Repeated measurements of depression and outcomes in patients receiving hemodialysis†
Article Category: Original Article
Publicado en línea: 31 dic 2022
Páginas: 353 - 359
Recibido: 07 mar 2022
Aceptado: 22 jun 2022
DOI: https://doi.org/10.2478/fon-2022-0045
Palabras clave
© 2022 Lu Zhang et al., published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 International License.
Depression is the most common psychological complication associated with physical illness, but it is often undetected and/or left untreated. Depression may aggravate disabilities, exacerbate pain, reduce patient compliance, and affect prognosis.1
In patients with end-stage renal disease (ESRD), depression is the most common psychological complication.2,3 Among renal replacement therapies, hemodialysis is the most prevalent, and >75% of patients with ESRD are treated using hemodialysis.4 Hemodialysis significantly reduces the physical and psychological quality of life (QOL), and life expectancy.5 The lives of patients receiving hemodialysis are significantly and negatively impacted by multiple losses, such as kidney function, family roles, work competence, sexual function, time, and mobility.6 Other factors, such as medication effects6 dietary constraints, fear of death, and dependency on treatment,7,8 may also affect the QOL.
Compared with previous studies, our study more comprehensively analyzed the relationship between depression and patient prognosis. Our study not only investigated changes in depression but also investigated whether depression can be used to predict outcomes in patients receiving maintenance hemodialysis (MHD).
Outcomes data contained data regarding survival and mortality. In our study, human participants were investigated. All procedures performed in studies involving human participants were per the ethical standards of the institution and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants included in the study. Study participation was voluntary. Before the investigation, we informed them that we did not require their names and only required their physiological and physical details. Furthermore, the identities of the study participants are not published.
The inclusion criteria were as follows: (1) MHD >3 months; (2) age >18 years; and (3) ability to provide written informed consent to participate in the study. The exclusion criteria were as follows: (1) inpatient MHD patients; (2) patients with dementia or any psychotic disorder.
At baseline, year 1, and year 2, we asked all enrolled patients to answer the following questions. Outcomes data were collected from baseline to the end of the last follow-up.
Demographic characteristics such as age, sex, education level, marital status, family per capita monthly income, medical expenses to be paid by oneself, and employment status were collected from the patients. Data regarding the primary renal diagnosis, dialysis duration, and outcomes were obtained from medical records.
We investigated patients receiving hemodialysis, and then assessed their answers, choosing the item closest to the patients’ answers according to the frequency and intensity of the events within specific days.9 Subsequently, the patients were divided into two groups according to the HAMDS scores: the lower score group, including those with low HAMDS scores (between 0 and 7), and the higher score group, including those with high HAMDS scores (>7). We then compared changes in the HAMDS score and percentage at three time points.
Mortality was used as a dichotomized outcome. Age, dialysis duration, body mass index (BMI), medical expenses to be paid by oneself, and HAMDS scores were classified as continuous variables, whereas sex, education level, marital status, employment status, family per capita monthly income, and the primary renal diagnosis were classified as categorical variables. For continuous variables, the mean and standard deviation (SD) were calculated. For categorical variables, the proportion of individuals in each category was assessed.
The
Repeated-measure variance analysis was used to analyze changes in depression at three time points.
Survival analysis was conducted to test the hypothesis that depression may be used to predict outcomes in patients receiving hemodialysis. Cox proportional hazard models were used to calculate mortality and its 95% confidence intervals (CIs). Log-rank tests were used to compare the Kaplan–Meier estimates of event rates between the groups. The time of origin was the date of the first evaluation of depression. The ending events defined were mortality.
Statistical significance was established for
This was a longitudinal study. Patients receiving hemodialysis were investigated three times. Each investigation was spaced by approximately a year. After three follow-up surveys, of the 317 patients, removing those patients who interrupted treatment, and transferred to other dialysis centers and renal transplantation, 258 patients were able to track the final results, in which 97 died and 161 survived. Of the 161 survivors, except for those who refused to participate in the survey and invalid questionnaires, 124 patients finally completed three surveys.
On completion of the study, excluding patients who discontinued treatment, transferred to another dialysis center, or received a kidney transplant, a total of 97 patients had died and 161 had survived.
Table 1 presents a comparison of the demographic and illness characteristics of the deceased and surviving groups of patients at the time of the first evaluation. As shown in Table 1, age was the only factor that was significantly different between the groups.
