Abstract
Adults with dialysis-dependent ESKD experience higher rates of depression than the general population, yet efficacy of depression treatments in this population is not well understood. We conducted a systematic review of the benefits and harms of depression treatment in adults with ESKD. We searched multiple data sources through June 2020 for English-language, controlled trials that compared interventions for depression in adults with ESKD to another intervention, placebo, or usual care, and reported depression treatment–related outcomes. Observational studies were included for harms. Two investigators independently screened all studies using prespecified criteria. One reviewer abstracted data on study design, interventions, implementation characteristics, and outcomes, and a second reviewer provided confirmation. Two reviewers independently assessed study quality and resolved any discords through discussion or a third reviewer. Strength of evidence (SOE) was assessed and agreed upon by review-team consensus. We qualitatively analyzed the data and present syntheses in text and tables. We included 26 RCTs and three observational studies. SSRIs were the most studied type of drug and the evidence was largely insufficient. We found moderate SOE that long-term, high-dose vitamin D3 is ineffective for reducing depression severity. Cognitive behavioral therapy is more effective than (undefined) psychotherapy and placebo for depression improvement and quality of life (low SOE), and acupressure is more effective than usual care or sham acupressure in reducing depression severity (low SOE). There is limited research evaluating treatment for depression in adults with ESKD, and existing studies may not be generalizable to adults in the United States. Studies suffer from limitations related to methodologic quality or reporting. More research replicating studies of promising interventions in US populations, with larger samples, is needed.
Systematic Review registry name and registration number: PROSPERO, CRD42020140227
Introduction
The incidence and prevalence of ESKD in the United States have increased steadily over the past four decades (1). Psychiatric and mental-health disorders are more common in adults with ESKD, and include issues such as depression, dementia, delirium, substance abuse, psychoses, anxiety, and personality disorders (2⇓–4). Adults with ESKD experience major depressive disorder at anywhere from three to over six times that of the general US population, depending on the method of assessment, and depression is the most common mental-health issue in this population (5,6).
The effect of depression can be extremely detrimental for a wide range of patient outcomes for those with ESKD. For instance, adults with ESKD who have depressive symptoms have a 12% higher rate of hospitalization than those without (7), and those with ESKD and psychiatric issues have a 40%–50% increased risk of all-cause mortality (8). Comorbid depression is associated with increased emergency-department visits, hospitalizations, hospital length of stay, suicide, and treatment nonadherence, along with decreased quality of life (QoL) and sleep (8⇓⇓⇓⇓–13).
There are no established guidelines for treating depression in adults with ESKD. Less than 25% of those on dialysis who are diagnosed with moderate or severe depression actually undergo treatment (12,14,15). There are few studies of treatment efficacy in this population and, while some studies include only participants with clinical depression, others include any adults with ESKD. Psychosocial treatments and cognitive behavioral therapy (CBT) are commonly used; however, interventions vary widely, the evidence is limited, and findings may vary on the basis of the presence and severity of depression.
Given the wide variation in depression treatment options for adults with ESKD, it is vital to understand the depression treatment–related outcomes for those in this population suffering from clinical depression. The purpose of this review is to better understand the benefits and harms of treatment for depression in these adults.
Materials and Methods
Topic Development
This systematic review was part of a larger review commissioned by the Veterans Health Administration (VHA) (16). A protocol describing the review plan was posted to PROSPERO, a systematic review registry, before study initiation (CRD42020140227). Our methods and reporting follow Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (17). See Supplemental Table 1 for the scope and parameters of the systematic review.
Data Sources and Searches
To identify trials examining the treatment of depression in adults with ESKD, we searched Ovid MEDLINE, PsycINFO, Elsevier EMBASE, and the Ovid Evidence-Based Medicine Reviews (Cochrane Database of Systematic Reviews, Database of Abstracts of Reviews of Effects, Health Technology Assessment, Cochrane CENTRAL, etc.) from database inception through June 2020. We reviewed the bibliographies of systematic reviews and other relevant articles, and contacted experts to identify additional studies. To identify unpublished literature, we searched the VHA Health Services Research and Development website, ClinicalTrials.gov, and the World Health Organization International Clinical Trials Registry Platform. Search strategies were developed in consultation with a research librarian and peer reviewed by another research librarian using the instrument for Peer Review of Search Strategies (see Supplemental Appendix 1) (18).
