Abstract
Background and objectives Up to 50% of patients undergoing hemodialysis suffer from symptoms of depression and/or anxiety. Access to traditional pharmacotherapies and psychotherapies for depression or anxiety in this patient population has been inadequate. The objective of this study was to investigate the feasibility and effectiveness of brief mindfulness meditation intervention for patients on hemodialysis with depression and anxiety symptoms.
Design, setting, participants, & measurements This study was a randomized, controlled, assessor-blinded trial conducted in an urban hemodialysis unit. Forty-one patients were randomly assigned to intervention (n=21) and treatment-as-usual (n=20) groups. The intervention group received an 8-week individual chairside meditation intervention lasting 10–15 minutes, three times a week during hemodialysis. Feasibility outcomes were primarily assessed: enrollment rates, intervention completion rates, and intervention tolerability. Symptoms of depression and anxiety were measured using the Patient Health Questionnaire (PHQ-9) and the General Anxiety Disorder-7 (GAD-7).
Results Of those deemed eligible for the study, 67% enrolled (41 of 61). Of the participants randomized to the intervention group, 71% completed the study, with meditation being well tolerated (median rating of 8 of 10 in a Likert scale; interquartile range=10–5 of 10). Barriers to intervention delivery included frequent hemodialysis shift changes, interruptions by staff or alarms, space constraints, fluctuating participant medical status, and participant fatigue. Meditation was associated with subjective benefits but no statistically significant effect on depression scores (change in PHQ-9, −3.0±3.9 in the intervention group versus −2.0±4.7 in controls; P=0.45) or anxiety scores (change in GAD-7, −0.9±4.6 versus −0.8±4.8; P=0.91).
Conclusions On the basis of the results of this study, mindfulness meditation appears to be feasible and well tolerated in patients on hemodialysis with anxiety and depression symptoms. The study did not reveal significant effects of the interventions on depression and anxiety scores.
Podcast This article contains a podcast at https://www.asn-online.org/media/podcast/CJASN/2017_10_12_CJASNPodcast_17_12_.mp3
- Meditation
- Mindfulness
- Anxiety
- depression
- dialysis
- Depressive Disorder
- Anxiety Disorders
- renal dialysis
- Surveys and Questionnaires
Introduction
Depression and anxiety symptoms affect approximately 50% of patients undergoing maintenance hemodialysis (1–3) and are associated with lower quality of life (3–9), increased hemodialysis nonadherence (10), suicidal behavior (11), medical comorbidity, and mortality (3–9). Despite the high burden posed by depression and anxiety (12), distress management has been inadequate in patients undergoing hemodialysis (13–15) due to low rates of detection (13–15), small scale studies of treatments with variable outcomes (16,17), and limited availability of psychologic interventions that are proven to be effective (13). Therefore, alternative, effective, and affordable strategies must be urgently explored to improve quality of life of the patients and their caregivers (3). Mindfulness meditation interventions, which allow patients to enter into “moment-to-moment nonjudgmental awareness” (18), have gained increasing popularity in recent years (19). Mindfulness meditation has been highly effective in several psychiatric conditions (19–21) and effective in chronic physical illnesses (21–30). For patients on hemodialysis specifically, one study demonstrated modest positive benefits for anxiety of caregiver-delivered Benson’s relaxation, a transcendental meditation technique. However, in this study, patients with major medical comorbidities, physical disabilities, and psychiatric histories were excluded, rendering their results hard to generalize to the typical hemodialysis population and to patients with significant depressive/anxiety symptoms (31). Additionally, analyses in that trial were not performed on an intention-to-treat basis and the study did not examine mindfulness meditation techniques, which are increasingly being found to be effective in depression and anxiety in medically-ill populations. Therefore, a methodologically sound, scalable, and generalizable clinical trial of mindfulness meditation for depression and anxiety symptoms in patients on hemodialysis is timely (32,33).
