aDefined as a 16-item Quick Inventory of Depressive Symptomatology–Clinician Rated (QIDS-C16) score of less than 11 after the placebo run-in phase.
bIncluded self-reported tremor, hypersomnolence, insomnia, diarrhea, and headache.
cDefined as having taken less than 65% of drug by pill count.
dEstimated glomerular filtration rate greater than 60 mL/min/1.73 m2 in 3 patients, ongoing substance dependence in 1 patient, suicidal ideation in 1 patient, and unable to obtain blood for laboratory tests in 1 patient.
eExited the trial prior to the first QIDS-C16 outcomes assessment after randomization (before week 2) and lacked any primary outcome data for analysis. Of the 5 patients in the sertraline group, 1 withdrew consent, 2 withdrew due to adverse effects, and 2 did not return for visits; of the 3 patients in the placebo group, 2 withdrew consent and 1 was hospitalized and no longer wanted to participate.
Participants completed at least 1 assessment after randomization and were included in the primary analysis. Error bars indicate SDs, which were calculated separately for each time point. Each of the 16 QIDS-C16 items can yield a score of 0 to 3 on a Likert scale. The score range is 0 to 27; higher scores indicate more severe depression; a score of 0 to 5 corresponds to a normal affect; 6 to 10 to a mild affect; 11 to 15 to a moderate affect; 16 to 20 to a severe affect; and 21 or greater to very severe depression.
Parallel line plot of the QIDS-C16 scores. Each patient is represented by a vertical line, with participants plotted by group assignment (sertraline, n = 97; placebo, n = 96) and sorted by baseline value.
Data analysis plan
eFigure. Randomized double-blinded placebo-controlled trial of sertraline treatment of depression in patients with chronic kidney disease
eTable. Treatment efficacy endpoints at exit
Customize your JAMA Network experience by selecting one or more topics from the list below.
Hedayati SS, Gregg LP, Carmody T, et al. Effect of Sertraline on Depressive Symptoms in Patients With Chronic Kidney Disease Without Dialysis Dependence: The CAST Randomized Clinical Trial. JAMA. 2017;318(19):1876–1890. doi:10.1001/jama.2017.17131
Does treatment with sertraline improve depressive symptoms in patients with stage 3, 4, or 5 non–dialysis-dependent chronic kidney disease and major depressive disorder?
In this randomized clinical trial that included 201 patients with non–dialysis-dependent chronic kidney disease and at least moderate depressive symptoms, the use of sertraline vs placebo did not result in a statistically significant difference in symptom improvement over 12 weeks (−4.1 points vs −4.2 points, respectively, of 27 possible points on the 16-item Quick Inventory of Depression Symptomatology–Clinician Rated).
Sertraline may not be an effective treatment for major depressive disorder in patients with non–dialysis-dependent chronic kidney disease.
Major depressive disorder (MDD) is prevalent among patients with chronic kidney disease (CKD) and is associated with morbidity and mortality. The efficacy and adverse events of selective serotonin reuptake inhibitors in these patients are unknown.
To determine whether treatment with sertraline improves depressive symptoms in patients with CKD and MDD.
Design, Setting, and Participants
The Chronic Kidney Disease Antidepressant Sertraline Trial (CAST) was a randomized, double-blind, placebo-controlled trial involving 201 patients with stage 3, 4, or 5 non–dialysis-dependent CKD, who were enrolled at 3 US medical centers. The Mini Neuropsychiatric Interview was used to establish MDD. The first participant was randomized in March 2010 and the last clinic visit occurred in November 2016.
After a 1-week placebo run-in, participants were randomized to sertraline (n = 102) for 12 weeks at an initial dose of 50 mg/d (escalated to a maximum dose of 200 mg/d based on tolerability and response) or matching placebo (n = 99).
Main Outcomes and Measures
The primary outcome was improvement in depressive symptom severity from baseline to 12 weeks determined by the 16-item Quick Inventory of Depression Symptomatology–Clinician Rated (QIDS-C16) (score range, 0-27; minimal clinically important difference, 2 points). Secondary outcomes included improvement in quality of life (Kidney Disease Quality of Life Survey–Short Form; score range, 0-100; higher scores indicate more favorable quality of life) and adverse events.