Study Selection
Studies were eligible if they (1) included adults with dialysis-dependent ESKD not slated for transplant and depression, defined by established thresholds for chronically ill populations (19⇓⇓⇓–23); (2) directly compared any pharmacologic or nonpharmacologic treatments for depression to placebo, wait-list control, or other intervention; (3) were randomized controlled trials (RCTs) or nonrandomized controlled trials in any setting (for harms outcomes, we also included observational studies); (4) assessed patient outcomes of interest (e.g., depression symptom severity, suicidal ideation, QoL); and (5) were published in English. We excluded studies examining adults with AKI or those with CKD stages 1–4. See Supplemental Appendix 2 and Supplemental Table 1 for complete selection criteria. All studies were reviewed for inclusion by two independent reviewers at the title/abstract and full-text levels. Discords were resolved through consensus or consultation with a third reviewer.
Data Abstraction and Quality Assessment
From included studies, we abstracted details on study design, sample size, setting, population characteristics, participant selection criteria, and intervention details (including the dosage, timing, and administration methods; duration of treatment and follow-up; outcomes; and relevant harms). Data from studies meeting inclusion criteria were abstracted by one reviewer and confirmed by an additional reviewer.
Two reviewers independently assessed the methodologic quality of each study using criteria established by the US Preventive Services Taskforce and adapted for depression interventions (24⇓–26). Disagreements were resolved by consensus or a third reviewer.
Data Synthesis and Analysis
We qualitatively synthesized the evidence for each treatment category and outcome of interest. We were unable to quantitatively synthesize the evidence because there were too few studies examining the same intervention and reporting the same outcome measure (52).
Using an established method by Berkman et al. (53), for each intervention, we assessed the overall strength of evidence (SOE) that considers study limitations, directness, consistency, precision, and reporting bias to classify the SOE for each outcome independently as high, moderate, low, or insufficient; we used separate guidance for the applicability (external validity) of the evidence to the clinical question (54). Supplemental Table 2 describes SOE domains and grading in more detail.
Results
We reviewed 8050 titles and abstracts, 192 of which qualified for full-text review. Of those, we included a total of 26 RCTs; nine examined pharmacologic interventions and 17 examined nonpharmacologic interventions for the treatment of depression in adults with ESKD. We also included three observational studies reporting harms of selective serotonin reuptake inhibitors (SSRIs; see Figure 1 for literature flow, and Table 1 for study characteristics).
The flow diagram of the literature-review process shows 26 RCTs and 3 observational studies. *After deduplication. CDSR, Cochrane Database of Systematic Reviews; DARE, Database of Abstracts of Reviews of Effects; EBM, Evidence-Based Medicine; HSR&D, Health Services Research and Development Service; HTA, Health Technology Assessment; RCTs, randomized controlled trials; VA, Veterans Affairs; WHO ICTRP, World Health Organization International Clinical Trials Registry Platform.
Characteristics of randomized controlled trials of interventions for depression and studies examining harms of depression treatment in adults with ESKD
Pharmacologic Treatments
SSRIs
SSRIs versus Placebo
Three studies comparing SSRIs with placebo report conflicting findings and provide insufficient evidence to draw conclusions about their effectiveness in treating depression in adults with ESKD (see Table 2 for study results and Table 3 for SOE ratings). Two small US RCTs (one 8-week, poor-quality trial of fluoxetine [N=14] [30], and one recent, fair-quality, 6-month trial of sertraline [A Study of Sertraline in Dialysis; ASSertID; N=30] [33]) found no difference in depressive-symptom reduction between those assigned to SSRIs or placebo. One fair-quality, 12-week, Iranian RCT (N=50) (46) found that participants who received sertraline reported a significant reduction in depressive symptoms compared with placebo. Small sample sizes and differences in depression assessment tools and statistical analyses detracted from the quality of these studies (see Table 4 for quality ratings).