The aim of this study was to test the feasibility and the effectiveness of a brief chairside mindfulness meditation intervention administered individually three times a week for 8 weeks during hemodialysis at a large urban dialysis unit. The primary goal was to determine whether the intervention could feasibly be performed in this population: whether enrollment rates, intervention completion rates, and intervention tolerability were reasonable. A secondary goal was to examine whether the intervention reduced depression and anxiety symptoms on well validated scales.
Materials and Methods
Patients on maintenance hemodialysis were recruited from the Jewish General Hospital hemodialysis unit, in Montreal, Canada, from March to July of 2016. Patients were included if they spoke English or French and had depression and/or anxiety symptoms as indicated by scores of ≥6 on the Patient Health Questionnaire (PHQ-9) and/or General Anxiety Disorder-7 (GAD-7) scales. Although summed scores of ≥10 are considered to represent moderate-to-severe depression and anxiety symptoms, and have been validated in patients on hemodialysis (34), scores ≥6 are felt to capture clinically significant subthreshold patients, because scores >5 are associated with lower quality of life, more disability days, and more primary care visits (35,36). Patients were excluded if they had significant cognitive impairment (determined by an abnormal score on the Mini-Cog (37)), current psychosis, or acute suicidal ideation with intent. Patients in both the intervention and control groups were asked to avoid changes in psychotropic pharmacotherapy or additional psychotherapy for the duration of the trial. The study was approved by the Jewish General Hospital ethics committee in accordance with the Declaration of Helsinki and was registered in ClinicalTrials.gov on Dec 30, 2015 (NCT02686333).
Randomization
Participants completed questionnaires with an independent assessor who then assigned each of them an anonymized code. The interventionists, who were not involved in the recruitment process and patient assessment, randomized the participant codes to the intervention group or the control group, using a simple 1:1 computer-generated sequence. Both the assessor and the statistical associate were blinded to randomization allocation.
Intervention
The intervention consisted of individually administered, guided, chairside meditative practices three times a week, during hemodialysis sessions, lasting 10–15 minutes. Four meditation techniques drawn from mindfulness-based cognitive therapy (body scan, guided meditation, silent meditation, and gentle arm movements) (38) were practiced in alternating fashion, on the basis of patient preference. In these techniques, the participant is guided to direct their attention toward specific elements of their experience, for example their breathing, their body sensations, their thoughts, or ambient sounds, all the while adopting a nonjudgmental attitude. Before and after each meditation, the interventionists performed a 1–2 minutes check-in with the patient, briefly exploring their experience. Curtains surrounding each patient allowed confidentiality and privacy to be easily preserved during interventions. Patients were encouraged to practice the techniques at home, between sessions, but did not maintain formal logs.
Interventionists
The two interventionists, blinded to patients’ baseline scores, were a fifth-year senior psychiatry resident with a 6-day Mindfulness-Based Practice Certification, and a senior psychologist with Mindfulness-Based Stress Reduction training and 40 years of clinical mental health experience. The interventionists reviewed the techniques together to ensure intervention consistency.
Controls
Patients in the control group received treatment as usual in the hemodialysis setting. Both control and intervention groups received psychoeducational literature on anxiety and depression.
Outcome Variables
The primary outcome variables examining feasibility were (1) the proportion of participants screened as eligible who enrolled, (2) the proportion of participants who completed the 8-week trial in the intervention arm (completed ≥13 sessions and stayed with the intervention until week 8), and (3) the tolerability of the mindfulness meditation intervention on a ten-point Likert scale. At the end of the intervention (8-week follow-up), participants completed a questionnaire in which they rated “how much they enjoyed” each mindfulness practice on a scale of 1–10, estimated their frequency of independent meditative practice over the last week, and reported any adverse events. Interventionists received qualitative subjective comments from intervention participants throughout the study on a spontaneous basis, and asked for overall feedback about the study after each participant’s last meditation session.