There were 201 patients (mean [SD] age, 58.2 [13.2] years; 27% female) randomized. The primary analysis included 193 patients who had at least 1 outcome assessment after randomization. The mean (SD) baseline QIDS-C16 score was 14.0 (2.4) in the sertraline group (n = 97) and 14.1 (2.4) in the placebo group (n = 96). The median participation time was 12.0 weeks and the median achieved dose was 150 mg/d, which was not significantly different between the groups. The QIDS-C16 score changed by −4.1 in the sertraline group and by −4.2 in the placebo group (between-group difference, 0.1 [95% CI, −1.1 to 1.3]; P = .82). There was no significant between-group difference in change in patient-reported overall health on the Kidney Disease Quality of Life Survey (median score, 0 in the sertraline group vs 0 in the placebo group; between-group difference, 0 [95% CI, −10.0 to 0]; P = .61). Nausea or vomiting occurred more frequently in the sertraline vs placebo group (22.7% vs 10.4%, respectively; between-group difference, 12.3% [95% CI, 1.9% to 22.6%], P = .03), as well as diarrhea (13.4% vs 3.1%; between-group difference, 10.3% [95% CI, 2.7% to 17.9%], P = .02).
Conclusions and Relevance
Among patients with non–dialysis-dependent CKD and MDD, treatment with sertraline compared with placebo for 12 weeks did not significantly improve depressive symptoms. These findings do not support the use of sertraline to treat MDD in patients with non–dialysis-dependent CKD.
clinicaltrials.gov Identifier: NCT00946998
Quiz Ref IDChronic kidney disease (CKD) affected 15% of the US population in 2016,1 and in 2013, depression was reported to be present in up to 25% of patients with CKD, which is 4 times higher than the prevalence in the general population.2,3 Comorbid depression was independently associated with hospitalization, cardiovascular events, and death among patients with non–dialysis-dependent CKD and among those dependent on dialysis.4-9 Depression also significantly worsened patient-centered outcomes and was associated with dialysis initiation, lower quality of life, greater symptom burden, sexual dysfunction, and nonadherence to medications and dietary restrictions.10,11
Evidence for placebo-controlled efficacy of commonly used antidepressants, such as selective serotonin reuptake inhibitors (SSRIs), is limited among patients with CKD, who generally have been excluded from large randomized trials of antidepressant medication treatment due to safety concerns.12-14 There is a need to establish whether antidepressant medications are safe and efficacious in patients with CKD who are at a disproportionately higher risk for both depression and its complications.
The purpose of this study was to determine whether treatment with sertraline would improve depressive symptoms in patients with non–dialysis-dependent stage 3, 4, or 5 CKD and major depressive disorder (MDD).
The Chronic Kidney Disease Antidepressant Sertraline Trial (CAST) was a double-blind, placebo-controlled, parallel-design, 12-week flexible-dose randomized clinical trial conducted at 3 large medical centers in Dallas, Texas (University of Texas Southwestern Medical Center, Parkland Hospital, and the Veterans Affairs North Texas Health Care System). Detailed methods were published.15 The trial protocol appears in Supplement 1. The data analysis plan appears in Supplement 2.
The study was approved by the institutional review boards at each site and conducted per good clinical practice guidelines and the principles of the Declaration of Helsinki. Written informed consent was obtained from every participant prior to enrollment. The study was monitored by an independent data and safety monitoring board.
Individuals with stage 3, 4, or 5 non–dialysis-dependent CKD presenting to primary care or CKD clinics were screened with the 16-item Quick Inventory of Depressive Symptomatology–Self-Reported (QIDS-SR16) and deemed eligible if they had a score of 11 or greater (score range, 0-27). This instrument was previously validated for depression screening and shown to be independently associated with outcomes among individuals with CKD.16,17
Individuals who provided consent completed the Mini International Neuropsychiatric Interview,18 which is based on the Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition) (DSM-IV), to confirm current unipolar MDD and exclude other psychiatric conditions (ie, bipolar disorder or psychosis). Quiz Ref IDParticipants then completed a 1-week single-blind placebo run-in to exclude those who were nonadherent and ensure each individual still met the eligibility criterion of having a 16-item Quick Inventory of Depression Symptomatology–Clinician Rated (QIDS-C16) score of 11 or greater. Those who successfully completed the run-in (score of ≥11 on the QIDS-C16 and took ≥65% of study drug by pill count) were randomized by concealed allocation to 50 mg/d of either sertraline or matching placebo.