Efficacy of interventions from randomized controlled trials for depression in adults with ESKD
Summary of the evidence on interventions for depression in adults with ESKD
Quality and applicability assessment of randomized controlled trials
SSRIs versus Active Comparators
SSRIs versus CBT
A recent, fair-quality, multisite, head-to-head RCT (N=120) in the United States (A Trial of Sertraline versus Cognitive Behavioral Therapy for ESKD Adults with Depression; ASCEND [44]) provides low-strength evidence that sertraline and CBT are similar when used for the reduction of depressive symptoms (Tables 3 and 4). Over the 12-week study period, both groups improved significantly. Participants who received sertraline experienced significantly greater improvement (effect estimate, −1.85; 95% CI, −3.55 to −0.16; Table 2) when assessed by a clinician (Quick Inventory of Depressive Symptomatology). There was no difference between groups in self-reported symptoms (i.e., Beck Depression Inventory-II; effect estimate, −2.9; 95% CI, −6.7 to 0.8; Table 2).
SSRIs versus Psychologic Training
A small (N=44), 3-month, poor-quality RCT (38), conducted in Iran, provides insufficient evidence for the comparison of citalopram to “psychologic training” in participants with ESKD and depression (Tables 3 and 4). Although both arms reported a reduction in depressive symptoms, there was no difference between groups (Table 2).
Harms of SSRIs
Adverse events (AEs) reported in the included trials of SSRIs in adults with depression and ESKD were similar in type and frequency to those reported by the general population on SSRIs (59). One trial reported a higher rate of dropout due to AEs associated with sertraline (33% versus 0%; P=0.04) (33). Across all four trials (30,33,44,46), a wide range of AEs (e.g., nausea, headaches, dizziness) were reported by participants in both treatment and control groups, but they were not consistently nor uniformly reported (Table 5).
Adverse events reported in depression treatment trials in patients with ESKD
Additionally, three observational studies focused specifically on potential harms related to SSRIs in adults with ESKD and depression (Table 6). Two examined Medicare beneficiaries who received SSRIs that were included in the US Renal Data System registry. A case-control study (N=54,032) found that SSRIs increased the risk of hip fracture regardless of dose or duration (58), and the second (57), a retrospective cohort study (N=65,654), found that adults, especially older adults and women, taking SSRIs with higher QT-prolonging potential (i.e., citalopram and escitalopram) were at higher risk for sudden cardiac death (adjusted hazard ratio, 1.18; 95% CI, 1.05 to 1.31). The final study was a follow-up to the ASSertID RCT (33) that examined SSRI practice patterns in adults with ESKD, and included participants who were already on SSRIs at baseline (N=41). Findings suggest poor medication management and both under- and over-treatment (56).
Adverse events reported in observational studies of adults with ESKD and depression
Supplements versus Placebo
High-dose Vitamin D3
A large (N=746), fair-quality, 52-week RCT (50) provides moderate-strength evidence that long-term, high-dose vitamin D3 does not reduce depression severity in adults with ESKD (Tables 3 and 4). In addition, no differences were reported by age, sex, body mass index, or plasma albumin level. Although overall differences in depressive-symptom reduction between groups were NS, a subgroup of participants with vascular depression (but not those with major depressive disorder) receiving vitamin D3 did report significantly greater reduction in depressive symptoms at 1 year (Table 2).
Harms of Vitamin D3
AEs associated with high-dose vitamin D3, including joint pain, diarrhea, nausea, and vomiting, resulted in study withdrawal of five participants. No statistical analyses of AEs or withdrawals due to AEs were reported (50).
Ω-3 Fatty Acids
A single, poor-quality RCT (N=54) (35), conducted in Iran, provides insufficient evidence to form conclusions about the effect of Ω-3 fatty acids versus placebo on depression severity in adults with ESKD (Tables 3 and 4). At 16 weeks, the Ω-3 fatty acids group reported a significantly greater reduction in depressive symptoms compared with both placebo and their own baseline. No differences were identified by age, sex, baseline depression severity, or length of time on hemodialysis (Table 2). No serious AEs (SAEs) were reported in this trial.
Synbiotic and Probiotic Supplements
One good-quality, 12-week RCT (N=49) (36) provides insufficient evidence for the use of synbiotic or probiotic supplements versus placebo on depression severity in Iranian adults with ESKD and Hospital Anxiety and Depression Scale–Depression (HADS-D) scores of eight or higher (Tables 3 and 4). HADS-D scores at the end of treatment were significantly lower in adults receiving synbiotics than those receiving placebo (P<0.001) and probiotics (P=0.01). HADS-D scores were also significantly reduced from baseline in the symbiotic compared with placebo (P<0.001) groups, but not in the other comparisons (Table 2). The reported AEs were few and mild (Table 5).