The secondary outcome variables exploring effectiveness were changes on the PHQ-9 and/or the GAD-7 during 8-week follow-up. The PHQ-9 is a brief nine-item self-report scale on the basis of the Diagnostic and Statistical Manual of Mental Disorders-IV criteria for major depressive disorder, in which each item is scored from 0 (not at all) to 3 (nearly every day). Construct validity studies have demonstrated its internal consistency (Cronbach’s α=0.89) and test-retest reliability (intraclass correlation=0.87) to be excellent (35). The GAD-7 is a brief seven-item self-report scale on the basis of Diagnostic and Statistical Manual of Mental Disorders-IV criteria for generalized anxiety disorder, with items scored from 0 (not at all) to 3 (nearly every day). Similar to the PHQ-9, it has excellent internal consistency (Cronbach’s α=0.92) and test-retest reliability (intraclass correlation=0.83) (36).
Statistical Methods
All statistical analyses were performed on an intention-to-treat sample by an independent member on the study team who had not directly participated in patient assessment. This included patients randomized to the intervention arm who then stopped the intervention, but accepted assessment at 8-week follow-up. Patients with missing data (e.g., refused assessment at 8-week follow-up) were removed from the analyses. Normality of the data was assessed with the Shapiro–Wilk test. Baseline characteristics were compared between groups to assess for randomization adequacy: chi-squared, t-, and Mann–Whitney U test as appropriate. Primary feasibility outcomes were described using counts, means, and percentages. The secondary effectiveness outcome measures, changes in PHQ-9 depression, and GAD-7 anxiety scores were compared between intervention and control groups using a two-sided independent t test or Mann–Whitney U test, as appropriate. We performed exploratory subgroup analyses to assess whether intervention was more effective in participants with higher baseline depression scores (PHQ-9≥10), with relatively less medical comorbidity (<10 medical problems), and in those already using psychotropic medication for a previously diagnosed mental health issue. Subgroup analysis for baseline depression scores was defined a priori; analyses stratifying for medical comorbidity and psychotropic medication usage were determined post hoc. The sample size recruited for this study was on the basis of Birkett’s theory that n=10 in each arm is sufficient to estimate effect sizes and future sample sizes needed for confirmatory randomized, controlled trials (39). Calculations were performed using SPSS 23 (IBM Corp., Armonk, NY). Effect sizes and 95% confidence intervals for the main effectiveness outcomes were calculated using an SPSS script (http://core.ecu.edu/psyc/wuenschk/SPSS/CI-d-SPSS.pdf). Measures of effect size were performed using Cohen’s d statistic.
Results
One-hundred and seventy-nine patients on hemodialysis were approached, 123 of which agreed to be screened for the study. Sixty-one patients met inclusion criteria, and 41 were enrolled into the trial (67% of those eligible). Of those who did not enter the trial, 15 were not interested, two had prominent cognitive deficits, and three were unable to complete the baseline questionnaires due to fatigue. Twenty-one patients were randomized to the intervention group, and 20 to the control group. Figure 1 displays participant flow over the course of the study.
Participant flow over the course of the study.
Participants’ mean age was 65±13 years, 34% were female, 49% were white, 49% were married, and 50% were living with family. On average, participants had 10±4 medical problems including kidney failure, with the most common comorbidities being hypertension, diabetes, dyslipidemia, coronary artery disease, arrhythmias, and peripheral vascular disease. At the time of the study, 46% of participants were on psychiatric medications, and 24% had active mental health follow-up. Further demographic details can be found in Table 1.
Baseline characteristics of patients on hemodialysis randomized to brief mindfulness meditation intervention versus treatment as usual (n=41)
Fifteen patients in the intervention arm completed ≥13 sessions and remained with the intervention until week 8, resulting in a 71% retention rate. In intervention completers, the mean number of sessions received was 20 out of 24 (82% of all sessions offered). Of missed sessions, 55% were due to logistic issues (switches in the location or time of assigned hemodialysis shifts) and 45% were due to refusals (most common reasons given were “too tired” or “too ill” on the given day). Five patients dropped out early in treatment (<2 sessions) for “feeling too medically ill” (n=1), “feeling already improved” (n=1), and “lack of interest” (n=3). One patient stopped after five sessions when they were transferred to home peritoneal dialysis therapy. Over the course of the study, ten participants were admitted to hospital.