Baseline demographic and illness characteristics.
| Characteristic | Deceased patients ( |
Surviving patients ( |
||
|---|---|---|---|---|
| 64.59 (12.27) | 59.98 (13.19) | −2.631 | 0.009 | |
| 2.282 | 0.157 | |||
| Male | 57 | 79 | ||
| Female | 40 | 82 | ||
| 3.892 | 0.143 | |||
| Primary school and below | 41 | 49 | ||
| Secondary school | 46 | 89 | ||
| College | 10 | 23 | ||
| 4.985 | 0.083 | |||
| Single | 3 | 5 | ||
| Married | 78 | 136 | ||
| Divorced/widowed | 16 | 20 | ||
| 2.807 | 0.246 | |||
| Working | 8 | 12 | ||
| Retired | 74 | 110 | ||
| Unemployed/students/ | 15 | 39 | ||
| 6.117 | 0.106 | |||
| <1000 | 14 | 10 | ||
| 1000–1999 | 46 | 77 | ||
| 2000–2999 | 19 | 45 | ||
| >3000 | 18 | 29 | ||
| 1731±1162 | 1867±1037 | 0.904 | 0.368 | |
| 39.91 (45.41) | 29.60 (34.99) | −1.903 | 0.057 | |
| 6.569 | 0.363 | |||
| Glomerulonephritis | 32 (33.0) | 50 (31.1) | ||
| Diabetic nephropathy | 21 (21.6) | 38 (23.6) | ||
| Hypertensive nephropathy | 21 (21.6) | 27 (16.8) | ||
| Obstructive nephropathy | 9 (9.3) | 7 (4.3) | ||
| Polycystic kidney | 4 (4.1) | 14 (8.7) | ||
| Unknown | 3 (3.1) | 6 (3.7) | ||
| Other* | 7 (7.3) | 19 (11.8) | ||
| 20.34 (3.22) | 20.77 (2.76) | −1.135 | 0.257 |
Repeated-measure variance analysis was conducted to analyze the changes in depression at three time points. The HAMDS scores increased at the three time points (Table 2). Moreover, the changes were statistically significant (
Changes in depression at three time points (
| Item | Baseline | 1-year follow-up | 2-year follow-up | ||
|---|---|---|---|---|---|
| HAMDS score (SD) | 4.74 (4.00) | 4.94 (3.84) | 6.50 (3.15) | 9.78 | <0.001 |
The patients were divided into two groups according to the HAMDS scores: the lower score group, including those with low HAMDS scores (between 0 and 7), and the higher score group, including those with high HAMDS scores (>7). The chi-square test was used to compare the percentages of the two groups at the three time points (Table 3). The results showed that the percentage of higher HAMDS scores increased. Moreover, the changes were statistically significant (
Changes in the percentage of depression at three time points,
| Items | Baseline | 1-year follow-up | 2-year follow-up | χ2 | |
|---|---|---|---|---|---|
| Lower score | 90 (72.6) | 87 (70.2) | 70 (56.5) | ||
| Higher score | 34 (27.4) | 37 (29.8) | 54 (43.5) | ||
| Total score | 124 (100) | 124 (100) | 124 (100) | 8.41 | 0.015 |
Table 4 shows that the surviving patients had significantly lower HAMDS scores (
Comparison of HAMDS scores between the deceased and surviving groups (n = 258).
| Items | Deceased patients ( |
Surviving patients ( |
||
|---|---|---|---|---|
| HAMDS score (SD) | 7.15 (3.14) | 6.30 (2.84) | −2.164 | 0.03 |
Age and HAMDS scores were analyzed based on the Kaplan–Meier method and log-rank test. As the age and HAMDS scores were all non-normally distributed, we chose the median as the cut-off value. Differences in age and HAMDS scores between the two groups were all statistically significant (
Hazard ratios for mortality (
| Groups | Total | Deceased patients | Incidence (%) | |
|---|---|---|---|---|
| Age (years) | 0.002 | |||
| 18–60 | 148 | 32 | 21.6 | |
| 61–91 | 169 | 65 | 38.5 | |
| HAMDS score | 0.037 | |||
| 1–6 | 175 | 46 | 26.3 | |
| 7–17 | 142 | 51 | 35.9 |
Using Cox proportional hazard models, as shown in Table 6, age and depression were found to be significantly associated with mortality, with mortality increasing by 1.032 for each point increase in age and increasing by 1.069 for each point increase in HAMDS score.