Radix Bupleuri Herbal Supplement
A poor-quality, 3-month RCT (N=160) (49) of Radix Bupleuri root powder supplements compared with placebo provides insufficient evidence for its use for depression in adults with ESKD (Tables 3 and 4). Participants who were being treated with antidepressant medications (about half of each group) continued on those treatments concurrently during this trial. Participants taking Radix Bupleuri experienced significantly more reduction in Montgomery–Åsberg Depression Rating Scale scores compared with the placebo group (P=0.02; Table 2). They also experienced improvement in QoL (measured by RAND-36) compared with the placebo group (P=0.04). The trial did not report on AEs.
Nonpharmacologic Treatments
CBT
CBT versus Psychotherapy
One fair-quality RCT (N=90) (32) provides low-strength evidence that small-group CBT is more effective than brief, individualized psychotherapy for reducing depression severity in participants with ESKD (Tables 3 and 4). Depressive symptoms improved significantly in both groups. Participants receiving CBT also experienced a significant within-group decrease in suicide risk and improved on several QoL domains (i.e., burden of kidney disease, quality of social interaction, sleep, overall health, and mental health) over the study period (Table 2). No study dropouts from serious AEs were reported.
CBT versus Psychoeducation
A poor-quality RCT (27), conducted in Jordan, provides insufficient evidence to form conclusions about the comparison of CBT with psychoeducation (Tables 3 and 4). Participants receiving CBT reported a significantly greater reduction in depressive symptoms (both groups reported significant improvement; Table 2).
CBT versus Sertraline
A head-to-head RCT (44) provides low-strength evidence that sertraline and CBT are similar when used for the reduction of depressive symptoms in participants with depression and ESKD. See the section SSRIs versus CBT above and Tables 2⇑–4 for more detail.
CBT versus Control
Two fair-quality RCTs (31,41) provide low-strength evidence that CBT is more effective than wait-list control for reducing depression severity and improving QoL in participants with ESKD (Tables 3 and 4). There is insufficient evidence to form conclusions about the benefit of CBT on fluid adherence (31).
One study compared outcomes (depressive symptoms, QoL, and fluid adherence) on the basis of the timing of the intervention (treatment first or after 90-day wait listing). Results indicated that participants receiving CBT experienced significantly greater benefit across all outcomes in both phases. However, findings suggest a sequence effect for depressive-symptom reduction (the treatment-first group experienced greater benefit than the wait-list group), but none for QoL or fluid compliance (Table 2) (31).
Acupressure
We found low-strength evidence that acupressure is more effective than both usual care and sham acupressure for reducing depression severity in participants with ESKD (Table 3). There is insufficient evidence to form conclusions about the comparison of acupressure to transcutaneous electrical acupoint stimulation (TEAS) for the reduction of depressive-symptom severity, fatigue, or sleep-quality improvement. A fair-quality, three-arm RCT (N=96) (39) compared acupressure with sham acupressure (i.e., pressure applied 1 cm from the acupressure point) and usual care. Participants receiving acupressure reported a significantly greater reduction in depression symptoms than those receiving sham acupressure or usual care (there was no difference between sham and usual care). A second three-arm RCT (poor quality; N=108) (48) compared acupressure with both TEAS and usual care. Participants in both the acupressure and the TEAS groups reported greater reductions in depressive symptoms and fatigue, and better-quality sleep than those who received usual care (Tables 2 and 4).
Intensive Nursing
Three poor-quality RCTs (42,43,55) provide insufficient evidence for the effect of intensive nursing interventions, compared with less-intensive nursing, for improving depressive symptoms in adults with ESKD (Tables 3 and 4). One study in China (N=72) compared home-nursing visits with telephone follow-up for adults on peritoneal dialysis (42). Although both groups’ Zung Self-Rating Depression Scale scores were improved from baseline (P<0.001), the home-nursing group’s depression scores were significantly lower than the control group after intervention (P<0.001). Another Chinese study of a comprehensive nursing intervention compared with usual care (N=128) (43) in adults on hemodialysis found that the intervention group had significantly greater reduction in Self-Rating Depression Scale scores from baseline (P=0.04), and those scores were significantly better after intervention compared with the control group (P=0.05). An Iranian RCT (N=60) using the Depression, Anxiety, and Stress Scale found that telenursing follow-ups compared with usual care (no telephone follow-ups) resulted in significantly lower depression scores (Table 2) (55). These studies did not report on AEs.