After randomization, participants attended clinic visits every 2 weeks for 6 weeks, then every 3 weeks for the remaining 6 weeks during the double-blind phase of the study (eFigure in Supplement 3). Dose was escalated by 50 mg/d at each successive visit to a potential maximum dose of 200 mg/d based on tolerability and response using a previously described measurement-based care protocol.12,13,15,19 The dose was kept constant for the last 6 weeks of the study. Adherence was ascertained by using pill counts at each visit, with nonadherence defined as having taken less than 80% of the drug. After 12 weeks, the drug was tapered at a rate of 50 mg/wk until completely discontinued. Participants were reassessed for depression 2 weeks after discontinuation and offered open-label sertraline or other therapy, which was managed by primary care or mental health physicians.
Quiz Ref IDAdults with non–dialysis-dependent CKD stage 3, 4, or 5 (estimated glomerular filtration rate <60 mL/min/1.73 m2) were invited to participate. The inclusion criteria were changed to an estimated glomerular filtration rate of less than 45 mL/min/1.73 m2 early during the study in response to the National Institutes of Diabetes and Digestive and Kidney Diseases recommendation to reduce sample heterogeneity. There was no lower eligibility threshold for estimated glomerular filtration rate. Patients were excluded for the following reasons: inability to provide consent; having a functioning kidney transplant or requiring maintenance dialysis; having a transaminase elevation 3 or greater times the upper limit of normal; receiving current treatment with an antidepressant or other serotonergic drugs, or previous sertraline treatment failure; receiving psychotherapy for depression within 3 months; having psychosis or bipolar disorder, dementia, or suicidal intent; or being pregnant. Race and ethnicity were recorded by self-report based on fixed categories to describe the diversity of participants.
Patients were randomized in a 1:1 ratio using block randomization (block size range, 4-8) based on a computerized random-number generator, stratified by hospital site ( the University of Texas Southwestern Medical Center and Parkland Hospital or  the Veterans Affairs North Texas Health Care System) and CKD stage (3, 4, or 5). Treatment assignments were conducted by research pharmacists at each site using the prespecified random allocation sequence. Participants, researchers, and clinicians were blind to treatment assignment. The QIDS-C16 outcome measure was administered by trained personnel who were blind to treatment assignments and measurement-based care algorithms.
The prespecified primary outcome was improvement in depressive symptom severity from baseline to 12 weeks determined by the QIDS-C16 at baseline and weeks 2, 4, 6, 9, and 12. The QIDS-C16 assesses the 9 DSM-IV criterion symptoms of MDD, with higher scores (range, 0-27) indicating more severe depression.20
Prespecified secondary outcomes included (1) response (≥50% decrease in baseline QIDS-C16 score) and remission (score decrease to ≤5); (2) quality-of-life end points and overall functioning assessed by the 5-item Work and Social Adjustment Scale (WSAS; in which each item is rated on a 0-8 Likert scale with 0 indicating no impairment; 8, severe impairment; score range, 0-40)21 and the Kidney Disease Quality of Life Survey–Short Form (KDQOL-SF version 1.3; score range, 0-100; higher scores signify more favorable quality of life),22 which were administered at baseline and at 6 and 12 weeks; (3) adverse events and tolerability assessed by the occurrence of serious adverse events (ie, death, dialysis initiation, hospitalizations, or bleeding requiring transfusion), maximal global adverse effects reported on the 56-item Systemic Assessment for Treatment Emergent Effects (SAFTEE; rated on a 0-4 ordinal scale with 0 indicating none; 4, marked adverse effects),23 and maximal frequency, intensity, and burden of adverse effects assessed with the Frequency, Intensity, and Burden of Side Effects Rating scale (each item rated on a Likert scale as present 0%, 10%, 25%, 50%, 75%, 90%, or all the time).24 Adverse events and tolerability were assessed at every visit.15
Data collected during the pilot phase showed that individuals with CKD and MDD had a mean (SD) decrease in QIDS-SR16 score of 5.6 (4.1) points after treatment with an SSRI.15 A mean antidepressant drug-placebo difference of 3 points or greater on the Hamilton Rating Scale of Depression is recommended as a criterion for clinical significance when establishing guidelines for the treatment of depression.25 This is equivalent to a difference of 2 points on the QIDS-C16, which was regarded as the smallest clinically meaningful difference between the sertraline and placebo groups.20 Assuming a similar SD of 4 points yielded an effect size of 0.5, a t test comparison using a 2-sided α of .05 was estimated to be able to detect an effect size of 0.5 with 80% power in a sample of 128 participants (64 per group). A sample size of 200 (100 per group) was proposed so that if the SD was unexpectedly larger at 5 points, resulting in an effect size of 0.4, there would still be 80% power to detect a between-group difference.
