Introduction

Hemodialysis (HD) is the most common therapy for managing patients with end-stage renal disease (ESRD). Patients with ESRD are in the last stage of chronic kidney disease (CKD), which is the point when kidney functions have ceased. Unless a kidney is available for transplantation, patients with ESRD require regular long-term dialysis to survive. Regardless of the form of dialysis (HD or peritoneal dialysis), patients with ESRD have a lower health-related quality of life than the general population. Several variables contribute to a lower health-related quality of life in patients with ESRD, including decreased physical function and social activities; increased risk of psychological distress; and symptom distress such as frailty, restless legs, itching skin, and fatigue (Ng et al., 2021). Moreover, the lifelong dialysis treatments needed to sustain the life of patients with ESRD significantly affect their physical and mental functions (Fradelos et al., 2021). According to a recent report, Taiwan had the highest prevalence of dialysis (3,593 per million population) and the second-highest incidence of treated ESRD (525 per million population) in the world in 2020 (U.S. Renal Data System, 2022).

Depression is one of the most common psychological problems faced by dialysis patients, with a prevalence rate ranging from 22.8% to 39.3% (Ma & Li, 2016), which is higher than the 12.9% reported for the general population (Lim et al., 2018). Dialysis patients with depression have higher mortality and hospitalization rates than patients without (L. Chan et al., 2017). Antidepressants, psychotherapy, dietary supplements, acupressure, and exercise therapy have been shown to be associated with reduced depressive symptoms, although the effect size is limited (Wen et al., 2020). Therefore, further studies examining additional factors associated with depression among dialysis patients with CKD and ESRD are needed.

A significant association between religion/spirituality and health-related conditions in the population of patients with CKD has been shown (Bravin et al., 2019). For example, Santos et al. (2017) surveyed 161 patients with CKD and found positive religious and spiritual coping scores to be negatively correlated with scores for depression and an independent protective factor for depression. Ramirez et al. (2012) reported similar findings, showing negative religious coping as positively associated with depression in HD patients. In addition, spiritual well-being has been shown to be negatively correlated with depression in patients receiving HD (Alradaydeh & Khalil, 2018). However, the studies on religion/spirituality in patients with CKD and dialysis are limited, and the definition of spirituality has been inconsistent, with spirituality and religion often used interchangeably and related measures not sufficiently comprehensive (Jugjali et al., 2018; Pilger et al., 2016). In the integrative literature review of Pilger et al., several studies were described that measured the concept of spirituality by organized and nonorganized religious activities, religiosity, or religious coping.

Although spirituality and religion are often considered synonymous, they are conceptually distinct and not interchangeable. First, spirituality emphasizes the realization of unique personal values and the meaning of life, whereas religion focuses on a set of beliefs that helps settle one's spirit and mind and is linked to a system of worship and a structured organization (Hsiao et al., 2013). In addition, most measurement scales for spirituality include spiritual and religious subscales that include items focused on religiosity, which may not be applicable to those who are not religious (Hsiao et al., 2013).

To the best of our knowledge, the influence of religion and spirituality on depression has not been investigated in the same study. Therefore, in this study, religion and spirituality were investigated separately. For religion, three categories used to measure religiosity in previous studies were used (Chiang et al., 2020a; Koenig et al., 2012), including religious affiliation, religious activities, and religious beliefs. Religious affiliation assesses if an individual associates themselves with a specific religious faith such as Catholicism, Judaism, Muslim, or Buddhism. Religious activities are defined as attending public or private activities associated with a religion such as praying or reading religious texts. Religious beliefs refer to the perception that adhering to articles of faith associated with a religion is an important aspect of one's life. These three domains were used to offer a more comprehensive assessment of the concept of religion in this study.

Although previous studies have explored the associations between depression and religion/spirituality in patients receiving HD, their participants were from primarily from Western countries (Bravin et al., 2019; Loureiro et al., 2018). Eastern countries, including Taiwan, are grounded in cultures and belief systems that differ from Western countries, which largely draw on Judeo-Christian religious beliefs. Thus, there are gaps in the literature regarding religion/spirituality-related variables that may impact depression in patients on long-term HD in Taiwan. To fill this gap, differences in religion and spirituality between patients receiving HD in Taiwan with and without depression were explored and elucidated.

Guided by the prior literature, the aims of this study were to (a) examine religion using measures of religious affiliation, religious activities, and religious belief; (b) examine the total score and the five subscale scores of a spiritual health instrument; and (c) determine the significant sociodemographic, disease-related, religion-related, and spiritual-health-related variables. Our three hypotheses were as follows: (a) Variables for the three religion-related measures differ between the two groups; (b) variables of spiritual health (total score and subscale scores) differ between the two groups; and (c) sociodemographic, disease-related, religion-related, and spiritual-health-related variables are predictors of depression in patients receiving HD.