Overall, intervention participants appreciated the mindfulness meditation intervention, as indicated by a median score of 8 of 10 on a Likert scale. Participants most appreciated the guided meditation (median, 8 of 10; interquartile range, 9–6 of 10) and least appreciated the gentle arm movements (median, 6 of 10; interquartile range, 8–0 of 10). On average, intervention participants reported practicing mindfulness meditation on their own an additional 3 days a week for 14 minutes per day. No adverse events were observed.
Table 2 displays secondary effectiveness outcome results. At baseline, PHQ-9 depression and GAD-7 anxiety scores did not differ significantly between groups. We did not observe any significant changes in the secondary outcome measures, change in depression (change in PHQ-9, −3.0±3.9 in the intervention group versus −2.0±4.7 in controls; P=0.45) or anxiety scores (change in GAD-7, −0.9±4.6 versus −0.8±4.8; P=0.91). There were no differences in baseline PHQ-9/GAD-7 scores between study completers and noncompleters (defined as those not having completed the end questionnaires) in the intervention group. For the control group, noncompleters (n=5) had significantly higher baseline PHQ-9, but not GAD-7 scores, compared with completers (n=15): PHQ-9, 16.8±4.9 versus 10.3±5.3; P=0.03; GAD-7, 9.6±5.7 versus 5.3±5.4; P=0.15. In exploratory subgroup analyses, higher baseline depression scores (PHQ-9≥10), and relatively less medical comorbidity (<10 medical problems), had similar negative results.
Effects of brief mindfulness meditation intervention on PHQ-9 depression and GAD-7 anxiety scores in patients on hemodialysis who completed both pretest (baseline) and post-test (8-week) assessments (n=32)
However, in the subgroup of participants using psychotropic medication for a previously diagnosed mental health issue (n=14), the mindfulness intervention was associated with a greater reduction in PHQ-9 scores: −4.3±3.5 versus +0.1±3.5, P=0.03.
Subjective reports made to the interventionists by participants in the mindfulness meditation group highlighted improvements in “courage,” “hope,” “dignity,” “self-confidence,” and “strength to face medical illness.” One participant noted that fatigue postdialysis was reduced from “4 hours to 1 hour.” Yet another participant attributed mindfulness meditation as a source of inspiration to propose marriage to his partner during the trial.
Interventionists noted several system-level barriers to intervention delivery, namely unpredictable and frequent patient shift changes, numerous interruptions by staff or alarms during meditations, and space constraints at chairside. Patient-level barriers were also identified, such as fluctuating medical status with intermittent deteriorations preventing meditation, and excessive participant fatigue.
Discussion
This study is the first to examine brief chairside mindfulness meditation interventions in patients on hemodialysis with depression and anxiety symptoms. We demonstrated that meditation interventions were feasible, enjoyable, and beneficial for patients on hemodialysis despite several systemic barriers. Retention rates were comparable to other psychotherapy trials; a notable outcome, considering participants had several factors associated with attrition, such as older age, comorbidities, and poor physical health (40). Whereas qualitative benefits and improvements were noted by participants, meditation seemed to have little effect on depression and anxiety symptoms assessed by standardized questionnaires; however, our sample was not powered to detect statistically significant differences in these outcome measures. The inclusion of patients with milder depression and anxiety symptoms may have similarly diminished the probability of finding statistically significant results.
There are only a limited number of psychotherapy studies available in patients on hemodialysis—these yielded positive results for cognitive behavioral therapy (CBT) and problem-solving therapy for depression (41–43). The problem-solving study examined a similar clinical population to this study (n=33), using 1-hour weekly sessions for 6 weeks, and obtaining significant improvements on Beck’s Depression Inventory and PHQ-9 scores (43). The two CBT studies that reported positive results had larger sample sizes than this study (n=90 and n=65, respectively), with one using 12 weekly group CBT therapy sessions of 1.5 hours each (42), and the other using 10 weekly therapy sessions of 1-hour individual chairside CBT (41). Patients in these studies seem to have had lower rates of medical comorbidities at baseline and were excluded for prolonged hospitalizations.