Using the modified intention-to-treat principle, participants were analyzed in the groups to which they were randomized if at least 1 QIDS-C16 assessment after randomization was available. The QIDS-C16 score was compared between groups using a repeated-measures mixed-effects model, with treatment group as the between-participants factor and random intercept and slopes as within-participant factors. Baseline QIDS-C16 was included as a covariate. The interactions between hospital site and CKD stage (3 vs 4 or 5) by treatment group and time were tested but not retained unless significant.
Response and remission were tested using the same procedure for a mixed-effects model, which is appropriate for a repeated binary outcome. A mixed-effects model for continuous outcomes as described for the primary analysis was used to compare changes in the WSAS and the KDQOL-SF, with baseline scores included as covariates. Given that (1) the amount of missing primary outcome data was small, (2) mixed-effects models allow inclusion of participants with missing data, and (3) these models are unbiased when data are missing at random, imputation of missing data was not done.26
The proportion experiencing an adverse event; the maximum SAFTEE global assessment of adverse effects; and the maximal frequency, intensity, and burden of adverse effects at any visit were compared between groups using χ2 tests. A binary outcome (any adverse effect at any visit vs none) was created for each participant and compared between groups using the χ2 test. The mean number of worsening SAFTEE symptoms was compared between groups using a t test. Statistical tests were 2-sided and P < .05 was considered significant. The statistical analyses were performed using SAS version 9.4 (SAS Institute Inc).
There were 201 patients (mean [SD] age, 58.2 [13.2] years; 27% female) randomized. The first participant was randomized in March 2010 and the last clinic visit occurred in November 2016. Of the 997 patients who qualified based on having a QIDS-SR16 score of 11 or greater and meeting other eligibility criteria, 697 did not provide consent. Of the 300 who provided consent, 261 met the Mini International Neuropsychiatric Interview MDD criteria and entered the placebo run-in. Sixty were excluded, leaving 201 to be randomized (Figure 1). Eight patients exited the study prior to the first QIDS-C16 assessment at week 2 and, therefore, had no outcome assessments after baseline and were excluded from the primary analysis. Therefore, 193 patients (97 in the sertraline group and 96 in the placebo group) constituted the modified intention-to-treat sample.
Of the 193 participants, 11% had stage 3A CKD; 36%, stage 3B; 36%, stage 4; and 17%, stage 5. The proportion in each CKD stage was not significantly different between treatment groups. Baseline characteristics were mostly balanced between groups, except that a larger proportion were receiving treatment with erythropoiesis-stimulating agents in the placebo group (7% vs 0% in the sertraline group; Table 1).
The mean (SD) baseline score on the QIDS-C16 (minimum score, 0; maximum, 27) was 14.0 (2.4) in the sertraline group and 14.1 (2.4) in the placebo group. There were no clinically meaningful between-group differences in the proportion with comorbid psychiatric illnesses, in the Mini-Mental State Examination scores, or in the quality-of-life scores (Table 1). Six participants in the sertraline group had a history of substance dependence compared with 1 participant in the placebo group.