Methods Study Design, Setting, and Participants

A cross-sectional correlational study was conducted on a convenience sample of patients with CKD at an HD clinic of a nephrology department in northern Taiwan. Data were collected from July to November 2020. The results were reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology checklist. The inclusion criteria were as follows: (a) over 20 years old, (b) receiving dialysis regularly for CKD for more than 3 months, (c) absence of a cognitive disorder and able to communicate verbally, and (d) voluntary participation. The exclusion criteria were as follows: (a) a diagnosis of psychosis or other psychiatric disorder and (b) use of an antidepressive medication. The sample size was estimated post hoc using G*Power Version 3.1.9.2. The statistical power for logistic regression analysis was > .80 (.87), which indicated a sample size of 137 patients was appropriate.

Instruments

The data were collected using a sociodemographic and disease-related questionnaire, religion measurement, spiritual health measurement, and depressive tendency scale. All of the measurements had established acceptable validity and reliability for self-reported data.

Sociodemographic and disease-related questionnaire

The sociodemographic and disease-related information collected included age, gender, marital status, educational level, job status, source of economic support, duration of HD, and number of chronic diseases.

Religion measurement

As described below, we used validated instruments to assess the following three aspects of Eastern religions considered components of “religion”: religious affiliation, participation in religious activities, and religious beliefs.

Religious affiliation

Religious affiliation was established using the Religious Affiliation Scale of Chang and Lin (1992), which distinguishes between three religious affiliation categories: primary, secondary, and atheist. The populations of Taiwan and many other ethnically Chinese societies tend to be multireligious. Therefore, participants were not categorized by a specific religious following such as Catholicism, Judaism, Muslim, Taoism, or Buddhism. Instead, participants were asked to rank themselves by the level of the religions they practiced. The participants were given three choices: primary, which is described as having a belief in a higher spiritual power or god and a very clear perception of their religious affiliation and religious practice; secondary, which is described as having no firm religious affiliation but having a belief in a higher spiritual power or god and participating in some religious activities; and atheist, which is described as having no belief in a god or higher spiritual power and not participating in religious activities.

Religious activities

Religious activities were assessed using a five-item instrument originally developed by Chiang et al. (2020a) to evaluate the spiritual life of clinical nurses. Five items are used to assess level of involvement in various religious activities over the past 6 months, including participating in religion-related events, praying, reading the bible/religious texts, meditating on religious incantations, and seeking religious guidance to make decisions. The validity of this instrument was established using exploratory factor analysis with one extracted factor explaining 48.0% of the variance, and factor structure validity was verified using confirmatory factor analysis. A 4-point Likert scale was used for scoring, with scores ranging from 4 = often to 1 = never. The total possible score ranges from 5 to 20, with higher scores indicating greater participation in religious activities. The Cronbach's alpha of the scale was .87 in this study.

Religious beliefs

The 17-item Religious Beliefs Scale (RBS) developed by Chiang et al. (2017) was designed to measure positive and negative religious beliefs. The RBS has acceptable reliability and validity, has a Cronbach's alpha of .87, and is composed of four subscales. The first two, “religious effects” and “divine,” are categorized as positive religious beliefs (with 12 items, e.g., “obtain support and assistance from religious groups,” “believe that happiness and peace are gifts from god/higher power”). The second two, “religious query” and “religious stress,” are categorized as negative religious beliefs (with five items, e.g., “Sometimes, I feel God is unfair to me,” “I worry about being punished by god if I do not comply with religious rules”). Each subscale item is scored on a 5-point Likert scale ranging from 1 = strongly disagree to 5 = strongly agree. The Cronbach's alpha of the RBS was .81 in this study.

Spiritual health

Spiritual health was defined in this study as one's own inner forces and resources that allow their developing a unique meaning of the “self,” which is reflected through their connections with themselves, others, and/or a higher being (Hsiao & Huang, 2005). The Spiritual Health Scale-Short Form developed by Hsiao et al. (2013) was used in this study to assess spiritual health. This instrument is composed of 24 items with the following five subscales: connection to others (four items), such as “I like to help solve problems for family members”; meaning derived from living (six items), such as “I think about how to have a more fulfilled life”; transcendence (six items), such as “I see setbacks as a type of challenge”; religious attachment (four items), such as “I hope to be blessed by God”; and self-understanding (four items), such as “I believe I am a person with value.” Each item is scored on a 5-point Likert scale. Responses range from 1 = strongly disagree to 5 = strongly agree, and total scores range from 24 to 120, with higher scores indicating better spiritual health. In this study, the Cronbach's alpha of the Spiritual Health Scale-Short Form was .92.

