cancers. Of 652 eligible patients for this study, 88 were excluded from analysis (> 30% missing data). Finally, we received effective responses from 564 PCa patients with effective response rate 86.5%. The Committee on Human Experimentation of our hospital reviewed this study, provided the ethics for the approval of this study, and determined that the study procedures were in accordance with the ethical standards.
Beside demographic and cancer-related variables, the following questionnaires were adopted:
Depression was measured by the Center for Epidemiologic Studies Depression Scale (CES-D), which is a 20-item measure of the severity of depressive symptoms . Items are ranked on a four-point frequency scale from 0 (never) to 3 (always). Higher scores reflect worse depressive symptoms, and a score ≥16 indicates probable clinical depression . The CES-D was validated in cancer patients , and Cronbach’s alpha for CES-D was 0.794 in this study.
Perceived social support
Perceived social support was measured by the Multidimensional Scale of Perceived Social Support (MSPSS) , which is 12-item measure of the sources of perceived social support, rated on a seven-point scale from 1 (very strongly disagree) to 7 (very strongly agree). It provides a summary score (12 to 84), as well as three subscales for perceived support from family, friends, and significant others. The MSPSS was validated and commonly used in Chinese cancer patients [12,14]. In this study, the Cronbach’s alpha was 0.942, 0.923, 0.896, and 0.931 respectively for MSPSS, family, friends, and significant others subscales.
For hope, we correlated this with “hope” and this was measured by the Adult Hope Scale (AHS) which included eight items and four filler items rated on 4-point scales (1 = strongly disagree, 4 = strongly agree) . The AHS contains four Agency and four Pathways items, and a high score denotes a higher level of pathways and agency. The hope level is the sum of the agency and pathways items. The AHS and its Chinese version have been used in cancer patients with acceptable validity and reliability . The Cronbach’s alpha was 0.742, 0.772, and 0.849 respectively for the pathway, agency and AHS.
The 14-items version of Resilience Scale (RS-14), a short version of the original RS (i.e. RS-25), was used to assess resilience . RS-14 consists of 14 items rated on a 7-point scale, ranging from 1 (strongly disagree) to 7 (strongly agree). The total score ranges from 14 to 98 scores, with higher scores indicating higher resilience. The Chinese version of RS-14 had a good validity and reliability among cancer patients . In this study, the Cronbach’s alpha was 0.959 for RS-14.
The Statistical Package for the Social Sciences (SPSS, version 13.0) was used to perform the statistical analyses, with two-tailed probability value of < 0.05 considered to be statistically significant. The distributions of CES-D in categorical variables were calculated using independent sample t-test and one way analysis of variance (ANOVA). When one-way ANOVA was found to be significant, least-significant-difference (LSD) was done to perform multiple comparisons. Pearson’s correlation was used to examine correlations among psychosocial variables. Hierarchical regression analysis was used to explore the effects of perceived social support, hope and resilience on depression with adjustment for demographics and clinical variables related to depression in univariate analysis (p < 0.05). There were two models (Model 1 and Model 2) in Step 2. Total score of MSPSS was added in Model 1, and three subscales of MSPSS (others, friend and family) were added in Model 2. Due to the high correlations among the MSPSS subscales, these variables were adjusted in the stepwise regression in Step 2 (Model 2).We provided data including R2, adjusted R2 (Adj.R2), R2-changes, F, standardized regression coefficient (β) and p value for each step in the regression model. Asymptotic and resampling strategies were used to examine the mediating roles (a*b product) of hope and resilience on the association between perceived social support and depression . In these equations, perceived social support was modeled as the independent variable, CES-D score as the dependent variable, hope and resilience as the mediators. The auxiliary routine estimate was based on 5000 bootstrap samples. Then, the bias-corrected and accelerated 95% confidence interval (BCa95% CI) for each a*b product was investigated, and a BCa95% CI not including 0 indicated a significant mediating role. All study variables were centralized before analysis to account for differences in scale scores. Moreover, tolerance (> 0.10) and variance inflation factor (< 10) were used to check for multicollinearity.