Exercise
There was insufficient evidence from two poor-quality studies (34,40) on exercise training for depression in adults with ESKD (Tables 3 and 4). One study (N=50) (34) examined the effect of 4 months of endurance and resistance training four times weekly in Tunisian male participants. Training sessions were on off-dialysis days and the exercise group significantly improved on HADS-D scores compared with sedentary controls. The other, a Greek RCT (N=50) (40), examined intradialytic exercise (during hemodialysis cycling and strength training) three times per week for a year and found significant improvement on both the Beck Depression Inventory-II and HADS-D with exercise compared with sedentary controls (Table 2). AEs were not reported.
Other Treatments
We included RCTs of six other therapies: Benson relaxation technique (37), guided imagery (29), hope therapy (45), Latihan Pasrah Diri (51), mindfulness-based stress reduction (47), and Quran readings for Muslim participants (28). All were small, single trials with methodologic challenges, and the evidence is insufficient for all interventions (Tables 2⇑–4). Two of these studies (Latihan Pasrah Diri [51] and mindfulness-based stress reduction [47]) reported no AEs, whereas the others did not report on AEs.
Discussion
In this systematic review, we examined nine RCTs of pharmacologic interventions and 17 of nonpharmacologic interventions for comorbid depression in adults with ESKD. We also found three observational studies that contributed to the evidence on harms of SSRIs. We found moderate-strength evidence that long-term, high-dose vitamin D3 is ineffective for reducing depressive symptoms, and low-strength evidence that both sertraline and CBT were similar for improving depressive symptoms. We also found low-strength evidence that CBT is more effective than both standard psychotherapy and wait-list controls, and that acupressure is more effective than usual care. We found insufficient evidence to draw conclusions about all other interventions. Overall, we found limited evidence for each intervention, sample sizes were small, and nearly all studies were hampered by methodologic flaws.
SSRIs, compared either with placebo or an active comparator, were the best-studied pharmacologic intervention. No other antidepressant categories were identified. Findings from placebo-controlled trials of SSRIs were mixed, but sertraline at least warrants further study. Among the three trials examining sertraline, one reported benefit over placebo, and another found no difference between sertraline and CBT (44). Harms related to SSRIs were not uniformly reported. The type and rate of harms reported and/or evaluated in included RCTs suggest little to no increase in risk for adults with ESKD compared with otherwise healthy adults using SSRIs. Primary side effects were minor (e.g., nausea, fatigue). However, observational studies suggest that clinicians should consider known risks (e.g., QT prolongation) when making prescribing decisions, and to coordinate care to avoid over- or under-treatment.
Vitamin D3 is an interesting intervention for adults with ESKD, due to the associated risks of hypercalcemia and hyperphosphatemia (60). Five adults withdrew from the study due to treatment-related AEs. Although not attributed to hyperphosphatemia, the reported AEs (i.e., joint pain, diarrhea, nausea, and vomiting) may be related. The negative finding regarding vitamin D3, along with its risks, suggests that clinicians should not recommend its use as a depression treatment in this population. Given the large size and length of the trial, the SOE for this finding is moderate, and future studies are unlikely to change conclusions.
Very few nonpharmacologic studies reported harms; however, most interventions presented minimal risk. Differences by subpopulation (i.e., demographic, clinical) were also reported in very few studies, and reported differences were insufficient to form conclusions. Future research should uniformly report harms and examine these subgroups.