Depressive tendency

The 21-item Beck Depression Inventory-II (Beck et al., 1996/2000) assesses depressive symptoms over the previous 2-week period. The Chinese version of the Beck Depression Inventory-II (C-BDI-II) translated by H. Y. Chen (2000) for use in Taiwanese populations was applied in this study. This 21-item self-report scale consists of questions aimed at measuring the presence and severity of depressive symptoms using a 4-point Likert scale with item scores ranging from 0 to 3. The total scores ranges from 0 to 63 points, with higher scores indicating greater depression severity. The total score may be used to distinguish among four different levels of depression: 0–13, no depression; 14–19, mild depression; 20–28, moderate depression; and 29–63, severe depression. In this study, participants with scores ≤ 13 were assigned to the nondepressed group, and those with scores > 13 were assigned to the depressed group. In this study, the Cronbach's alpha for the C-BDI-II was .71.

Data Collection Procedures

After providing informed consent, the participants completed the questionnaire during a clinic visit. If the patient was unable to complete the questionnaire in writing, a researcher asked each question orally and filled in the answers. All of the questionnaires were coded with a number to ensure data anonymity.

Ethical Considerations

This study was approved by the research ethics committee of the participating hospital (IRB No. 202000959B0). All of the eligible patients were provided with oral and written information by the first author about the study and procedures and were assured that participation was voluntary and that no impact on their care would result because of their decision to participate or not. Furthermore, they were assured they could withdraw from the study without consequence at any time and for any reason and that the anonymity of their data would be maintained. All of the collected questionnaires were stored in a locked box in the office of the principal investigator.

Data Analysis

The data were analyzed using SPSS 23.0 software (IBM Inc., Armonk, NY, USA). Descriptive statistics included mean and standard deviation (SD) for continuous variables and frequencies and percentages for categorical variables. With regard to inferential statistics, independent t tests and chi-square tests were used to compare sociodemographic and disease-related characteristics between the nondepressed and depressed groups. Independent t tests were used to compare differences between the groups, whereas point-biserial correlations (rpbi) were employed to measure the relationship between groups on religious activities, positive and negative religious beliefs, and total and subscale scores for spiritual health. Logistic regression analysis was used to examine the relationship between depression and potential predictor variables. Significant predictors for risk of depression in the model were presented as odds ratios (ORs) to indicate the odds of depression compared with the odds of no depression. The potential predictors of depression were identified as variables that differed significantly between the two groups. Significance was set at p < .05.

Results Participant Characteristics

One hundred thirty-seven HD patients participated in the study. The mean age was 60.28 (SD = 13.83) years. Most were male (63.5%) and unemployed (65.7%) and had a primary or secondary religious affiliation (57.7% and 22.6%, respectively). Half (51.1%) of the participants had 7–12 years of formal education. The 37.5% who scored ≤ 13 on the C-BDI-II were assigned to the non-depressed group, and the 87 (63.5%) who scored > 13 were assigned to the depressed group. The participants in the depressed group were significantly older, less educated, and more likely to be male, unemployed, and atheist than their nondepressed peers. No significant difference in source of economic support or number of chronic diseases was found between the groups. More than half (59.8%) of the depressed group had been on dialysis for 1–5 years, compared with 32% for the nondepressed group. Details of the characteristics for all participants and for each group are shown in Table 1.

Table 1. - Characteristics of All Participants and Differences Between Nondepressed and Depressed Groups (N = 137) Variable All
(N = 137) Nondepressed a
(n = 50) Depressed b
(n = 87) t/χ2 n % n % n % Demographic characteristics  Gender 4.50*   Male 87 63.5 26 52.0 61 70.1   Female 50 36.5 24 48.0 26 29.9  Age (years; mean and SD) 60.28 13.83 56.30 13.49 62.56 13.58 −2.61*   Range 20–92 23–88 20–92  Marital status 0.02   Single 17 12.4 6 12.0 11 12.6   Married 95 69.4 35 70.0 60 69.0   Other c 25 18.2 9 18.0 16 18.4  Years of education 7.72*   0–6 35 25.5 6 12.0 29 33.3   7–12 70 51.1 31 62.0 39 44.9   ≥13 32 23.4 13 26.0 19 21.8  Employment status 6.55*   Unemployed 90 65.7 26 52.0 64 73.6   Employed 47 34.3 24 48.0 23 26.4  Source of income 3.22   Self 94 68.6 39 78.0 55 63.2   Others 43 31.4 11 22.0 32 36.8  Religious affiliation 11.39**   Primary 79 57.7 37 74.0 42 48.3   Secondary 31 22.6 10 20.0 21 24.1   Atheist 27 19.7 3 6.0 24 27.6 Disease characteristics  Duration of hemodialysis 10.56**   ≥ 3 months to ≤ 1 year 34 24.8 15 30.0 19 21.8   > 1 year to ≤ 5 years 68 49.7 16 32.0 52 59.8   > 5 years 35 25.5 19 38.0 16 18.4  Number of chronic diseases (mean and SD) 1.66 0.86 1.50 0.86 1.75 0.85 −1.63

Note. Religious affiliation = identified on the Religious Affiliation Scale.

a A score ≤ 13 on the Chinese Beck Depression Inventory-II. b A score > 13 on the Chinese Beck Depression Inventory-II. c Divorced, widowed, or separated.