The results of these additional studies suggest that cognitive elements may be important therapeutic components to effective psychologic interventions in patients on hemodialysis. In our study, participants demonstrated substantial interest in discussing their psychologic states, such that keeping the conversations surrounding meditations brief and limited proved to be challenging for the interventionists. Furthermore, others have stated that, in more advanced disease, issues related to hope and existential wellbeing may be more prominent (44,45), effects that may not be conveyed in standard questionnaires employed in the previous literature or this study.
Strengths of this study were inclusion of patients on hemodialysis with severe medical illness, rendering findings generalizable, as well as its randomized design and assessor blinding. An additional strength was the chairside mode of delivery, which enhanced patient convenience and provided increased access to the interventions. Limitations of this study included modest sample size, although it was comparable to a previous study of psychotherapy in patients on hemodialysis. Other limitations were the lack of an active control group and the fact we did not assess long-term outcomes (41).
Considering the need for treatments for depression and anxiety in this population, the obstacles encountered when attempting to engage patients on dialysis in additional appointments, and the widespread use of polypharmacy in these patients rendering psychopharmacologic treatments less desirable, an intervention as scalable as mindfulness meditation may be a promising area of study (3,13,15–17,46). Further trials may restrict inclusion criteria to baseline PHQ-9 scores of 10 and above to select for more severely depressed patients. Headphones could be used in future studies to facilitate the scalability of guided meditation at the hemodialysis chairside. A cognitive and group therapy component could also be integrated (38). In order to determine efficacy of such an intervention for improving depressive symptoms, future trials could aim for a sample size of at least n=143 (on the basis of our effect size 0.23, at two-tailed α=0.05, and power 80%; G*Power 3.1.9.2; Universitat Kiel, Germany). On the basis of mindfulness-based studies in similar depressed samples with chronic severe physical illness, including an active comparator group controlling for clinical attention and social support elements of the intervention, a sample size of 175–300 may be needed (20,29). Measurements of fatigue and sleep may be examined, as well as scales examining hope and demoralization (e.g., Herth Hope Index [47]), and qualitative assessments.
In conclusion, we demonstrated that brief meditation interventions were feasible in an in-center hemodialysis setting. Although meditation seemed associated with self-reported subjective benefits, we did not detect statistically significant effects on depression and anxiety symptom scores with our small sample. Future mindfulness research in patients on hemodialysis can better determine effectiveness by recruitment of a larger sample, provision of an active control group, inclusion of only more severely depressed patients, and assessment of qualitative measures.
Disclosures
N.H. has research contracts funded by F. Hoffman-La Roche Ltd., and Lundbeck Canada Inc. All other authors have no disclosures to declare.
Acknowledgments
We are grateful to the dialysis center staff, particularly the nurses, who allowed us to conduct this study. We extend particular thanks to M.L. and Mari Srian, who facilitated our involvement in the dialysis center. We thank our volunteers, Corina D’Alesio, Ian Bernstein, Ghizlane Moussaoui, Roya Taheri, and Marina Khalil, who helped us with recruitment and data entry. We are also grateful to Dr. Vanita Jassal and Dr. Charmaine Lok, who collaborated with project development.
S.R. is supported by the Canadian Institute of Health Research Fellowship Award and Fonds de Recherche Santé Québec Chercheur-Boursier Clinicien Junior Investigator Award. This study was also supported by charitable donations to the Jewish General Hospital Division of Geriatric Psychiatry.
Some data from this paper had been presented at the American Association for Geriatric Psychiatry Annual Scientific Meeting, Dallas, Texas, March 24, 2017, and at the European Association for Psychosomatic Medicine Annual Scientific Conference, Barcelona, Spain, June 30, 2017.
Footnotes
Z.T. and M.N. shared first authorship.
Published online ahead of print. Publication date available at www.cjasn.org.
- Received April 9, 2017.
- Accepted August 21, 2017.
- Copyright © 2017 by the American Society of Nephrology
Podcast