The median treatment duration was 84.0 days (interquartile range, 83.0 to 87.0 days; between-group difference, 0 days [95% CI, 0 to 2.0 days], P = .19; Table 2). Altogether, 92% completed at least 6 weeks and 84% completed all 12 weeks of the study. The number of postrandomization visits and the proportion that completed 4, 6, 9, or 12 weeks of the study were not significantly different between the groups (Table 2). The median achieved sertraline dose was 150 mg/d (interquartile range, 100 to 150 mg/d; between-group difference, 0 mg/d [95% CI, −50.0 to 0 mg/d], P = .10). The mean percentage of drug taken ascertained by pill count was 94% in the sertraline group and 96% in the placebo group (between-group difference, −1.6% [95% CI, −4.7% to 1.4%], P = .29). Three participants in the sertraline group and 2 in the placebo group were nonadherent (P = .66).
The mixed-effects model showed no significant treatment group main effect (P = .82) or interaction with time (P = .71) in the QIDS-C16 scores (Figure 2). Quiz Ref IDThe mean change from baseline to study exit in the QIDS-C16 score was −4.1 in the sertraline group and −4.2 in the placebo group (between-group difference, 0.1 [95% CI, −1.1 to 1.3], P = .82; Table 3). Figure 3 illustrates changes in the QIDS-C16 score from baseline to study exit for individual participants.
The binary mixed-effects model for remission showed no treatment group main effect (P = .57) or interaction with time (P = .58). The proportion with remission was 15.5% in the sertraline group and 14.6% in the placebo group (between-group difference, 0.9% [95% CI, −9.2% to 11.0%], P = .86; Table 3). The mixed-effects model for response showed no treatment group main effect (P = .54) or interaction with time (P = .97). The proportion of participants with a treatment response was 32.0% in the sertraline group and 25.0% in the placebo group (between-group difference, 7.0% [95% CI, −5.7% to 19.6%], P = .28; Table 3).
The mixed-effects model for the WSAS showed no significant treatment group main effect (P = .64) or interaction with time (P = .33). The mean change from baseline to study exit in overall functioning by WSAS score was not significantly different (between-group mean difference, −1.8 [95% CI, −5.4 to 1.8], P = .32; Table 3).
There was no significant difference in change in patient-reported overall health on the KDQOL-SF (median, 0 vs 0; between-group median difference, 0 [95% CI, −10.0 to 0]; P = .61). The mixed-effects model for the physical component of the KDQOL-SF also found no significant treatment group main effect (P = .55) or interaction with time (P = .98). No differences were observed in quality-of-life components from baseline to study exit, except for improvement in sleep, which was significantly greater in those receiving sertraline (between-group median difference, 5.0 [95% CI, 0 to 12.5], P = .03; Table 3). The 12-week or study exit efficacy and quality-of-life measures for each treatment group are included in the eTable in Supplement 3.
The groups did not significantly differ in the proportion that experienced 1 or more prespecified serious adverse events (Table 4). One participant receiving placebo developed acute suicidal intent; however, there were no suicides. Another participant receiving 200 mg/d of sertraline was hospitalized with elevated liver enzyme levels. Liver biopsy indicated drug-induced acute liver injury, and sertraline was discontinued. Between randomization and the first assessment of outcomes, 2 participants assigned to sertraline withdrew due to adverse effects (Figure 1).
Of the 193 participants included in the modified intention-to-treat analysis, 3 (3.1%) assigned to sertraline and 3 (3.1%) to placebo discontinued treatment due to drug intolerance (between-group difference, 0% [95% CI, −4.9% to 4.9%], P > .99) (Table 4). Although there were no statistically significant differences for worsening maximal adverse effect frequency, intensity, or burden, the proportion with any vs no adverse effects was higher in the sertraline group (76.1%) compared with the placebo group (60.0%; between-group difference, 16.1% [95% CI, 2.9% to 29.2%], P = .02). Quiz Ref IDNausea or vomiting was reported by 22.7% of the sertraline group compared with 10.4% of the placebo group (between-group difference, 12.3% [95% CI, 1.9% to 22.6%], P = .03) and diarrhea by 13.4% vs 3.1%, respectively (between-group difference, 10.3% [95% CI, 2.7% to 17.9%], P = .02; Table 4).