*p < .05. **p < .01.

Between-Group Differences in Terms of Religion and Spiritual Health

Scores for religious activities and positive or negative religious beliefs as well as total and subscale scores for spiritual health for all of the participants and for the two groups are shown in Table 2. Mean scores for the nondepressed group were higher than those for the depressed group for religious activities (t = 3.42, p < .001), with scores positively correlated with nondepression status (rbpi = .28, p < .01). Positive religious beliefs were found to be higher in the nondepressed group than the depressed group (t = 4.14, p < .001) and positively correlated with nondepression (rbpi = .34, p < .001). Conversely, negative religious beliefs were found to be lower (t = −4.80, p < .001) in the nondepressed group and negatively correlated with nondepression (rbpi = −.38, p < .001). In addition, mean total scores for spiritual health (t = 9.17, p < .001) and the five subscale scores (t = 4.69–7.40, all ps < .001) were all higher in the nondepressed group and were all correlated positively with nondepression (all ps < .001).

Table 2. - Scale Scores for Religion and Spiritual Health: All Participants (N = 137) and Comparison Between Nondepressed and Depressed Participants Scale No. of Items Score Range Min–Max Scale Scores t r pbi All Participants
(N = 137) Nondepressed
(n = 50) Depressed
(n = 87) Mean SD Mean SD Mean SD Religious activities 5 5–20 9.04 3.78 10.44 3.92 8.23 3.47 3.42*** .28** Religious beliefs  Positive 12 12–60 36.62 7.35 39.8 7.33 34.76 6.72 4.14*** .34***  Negative 5 5–30 13.12 2.60 11.82 2.62 13.87 2.29 −4.80*** −.38*** Spiritual health, total score 24 24–120 78.12 11.21 86.94 7.83 73.06 9.63 9.17*** .60***  Subscale scores   Connection to others 4 4–20 14.30 2.46 15.74 1.71 13.47 2.45 6.35*** .45***   Meaning derived from living 6 6–30 17.90 3.40 20.14 2.63 16.68 3.14 6.57*** .49***   Transcendence 6 6–30 19.80 3.02 21.92 2.62 18.56 2.52 7.40*** .54***   Religious attachment 4 4–20 12.99 3.20 14.56 2.87 12.08 3.04 4.69*** .37***   Self-understanding 4 4–20 13.11 2.45 14.58 1.70 12.26 2.42 5.96*** .46***

Note. Min = minimum; Max = maximum; rpbi = point-biserial correlation coefficient.

**p < .01. ***p < .001.

Association Between Religion, Spiritual Health, and Depression

A logistic regression of the significant demographic, clinical characteristic, religion-related, and spiritual-health-related variables was used to identify the predictors of depression (Table 3). The model was significant (χ2 = 76.15, p < .001), providing evidence that all of the variables account for variance in depression, a moderate level of correlation (Nagelkerke R2 = .58). The goodness-of-fit index based on the Lemeshow and Hosmer test (used for the logistic regression analysis) was 7.36 (p > .05). Regression results showed risk of depression as higher for participants with 1–5 years of HD treatment than for those with treatment durations > 5 years (OR = 3.64, 95% CI [1.01, 13.15]). Religion-related variables were not found to correlate with depression. Risk of depression was shown to be lower in participants with higher spiritual health total scores (OR = 0.82, 95% CI [0.75, 0.90]).

Table 3. - Logistic Regression Analysis: Factors Associated With Depression Variable Odds Ratio 95% CI Lower Upper Gender, male (female a) 1.79 0.56 5.73 Age 1.00 0.95 1.04 Years of education (> 13 years a)  0–6 0.84 0.14 4.99  7–12 0.55 0.17 1.79 Employment status, unemployed (employed a) 2.76 0.79 9.63 Religious affiliation (atheist a)  Primary 2.71 0.26 28.78  Secondary 1.86 0.25 14.18 Duration of hemodialysis (> 5 years a)  ≥ 3 months to ≤ 1 year 0.79 0.20 3.14  > 1 year to ≤ 5 years 3.64* 1.01 13.15 Religious activities 0.94 0.77 1.13 Positive religious beliefs 1.02 0.91 1.14 Negative religious beliefs 1.11 0.89 1.39 Spiritual health 0.82** 0.75 0.90 Model χ2 (df) b 76.15*** (13) Goodness-of-fit χ2 (df) c 7.36 (8), p = .498 Nagelkerke R