Many of the studies were hampered by small sample sizes, posing challenges related to group comparability and statistical power. The duration of treatment and follow-up varied widely across studies, making it more challenging to compare results. For instance, between two studies of sertraline, one was twice as long as the other (3 versus 6 months). The methods used to screen for and diagnose depression were heterogeneous, as was the implementation of interventions (e.g., timing, doses, comparators, modes of delivery). For example, among CBT studies, although all of them met in person, half of the studies used private sessions, and sessions were conducted in groups in the other studies. Session lengths varied and, in one study, CBT was delivered while the participants were receiving hemodialysis. Given the association with hemodialysis and somatic complaints, hemodynamic changes, and alterations in mental acuity, treatments (particularly psychologic ones) that are administered during hemodialysis may be confronted with these challenges; although treating adults outside of hemodialysis sessions has its own challenges, including the increased time burden and physical and mental fatigue. Among other psychologic treatments, some were applied in person, whereas others were practices expected to be followed at home. In addition to these issues, the lack of methodologic detail reported in many of the studies resulted in poor-quality ratings and uncertainty about study processes.
Another important limitation to the current evidence base is that most of the studies were conducted outside of the United States and examined participants and health systems that differ greatly from the general US population. These differences may be reflected in both the patient demographics and medical disease burden, and in how care is delivered, particularly because the majority of US adults on dialysis receive their care from large dialysis organizations in coordination with their nephrology provider.
This is the only systematic review to date that examines both pharmacologic and nonpharmacologic treatment of depression in adults with ESKD. This review confirms and adds to a 2016 Cochrane review of antidepressants in adults with ESKD, which included meta-analyses of harms reported in trials included in our report (59). Although we also included more recent trials, outcomes were not reported in a way that allowed for a quantitative synthesis of harms. Our review adds to the pharmacologic evidence by including studies of dietary supplements. A Cochrane review examining psychosocial interventions for adults with ESKD was recently published (61); however, it includes studies of participants with and without clinical depression. How a participant with ESKD responds to an intervention may vary widely depending on the severity of their baseline depressive symptoms. To reduce clinical heterogeneity, we included only studies with participants meeting established depression-screening thresholds for chronically ill populations (19⇓⇓⇓–23). Finally, additional trials included in our review, particularly the ASCEND (44) and ASSertID trials (33,44), add to both the pharmacologic and nonpharmacologic evidence.
Future research, particularly in the United States, is needed to better evaluate both pharmacologic and nonpharmacologic interventions for this population. In particular, larger replication studies of CBT, acupressure, and SSRIs (such as sertraline), and examination of their effects in different subgroups of adults with ESKD and depression (such as severity of depression, duration of ESKD diagnosis, peritoneal versus hemodialysis, and presence of comorbidities), will help decision makers to implement depression treatments that are not only evidence based, but are also the best fit for their patient population.
Disclosures
All authors have nothing to disclose.
Funding
This research was funded by the US Department of Veterans Affairs, VHA, Office of Research and Development, Health Services Research and Development.
Supplemental Material
This article contains supplemental material online at http://kidney360.asnjournals.org/lookup/suppl/doi:10.34067/KID.0003142020/-/DCSupplemental.
Supplemental Appendix 1. Search strategies (parent VA review).
Supplemental Appendix 2. Study selection criteria (parent VA review).
Supplemental Table 1. PICOTS by key question.
Supplemental Table 2. Strength of evidence domains and grading.
Acknowledgments
The authors wish to thank Ms. Robin Paynter (MLIS) for developing the search strategy and running electronic searches. We also thank the members of our technical expert panel for their expertise.
The findings and conclusions in this document are those of the authors who are responsible for its contents, the findings and conclusions do not necessarily represent the views of the Department of Veterans Affairs or the US Government.
Each author contributed important intellectual content during manuscript drafting or revision, accepts personal accountability for the author’s own contributions, and agrees to ensure that questions pertaining to the accuracy or integrity of any portion of the work are appropriately investigated and resolved.
Author Contributions
J.R. Antick, D. Kansagara, and K. Kondo provided supervision; C.K. Ayers was responsible for project administration; C.K. Ayers, P. Chopra, D. Kansagara, and K. Kondo reviewed and edited the manuscript; C.K. Ayers, P. Chopra, and K. Kondo were responsible for investigation; C.K. Ayers and K. Kondo were responsible for data curation and methodology; and all authors conceptualized the study, were responsible for formal analysis, and wrote the original draft.
- Received May 19, 2020.
- Accepted January 4, 2021.
- Copyright © 2021 by the American Society of Nephrology