Treatment with sertraline did not improve depressive symptoms or quality of life in patients with stage 3, 4, or 5 non–dialysis-dependent CKD and resulted in increased adverse effects compared with placebo. To our knowledge, this is the largest randomized, double-blind, placebo-controlled trial to provide evidence about MDD treatment with a commonly used SSRI in patients with non–dialysis-dependent CKD, a chronically ill population that is not only at significantly increased risk for developing depression,2,3 but also its serious complications.4-9
Although the risk of serious adverse events was not higher in patients receiving sertraline vs placebo, those treated with sertraline did experience a significantly higher incidence of gastrointestinal adverse effects. These results provide evidence regarding the lack of efficacy of sertraline among patients with non–dialysis-dependent CKD that would change clinical practice.
Previous studies investigating the efficacy of antidepressant medications among patients with CKD were performed in dialysis-dependent samples and were limited by lack of randomization and control, small sample sizes, short treatment durations, low adherence, high dropout rates, and nonstandard criteria for MDD diagnosis.27-34 Only 2 trials appropriately diagnosed MDD by using a psychiatric interview, one of which included only 14 patients treated over 8 weeks32 and the other, 30 patients over 6 months.27 Neither showed an antidepressant treatment effect. In one of these trials,27 there was a higher withdrawal rate in the sertraline group due to adverse events. However, all participants in the present study underwent a structured psychiatric evaluation to rule in MDD and received an adequate dose and duration of treatment to elicit a clinically meaningful effect.
The retention rate in the current study was much higher than in other antidepressant trials involving patients with CKD,27-34 comparable with large trials among participants without CKD.12-14 There was a high percentage (92%) of patients who completed 6 weeks and 84% who completed 12 weeks (the entire trial duration). In addition, the run-in period successfully minimized nonadherence.
A high placebo response has been postulated to be associated with negative trials of antidepressant medications.35 However, the 4-point decrease in the QIDS-C16 score in the placebo group is lower than what is considered a “high” placebo response rate observed in depression treatment trials. A meta-analysis of 169 randomized placebo-controlled trials of antidepressant monotherapy concluded that a placebo response rate of less than 30% provided the best chance of finding efficacy in drug-placebo comparisons.36 Thus, a “high” placebo response rate is an unlikely explanation for a lack of sertraline efficacy in this study. If sertraline did have any effect, it was significantly smaller than postulated in the sample size calculations and clinically irrelevant. In addition, use of a measurement-based care design to bring each participant to the highest individually tolerable dose resulted in adequate dosing (median, 150 mg/d) compared with lower achieved SSRI doses in other large trials among patients with chronic diseases that did not show SSRI treatment efficacy compared with placebo. The mean (SD) sertraline doses were 65 (29) mg/d and 68.8 (40.1) mg/d in trials of patients with congestive heart failure36 and ischemic heart disease,12 respectively, and the mean (SD) escitalopram dose was 15.8 (6.4) mg/d in a more recent large congestive heart failure trial.14 Therefore, inadequate dosing would not explain the lack of effect, and it is unlikely that sertraline is effective for depression treatment in this patient population.
Another reason for lack of efficacy could be that depression comorbid with chronic disease is a different clinical entity than psychiatric MDD among those without comorbidity and depression that is not responsive to SSRI medications. The current study included participants with medical multimorbidity in addition to non–dialysis-dependent CKD (58% with diabetes, 27% with metabolic syndrome, and 32% with congestive heart failure), which makes it challenging to differentiate whether lack of response to sertraline was due to the presence of CKD or other chronic illnesses. These findings are in line with increasing evidence from well-powered trials among patients with other chronic medical conditions, such as asthma, ischemic heart disease, and congestive heart failure, that found SSRIs were no more efficacious than placebo for treating depression.12,14,36,37 Another potential reason for lack of efficacy is that 84% of participants were unemployed, and socioeconomic factors such as unemployment were shown to be associated with a low likelihood of antidepressant treatment response.38
In addition to investigating the efficacy of a commonly prescribed SSRI among patients with CKD, this study assessed adverse events and tolerability. Few data exist regarding SSRI safety among patients with CKD because such patients have been excluded from large antidepressant medication trials.12,13,39 Patients with advanced CKD are at a higher risk for antidepressant adverse events due to the potential accumulation of toxic metabolite effects in the setting of reduced glomerular filtration rate; increased risk of drug-drug interactions; cardiac sequelae (ie, QTc-interval prolongation and orthostatic hypotension); and increased bleeding from SSRI-induced decreased platelet aggregability, which is particularly problematic among patients with advanced CKD.