Results of Cox regression analysis models.
| Variable | Coefficient | Adjusted RR of death (95% CI) | |
|---|---|---|---|
| Age | 0.032 | <0.001 | 1.032 (1.015–1.050) |
| HAMDS score | 0.067 | 0.034 | 1.069 (1.005–1.138) |
Economic and social development has led to significant changes in demographics in all industrialized countries. The proportion of elderly individuals within the population has progressively increased. Similar demographic changes have also been noted in patients receiving hemodialysis. Our study showed that deceased patients were older than surviving patients; moreover, the age differences were significant between the groups (Table 1). Our study also showed that the mortality rate among elderly patients was higher than that among young patients (Table 5). Similarly, Ríos et al.’s10 study showed a similar significant association between increasing age and mortality. Pladys et al.’s11 showed that elderly patients receiving hemodialysis exhibit more comorbidities than those in young patients. Moreover, at the end of their study’s follow-up period, 69% of the patients had died in the elderly group. Blake et al.12 also showed that age is an independent predictor of poor physical function in patients with ESRD. With an increase in age, various functions of the body gradually degrade. At the same time, the compensatory ability and ability to withstand diseases also decline, and mortality increases significantly.
Depression is a frequently encountered complication in patients receiving hemodialysis. Depression may affect immune function, nutrition, and compliance factors, which may affect the therapeutic efficacy and outcomes of hemodialysis. In a similar study using the Hospital Anxiety and Depression Scale, morbid and borderline depression were observed in 33% and 55% of patients receiving hemodialysis, respectively.13 Despite being under diagnosed, depression appears to have an important influence on clinical outcomes. As depression has a high prevalence, it is the most important factor potentially increasing mortality and reducing QOL.14 Another similar study also showed that patients with chronic kidney disease receiving hemodialysis with depression have higher risks of death and hospitalizations compared with those in patients without depression.15
Our study was designed to investigate the association between depression and mortality in patients receiving hemodialysis, and the results showed that surviving patients have lower scores than those of deceased patients. Other studies have shown similar outcomes. A longitudinal study lasting for 1 year shows that deceased patients have higher Beck Depression Inventory (BDI) scores than those of surviving patients.16 Another longitudinal study showed that, after 24 months of follow-up, survival rates were 39% for patients with higher BDI scores and 95% for patients with lower BDI scores.17 Relevant interventions to alleviate depressive symptoms may improve patients’ outcomes.18 Depression may also affect the QOL in patients receiving hemodialysis. Belayev et al.19 showed that depression can independently reduce health-related QOL and global QOL. Thus, interventions to alleviate depression may potentially improve patients’ health-related QOL and global QOL. Depression affects not only mortality and QOL, but also other medical aspects. A study showed that a higher hospital admission rate as well as a higher likelihood of emergency department visits is observed in patients with depression.20 Because depression can lead to poor outcomes. More methods such as cognitive-behavioral therapy, exercise, and relaxation techniques probably decrease depressive symptoms for dialysis patients.21
Our study also showed that HAMDS scores increased (Table 2) and the percentage of higher HAMDS scores increased at the three time points (Table 3). The changes were statistically significant. A similar study showed that the symptoms of depression worsen over time in patients receiving hemodialysis. The authors of this study also suggest carefully monitoring patients receiving hemodialysis to avoid the under-recognition and under-treatment of depression.22 To the best of our knowledge, there have been a few longitudinal studies in which the effects of changes in depression and treatment for depression on mortality have been evaluated in patients receiving hemodialysis. Therefore, it is reasonable to hypothesize that treatment for depression in patients receiving hemodialysis might affect mortality.
Although our study showed that depression is associated with mortality in patients receiving hemodialysis, some other studies did not confirm this relationship. Zimmermann et al.23 showed that depression is not associated with mortality in patients with kidney disease. However, depression is a strong predictor of QOL. Compared with our study, Zimmermann’s study included fewer samples: a sample size of only 125. However, our study included 317 patients. Therefore, more similar studies with large sample sizes are needed.
The present study had some limitations. First, we investigated patients receiving hemodialysis at only two hospitals in Xi’an, China. However, the sample was representative of the patients receiving hemodialysis in the city of Xi’an. Thus, we are unaware if patients from other cities have a similar mortality rate. Second, we only used one depression scale (the HAMDS) to assess the presence and severity of depression. Therefore, other depression scales may be used to evaluate the mental state of patients receiving hemodialysis.
This study showed that depression in patients receiving hemodialysis was worse, and depression can be used to predict outcomes in patients receiving hemodialysis. Patients receiving hemodialysis should focus on improving their psychological complications.
The HAMDS may be used easily. According to the results of our evaluation, we can also predict outcomes in patients. More studies about depression related to hemodialysis are also needed. Despite the high prevalence rates, only a few patients are appropriately treated for depression. Further studies of the associations among depression, treatment, and outcomes in patients receiving hemodialysis are needed. We can then establish a psychological health promotion plan for patients receiving hemodialysis as part of routine practice.