Sertraline was chosen because it is primarily metabolized by the liver. Its active metabolite, N-desmethylsertraline, is further metabolized to an inactive form before renal excretion.15 In addition, the Sertraline Antidepressant Heart Attack Randomized Trial found sertraline to be safe among patients with cardiovascular disease.12 Although sertraline did not result in increased serious adverse events or bleeding, it did increase the rate of adverse gastrointestinal symptoms. These types of adverse events are particularly undesirable among patients with advanced CKD because they are already prone to uremic symptoms such as nausea and vomiting.
This study has several limitations. Although the duration was longer than previous studies, it did not assess the long-term effects of antidepressant treatment in CKD. The duration was designed to minimize potential adverse events while allowing adequate drug exposure for a meaningful response. The largest trial of MDD treatment in the general population also assessed efficacy over 12 weeks and showed that a substantial portion of patients who achieved response or remission did so after 8 weeks of treatment with an SSRI.13 In addition, trials with longer durations (6 and 18 months) among patients with other chronic diseases did not reveal efficacy of SSRI vs placebo.14,36 The potential benefit of SSRIs for the treatment of more severe depression may also deserve exploration. Although patients with dialysis-dependent CKD were not included, there is a substantially larger population of patients with non–dialysis-dependent CKD, and the majority of patients with CKD die of cardiovascular disease before reaching end-stage kidney disease. Identifying interventions that could affect outcomes earlier during the disease duration is imperative.
The efficacy and safety of interventions for the treatment of depression among dialysis-dependent patients, a population in which the burden of symptoms, demands of treatment, and propensity for adverse drug reactions are substantially higher than among patients with non–dialysis-dependent CKD, deserve further investigation. Although the current study involved a single geographic area, the demographic breakdown was racially and ethnically diverse, regionally representative, and consistent with published data from other large nationally representative CKD cohorts.40 The current study did include a larger proportion of black participants than previously reported in some CKD cohorts. However, given that the ratio of adjusted incidence rates for CKD in 2014 for blacks vs whites was 3.1, inclusion of a large proportion of blacks is clinically relevant.1 Twenty-two percent of study participants had a history of drug abuse; however, it is uncertain whether this factor is generalizable. The inclusion of non-Veteran patients from the University of Texas Southwestern Medical Center and Parkland Hospital increases the generalizability to females with CKD.1 Future studies need to explore whether non-SSRI medications or nonpharmacological therapies can result in improvement in depressive symptoms among patients with non–dialysis-dependent CKD, who are at a disproportionately higher risk for depression and its complications than the general population.
Among patients with non–dialysis-dependent CKD and MDD, treatment with sertraline compared with placebo for 12 weeks did not significantly improve depressive symptoms. These findings do not support the use of sertraline to treat MDD in patients with non–dialysis-dependent CKD.
Corresponding Author: S. Susan Hedayati, MD, MHSc, University of Texas Southwestern Medical Center, MC 8516, 5939 Harry Hines Blvd, Dallas, TX 75390 (firstname.lastname@example.org).
Accepted for Publication: October 16, 2017.
Published Online: November 3, 2017. doi:10.1001/jama.2017.17131
Author Contributions: Drs Hedayati and Carmody had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Hedayati, Rush, Toto, Trivedi.
Acquisition, analysis, or interpretation of data: Hedayati, Gregg, Carmody, Jain, Toups, Trivedi.
Drafting of the manuscript: Hedayati, Gregg, Carmody.
Critical revision of the manuscript for important intellectual content: Hedayati, Gregg, Jain, Toups, Rush, Toto, Trivedi.
Statistical analysis: Carmody.
Obtained funding: Hedayati.
Administrative, technical, or material support: Gregg, Rush, Toto.
Supervision: Hedayati, Rush, Toto, Trivedi.
Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Rush reported receiving consulting fees from Brain Resource Ltd, Curbstone Consultant LLC, Eli Lilly, Emmes Corp, MindLinc, Montana State University, University of Texas Southwestern Medical Center, Sunovion, and Texas Tech University; speaking fees from Stanford University, Global Medical Education Inc, Sing-Health, University of Montana, Montana State University, and John Peter Smith Health Network; and royalties from Guilford Press and the University of Texas Southwestern Medical Center. Dr Toto reported receiving consulting fees from Amgen, AstraZeneca, Bayer, Boehringer-Ingelheim, Novo Nordisk, Reata, Relypsa, and ZS Pharma. Dr Trivedi reported serving as an advisor or consultant to Abbott Laboratories, Akzo (Organon Pharmaceuticals), Allergan Sales LLC, Alkermes, Arcadia Pharmaceuticals Inc, AstraZeneca, Axon Advisors, Brintellix, Bristol-Myers Squibb, Cephalon Inc, Cerecor, Eli Lilly & Co, Evotec, Fabre Kramer Pharmaceuticals Inc, Forest Pharmaceuticals, GlaxoSmithKline, Global Medical Education Inc, Health Research Associates, Janssen, Johnson & Johnson, Lundbeck, MedAvante, Medscape, Medtronic, Merck, Mitsubishi Tanabe Pharma Development America Inc, MSI Methylation Sciences Inc, Nestle Health Science-PamLab Inc, Naurex, Neuronetics, One Carbon Therapeutics Ltd, Otsuka Pharmaceuticals, Parke-Davis Pharmaceuticals, Pfizer, PgxHealth, Phoenix Marketing Solutions, Rexahn Pharmaceuticals, Ridge Diagnostics, Roche Products Ltd, Sepracor, SHIRE Development, Sierra, SK Life and Science, Sunovion, Takeda, Tal Medical/Puretech Venture, Targacept, Transcept, VantagePoint, Vivus, and Wyeth-Ayerst Laboratories. No other disclosures were reported.
Funding/Support: This work was funded by grant R01DK085512 from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) and Department of Veterans Affairs MERIT grant CX000217-01 (awarded to Dr Hedayati). Support was also provided by the University of Texas Southwestern (UTSW) Medical Center O’Brien Kidney Research Core Center via NIDDK grant P30DK079328, grant UL1TR001105 from the Center for Translational Medicine, National Center for Advancing Translational Sciences of the National Institutes of Health (NIH), the Center for Depression Research and Clinical Care at UTSW, and grant K23MH104768 from the National Institute of Mental Health (awarded to Dr Toups).
Role of the Funder/Sponsor: Neither the NIDDK, the NIH, the Department of Veterans Affairs, or the National Institute of Mental Health had a role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIDDK, the NIH, the Department of Veterans Affairs, or the National Institute of Mental Health.
Meeting Presentation: Presented in part at the American Society of Nephrology’s Kidney Week; November 3, 2017; New Orleans, Louisiana.
Additional Contributions: We are grateful to all of the patients who consented to be screened and who participated in the CAST study. We also thank the following individuals who did not receive compensation for their role in the study: members of the data and safety monitoring board at the UTSW site (Beverley Adams-Huet, MS, Sherwood Brown, MD, Kevin C. Kelly, PharmD, and Robert F. Reilly, MD); the Central VA central data monitoring committee (Clinical Science Research and Development, Hines Cooperative Studies Program Coordinating Center); psychiatry consultants (Benji Kurian, MD, UTSW, and Collin Vas, MBBS, Veterans Affairs North Texas Health Care System); a nephrology faculty member (Peter Van Buren, MD, MSc, UTSW); a biostatistician (Charles South, PhD); and UTSW nephrology residents for contributions in recruiting CAST participants (Masoud Afshar, MD, Lei Chen, MD, Michael Concepcion, MD, Vishal Jaikaransingh, MD, Naseem Sunnoqrot, MD, and Venkata Yalamanchili, MBBS). We also thank the following individuals who were compensated for their role in the study: Anuoluwapo Adelodun, MBBS, MPH, Patricia Alvarez, MSW, Mieshia Beamon, MS, Susamei Khamphong, BA, Ammar Nassri, MD, Michael Phan, PharmD, MBA, MHSM, Staci Schwartz, BA, Francisco Sanchez, BS, and Kyle West, MS (all research personnel with UTSW); David Rezaei, PharmD (Veterans Affairs North Texas Health Care System); and the UTSW Nephrology Clinical and Translational Research Center.
Create a personal account or sign in to: