Does Native Vitamin D Supplementation Have Pleiotropic Effects in Patients with End-Stage Kidney Disease? A Systematic Review of Randomized Trials

Vitamin D has been shown to have multiple pleiotropic effects beyond bone and mineral metabolism, with purported roles in cardiovascular disease, cancer, and host immunity. Vitamin D deficiency is common in patients with end-stage kidney disease (ESKD); however, current clinical practice has favored the use of the active hormone. Whether vitamin D deficiency should be corrected in patients with ESKD remains unclear, as few randomized trials have been conducted. In this systematic review, we summarize the current evidence examining whether vitamin D supplementation improves outcomes, beyond mineral metabolism, in patients with ESKD. Data from randomized controlled trials of adults with ESKD were obtained by searching Ovid MEDLINE, Embase, the Cochrane Central Register of Controlled Trials, and the Web of Science Core Collection from inception to February 2023. Twenty-three trials composed of 2489 participants were identified for inclusion. Data were synthesized by two independent reviewers and summarized in tables organized by outcome. Outcomes included measures of mortality, cardiovascular disease, inflammation, muscle strength/function, nutrition, patient well-being, and outcomes specific to ESKD including erythropoietin usage, pruritus, and dialysis access maturation. The Cochrane risk of Bias Tool (RoB 2, 2019) was used to assess study quality. Overall, our findings indicate a minimal and varied benefit of native vitamin D supplementation. From the largest studies included, we determine that vitamin D has no demonstrated effect on patient-reported measures of well-being or utilization of erythropoietin, nor does it change levels of the inflammation biomarker C-reactive protein. Included trials were heterogeneous with regards to outcomes, and the majority studied small participant populations with a relatively short follow-up. We conclude that vitamin D supplementation corrects vitamin D deficiency and is safe and well-tolerated in humans with ESKD. However, it is not clear from clinical trials conducted to date that a causal pathway exists between 25(OH)D and pleiotropic effects that is responsive to vitamin D treatment.


Introduction
Vitamin D may have actions beyond the expected effects on mineral homeostasis and bone disease. Local production of 1.25(OH) 2 D in cells of the immune system, gastrointestinal system, breast, and skin has expanded the understanding of the vitamin D endocrine system to include potential pleiotropic effects that might benefit cardiovascular (CV) disease, host immunity, glucose homeostasis, cancer, and inflammation [1,2]. of treatment, length of follow-up, number of randomized versus analyzed, and number lost to follow-up. Summary data regarding the age of participants and the outcome measures were obtained. No formal assessment of agreement between raters was obtained.

Data Synthesis and Analysis
The data were synthesized in a narrative form and then sub-grouped based on themes of outcomes. These themes included mortality, CV disease (surrogate markers (imaging and laboratory) and clinical events), inflammation (C-reactive protein (CRP)), metabolic outcomes (lipids and albumin), musculoskeletal outcomes (grip strength, bone mineral density (BMD), and fracture), patient well-being outcomes (health-related quality of life (HRQOL) and depression), and outcomes specific to dialysis/ESKD (e.g., anemia management, pruritis, and arteriovenous fistula maturation). Study descriptions and clinical characteristics are summarized by summary tables and text. Overall, we sought to summarize the direction of any observed effects of vitamin D treatment across studies. Based on substantial heterogeneity between studies, meta-analysis was not performed.

Assessment of Risk of Bias
The risk of bias for each study was assessed using the Cochrane Risk of Bias tool version 2 [14] by two review authors (NP and ON). A third reviewer (RH) resolved disagreements. We used the following six criteria to determine whether the risk was low, high, or had some concerns: randomization process, effect of assignment to intervention, effect of adherence to intervention, missing outcome data, measurement of outcome, and selection of reported results.

Measurement of Treatment Effect
Many trials reported continuous measures at baseline and study end and determined the treatment effect using tests of statistical significance. Some trials reported event rates at study end (e.g., mortality and CV events). For ease of reporting, we categorized trials into tables that encompassed the themes described above. Some (n = 11) trials are reported in more than one table if the reported outcomes in the study encompassed more than one theme (e.g., Wang et al. reported CRP and psychological health) [15]. Measures of statistical significance of treatment effect were reported directly from the respective studies.

Safety
The trials were examined for reported events of hypercalcemia and vitamin D toxicity.

Study Selection
The results of the search are presented in Figure 1. Of the 144 studies included for full-text review, 23 studies were conducted in patients with ESKD and included in this review.

Study Characteristics
All 23 studies selected for inclusion in this review were randomized trials published in English. One trial was excluded as the body of text and abstract were not in English [16]. Three trials were excluded due to multiple expressions of concern published by journals [17,18]. One study was excluded as data were presented by sub-categories within treated and control groups, based on baseline 25(OH)D level, and the number of patients per stratification was not reported [19]. Two studies that were included were only available in abstract form [20,21].

Study Characteristics
All 23 studies selected for inclusion in this review were randomized trials published in English. One trial was excluded as the body of text and abstract were not in English [16]. Three trials were excluded due to multiple expressions of concern published by journals [17,18]. One study was excluded as data were presented by sub-categories within treated and control groups, based on baseline 25(OH)D level, and the number of patients per stratification was not reported [19]. Two studies that were included were only available in abstract form [20,21].
The 23 studies included in this review are presented in Table 2. The studies are presented in alphabetical order by the surname of the first author. Overall, 2489 participants were included in these studies (1281 randomized to vitamin D and 120mt8 randomized to The 23 studies included in this review are presented in Table 2. The studies are presented in alphabetical order by the surname of the first author. Overall, 2489 participants were included in these studies (1281 randomized to vitamin D and 120mt8 randomized to control). Nineteen trials were randomized, and placebo-controlled, whereas the control arm was standard of care in four trials [20][21][22][23]. The geographical regions where the trials were conducted included Europe (n = 7), North America (n = 6), Asia (n = 2), Gulf States (n = 4), South America (n = 2), and Australia (n = 2). The majority of trials were conducted in a single center, whereas six trials included more than one center, and in four trials it was not specifically reported. Overall, the sample size studied was fewer than 100 participants in 19 trials and was 60 or fewer in 12 of these. In 14 trials, participants were randomized only if vitamin D insufficiency was present, whereas baseline vitamin D status was not considered for eligibility in 9 trials. Elevated PTH and pruritis were required for inclusion in two studies [24,25]. The vitamin D intervention included cholecalciferol (n = 20), ergocal-ciferol (n = 2), and calcifediol (n = 1). In trials of cholecalciferol, the most common dosing regimen was 50,000 units weekly, with some studies employing a step-down to a lower weekly dose once a sufficient 25(OH)D level was obtained. One three-week study provided 200,000 IU weekly [24], and the study of Khajehdehi reportedly provided 50,000 units daily for 3 months [26]. The two studies conducted in the United States using ergocalciferol employed similar dosing regimens (50,000 units weekly with a step-down to monthly) [27,28]. The study of calcifediol provided 40 mc thrice weekly [22]. Follow-up ranged from 3 weeks to 24 months. The primary outcome of the trial is presented in Table 1, and the timing of the various outcome measures is presented at baseline and study end. In two trials, the duration of follow-up for ascertainment of clinical outcomes extended beyond the treatment duration [28,29]. We report only on measured outcomes and/or clinical events not directly related to laboratory measures of CKD-MBD. A component of laboratory assessment of CKD-MBD or 25(OH)D levels was the primary outcome of the trial in seven cases [21,[23][24][25]27,30,31].

Risk of Bias Assessment
In Table 3, we present the quality measures of the studies. There was a low risk of bias across all parameters in the studies by Bhan, Brimble, Hewitt, Miskulin, Seibert, and Singer. Otherwise, bias due to the effect of assignment to intervention and adhering to intervention was deemed to be high if not adequately reported. A number of trials were deemed to have concerns or be at high risk of bias based on missing outcomes if drop-out from the study was unacceptably high or, as in several cases, not specifically reported upon. Several studies were deemed to be at high risk of bias across several of the assessed domains.

Vitamin D Supplementation and Mortality Outcomes
In the four trials that reported mortality outcomes, the treatment duration ranged from 4 to 24 months; however, follow-up in two of the trials extended beyond the treatment period (Table 4) [22,27,29,40]. The formulation of vitamin D included ergocalciferol (n = 1), the pre-hormone calcifidiol (n = 1), and cholecalciferol (n = 2). Mortality was only pre-specified within the composite primary outcome in the study where supplementation was provided by the pre-hormone, calcifediol, in an open-label fashion [22]. Although this phase-III multicenter study was the largest trial, and included 284 patients receiving HD, the study did not reach the projected sample size to address the primary outcome due to funding issues. Over 24 months, supplementation with calcifediol did not prevent mortality compared with standard care [22]. However, a large proportion of study patients did not achieve vitamin D sufficiency (36% in the treatment group versus 11% in the placebo group). Bhan et al. conducted a three-arm, 12-week trial in incident patients receiving HD that compared weekly ergocalciferol to monthly administration and placebo [27]. A large separation between the groups in terms of 25(OH)D levels was achieved. The mortality outcome was assessed at 12 months where a trend towards a benefit of monthly ergocal-ciferol supplementation on mortality (p = 0.08) was observed as well as a non-significant trend favoring the combination of weekly and monthly ergocalciferol arms compared to placebo (HR 0.28; 95% CI, 0.07 to 1.19, p = 0.07) [27]. Brimble et al. found a significant effect of cholecalciferol treatment on all-cause mortality (12% in the vitamin D group versus 39% in the placebo group (p = 0.004)) and death from CV cause (3% in the vitamin D group versus 19% in the placebo group, p = 0.03) in patients receiving PD [29]. At the end of this trial, 96% of participants were replete (defined as 25(OH)D level ≥ 50 nmol/L) compared to only 14% of the placebo group indicating successful correction of vitamin D deficiency [29]. However, the number of events was small and the length of follow-up for the mortality outcome extended beyond the 1-year treatment with vitamin D. The trial by Hewitt et al. included 68 vitamin D insufficient patients with ESKD on either HD or PD and was designed to determine the impact of 12 months of cholecalciferol supplementation on hand grip strength and quality of life. Only one participant died during the follow-up and, overall, adverse events were similar between groups [33]. No difference between the groups in rates of hospitalization was reported in any of the trials.  CV: cardiovascular; HR: hazard ratio; IRR: incidence rate ratio; ns: not significant. * p value represents the significance level of the reported treatment effect between groups; 1 treatment was 4 months and follow-up was 12 months; 2 treatment was 3 months and follow-up was >1 year.

Vitamin D Supplementation and Cardiovascular Disease-Related Outcomes
Of the 23 studies included in this review, 4 trials evaluated the effect of vitamin D therapy on surrogate measures of CV disease including left ventricular mass and function, abdominal aortic calcification (AAC), and pulse wave velocity (PWV) [29,31,33,35]. Four trials reported clinical CV events [22,[27][28][29]. Two trials evaluated the response of brain natriuretic peptide (BNP), a biomarker of congestive heart failure, to vitamin D treatment [35,38]. These results are summarized in Table 5. All trials were placebo-controlled with the exception of one open-label study where standard therapy was the control arm [22]. Follow-up ranged from six months [28] to two years [22]. Overall, no treatment difference was observed in left ventricular mass or function, PWV or AAC [29,31,33,35]. In fact, PWV and AAC increased over time in both vitamin D-treated and untreated control participants [31,33,35]. Vitamin D treatment did not modify 24 h blood pressure, nor did it change levels of BNP [35,38]. There were significantly fewer deaths related to CV disease in the vitamin D-treated patients in the study by Brimble et al.; however, this was an exploratory analysis where the overall event rate was small and follow-up extended beyond the treatment period [29]. In the largest trial that included 284 patients and 24 months of follow-up, no difference was observed in CV death, nonfatal MI, nonfatal CVA, fatal MI, or fatal stroke [22]. The second largest study of 276 participants did not report any difference in the rate of hospitalization for CV disease over the 6-month treatment period [28].

Vitamin D Supplementation and Inflammation
Eight trials reported changes in C-reactive protein (CRP), a clinically used biomarker of inflammation (Table 6) [15,28,30,[32][33][34][35]37]. One study demonstrated a significant betweengroup difference in CRP favoring vitamin D treatment [34]. However, this was a small 12-week study that randomized 55 patients but only included 38 participants in the final analysis. The much larger study by Wang et al. randomized 746 patients on hemodialysis with elevated depressive scores to 50,000 IU/week of cholecalciferol versus placebo and reported no difference in hs-CRP over the duration of the 12-month trial [15]. The second largest trial conducted by Miskulin et al. found a significant within-group decrease in hs-CRP in the group treated with ergocalciferol (p = 0.02), but no difference was observed in the change between the two groups [28]. The remaining trials randomized small numbers of patients and found no difference in CRP with vitamin D treatment [30,32,33,35,37].

Vitamin D Supplementation and Musculoskeletal Outcomes
Two trials reported outcomes related to muscle strength [33,40], and one trial evaluated changes in serum levels of testosterone [41]. In two studies, the change in grip strength was the primary outcome of the study. Neither study demonstrated any treatment benefit over 6 months and 12 months, respectively. Hewitt also assessed functional capacity and timed walking with no differences found between treatment arms [33]. Baseline 25(OH)D levels correlated with the distance walked in 6 min, but not muscle strength; however, no changes were detected over the 6-month supplementation period. One trial was designed to determine whether cholecalciferol treatment increases serum testosterone levels in patients with ESKD. Despite normalization of serum 25(OH)D levels, testosterone levels did not change [41]. Further, there was no correlation between testosterone and 25(OH)D levels at baseline or at the end of the study, suggesting that testosterone levels were independent from vitamin D status. Results were similar in males and females. No difference was reported in any study reporting fracture outcomes or falls. One small trial, published as an abstract only, reported a benefit in BMD preservation in favor of vitamin D [21]. In Zheng et al., BMD increased in the vitamin D and placebo groups; however, both arms received co-treatment with a vitamin D analog and a calcimimetic [25]. The results of these trials are summarized in Table 7. CRP: C-reactive protein; Hs-CRP: high-sensitivity C-reactive protein; NR: not reported; ns: not significant. * p value represents the significance level of the reported treatment effect between groups; ** p value represents treatment effect within groups from baseline (p < 0.05 in ergocalciferol patients at 6 months compared to baseline).

Vitamin D Supplementation and Anemia, Pruritis, and Arteriovenous Fistula Maturation
One trial was designed, and thus powered, to determine whether vitamin D supplementation modified erythropoetin (EPO) dosage [28]. The overall results of this trial were negative. Two smaller trials, including fewer than 100 participants in each, also reported a change in EPO dose. The study conducted by Mehorotra et al., reported in abstract form only, reported a significant decrease in EPO dosage in the treatment group [20]. Naini et al. reported no significant effect of treatment on EPO dose [36]. One trial measured change in pruritus symptoms over 12 weeks. Pruritus severity was measured using a survey and corresponding score. In both the treatment and placebo groups, there was a decrease in severity of itch over the time course of the study and there was no significant effect of treatment [39]. Wasse et al. measured the patency success of created hemodialysis access (arteriovenous fistula (AVF) or arteriovenous graft (AVG)) in patients randomized to either vitamin D or matching placebo [24]. After 6 months, the percentage of patients with successful cannulation of their AVF or AVG was not significantly different between the two groups [24]. These results are summarized in Table 8.

Vitamin D and Metabolic/Nutritional Measures
Five trials included a measure of nutritional or metabolic health as an outcome. The results of these outcomes are summarized in Supplemental Table S1. Serum albumin was measured in five of the trials. In each of these trials [15,29,30,32,38] there was no significant change in albumin between placebo or treatment groups, although there was a trend towards higher levels of albumin and prealbumin in the vitamin D-treated group in the largest study [15].     1 Abstract only; 2 follow-up was 6 months following AVF; 3 data presented graphically only. * p value represents the significance level of the treatment effect between groups; ** p value represents the significance of the treatment effect within groups from baseline.
Khajehdehi et al. also measured lipids in patients randomized to either vitamin D, vitamin C, or vitamin E with each group compared to placebo [26]. They found that, compared to the placebo group, participants in the vitamin D group had significantly lower triglycerides and a lower triglyceride to HDL-c ratio at study end. There were no differences at study exit from baseline in either the vitamin D or placebo group in LDL-c/HDL-c ratio, cholesterol/HDL-c ratio, cholesterol and LDL-c, and HDL-c [26].

Vitamin D Supplementation and Well-Being
Ayub et al. used a participant-reported survey to evaluate chronic pain. They found a significant reduction in reported pain in both the treatment and placebo groups [30]. Hewwit et al. used the Kidney Disease Quality of Life (KDQOL)-36 to assess kidneydisease-related quality of life (Supplemental Table S2) [33]. Baseline 25(OH)D levels did not correlate with HRQOL at baseline, and there was no difference between treated participants and controls after 6 months [33]. Singer et al. used the KDQOL-Short Form (KDQOL-SF) to assess quality of life related to kidney-failure-specific symptoms [40]. After 12 months of cholecalciferol treatment, there was no difference in the KDQOL-SF scores between the treated and placebo groups. Re-analysis restricted only to those with more severe baseline vitamin D deficiency (25(OH)D < 27.5 nmol/L) did not change these results. Further, no differences were observed in any of the KDQOL-SF domains [40]. Only one trial measured psychological health as an outcome. Wang et al. used the Chinese version of Beck's Depressive Inventory II (BDI-II) as a measure of depressive symptoms in hemodialysis patients [15]. The BDI-II evaluates 21 self-reported items on a scale of 0-3, giving a total possible score of 63. Scores over 11 reflect the presence of depressive symptoms [42]. Participants in the study required a score of 16 or higher to be eligible. At baseline, there were no significant differences between the vitamin D and placebo groups. After 12 months, BDI-II scores were significantly lower compared to baseline in both the vitamin D and placebo groups, and there was no significant difference in the mean change in score between the two groups [15]. In a sub-analysis, participants were stratified by type of depression. In participants with a diagnosis of vascular depression at baseline, there was a significantly larger mean decrease in BDI-II scores (−5.5 ± 0.6 vs. −1.4 ± 0.3, p = 0.047) in patients treated with vitamin D (n = 150) compared to patients taking the placebo (n = 160) [15]. In participants with diagnosed major depressive disorders (MDDs) at baseline, there was no significant difference in mean change in BDI-II score by treatment group.

Vitamin D and Changes in 25(OH)D, Hypercalcemia, and Vitamin D Toxicity
All studies but one [26] included in this review measured the response of serum 25(OH)D levels to vitamin D treatment (Supplemental Table S3). Overall, a significant treatment effect on the increase in 25(OH)D was observed. In only one trial, a substantial proportion of study participants did not reach sufficient vitamin D levels at study end [22]. We determined the frequency of hypercalcemia and vitamin D toxicity (25(OH)D > 250 nmol/L). Vitamin D toxicity was only specifically addressed in three trials, and a vitamin D level > 250 nmol/L was reported in three patients (one treated patient and two controls). Hypercalcemia was specifically mentioned in 15 trials but occurred infrequently.

Discussion
To date, this is the largest systematic review examining native vitamin D therapy in patients with ESKD requiring dialysis that focuses on the potential pleiotropic benefits of vitamin D beyond those related to mineral homeostasis. Our search criteria started in 1947, yet 20 of these trials were published within the past 10 years, indicating a renewed interest in native vitamin D in this patient population. Overall trends in the data show limited effectiveness of vitamin D therapy on a variety of clinical outcomes. Previous systematic reviews confirmed that treatment with native vitamin D corrects vitamin D deficiency in patients with ESKD [43,44]. The results of this review suggest that repletion of vitamin D does not appear to parallel subsequent changes in outcomes beyond measures of vitamin D sufficiency. However, the majority of trials (19 of 23) randomized 100 or fewer participants, indicating that, overall, the studies were quite small with limited power to address the outcomes. Further, many trials were of low-to-moderate quality. From the two largest, and unbiased, trials, we find no evidence that vitamin D replacement decreases erythropoietin usage or improves depressive symptoms or levels of CRP in patients with ESKD [15,28].
In a meta-analysis of 50 randomized trials that included almost 75,000 participants without CKD, vitamin D supplementation was not associated with a reduced risk of all-cause mortality risk [45]. Whether these results can be applied to a population with ESKD, where vitamin D metabolism is dysregulated and vitamin D deficiency is frequently encountered, is unknown. In a meta-analysis of observational studies, vitamin D treatment was associated with survival in patients with earlier stages of CKD [46]. The small sample sizes of the randomized studies included in our review would suggest that none of these trials were powered adequately to detect an impact of vitamin D supplementation on overall survival, should one exist. The nephrology community awaits the results of the simplified trial. This large, pragmatic trial of over 4000 dialysis patients in the United Kingdom is a prospective, randomized, open-label blinded-endpoint superiority trial comparing cholecalciferol versus standard care in patients on dialysis [47]. The primary outcome for the simplified trial is patient survival.
CV disease is much more common in patients with ESKD than in the general population; however, this prevalence is not explained by traditional CV risk factors [48]. Other factors contribute, including vascular calcification related to dysregulated mineral metabolism and inflammation [48]. Vitamin D receptors are expressed widely, and CYP27B1 has been found in cardiac tissue, vascular smooth muscle cells, and endothelium [10]. Stimulation of vitamin D receptors in the myocardium can prevent cardiac hypertrophy and decrease secretion of BNP, a biomarker used clinically to evaluate congestive heart fail-ure [49]. In the general population, vitamin D supplementation does not appear to reduce CV events [50,51]. However, the majority of these trials were primary prevention trials in populations where vitamin D deficiency is uncommon, and the CV event rate was typically a secondary outcome. The duration of follow-up may not have been sufficient to capture events related to a chronic disease in healthier people. In our review, a surrogate outcome of CV disease was the primary outcome in four trials. In prospective and non-randomized studies of patients receiving dialysis, vitamin D treatment decreased left ventricular mass and reduced levels of BNP [52,53]. However, the two randomized and controlled studies that included a measurement of left ventricular mass showed no difference between vitamin D treatment versus placebo [29,35]. Similarly, we found no treatment effect of vitamin D on BNP, PWV, abdominal aorta calcification, or 24 h BP [31,33,35,38]. Taken together, there is no available evidence from these relatively small trials to suggest that vitamin D alters CV structure and/or function in patients receiving dialysis.
Clinical CV event rates were assessed in four studies. In the largest trial of 276 patients receiving HD that compared 6 months of ergocalciferol supplementation to placebo, there was a trend toward reduced hospitalization for a CV cause, favoring those receiving vitamin D [28]. However, estimates would be expected to be imprecise given the sample size, and the study was not powered to detect this outcome. In a sub-group analysis, there was no difference by subgroups of baseline 25(OH)D concentrations. In a much smaller trial of PD patients, there was a significant decrease in death from CV disease in patients treated with vitamin D, but caution is required when interpreting this due to the overall very small number of events and extension of follow-up beyond the treatment period [29].
Patients requiring hemodialysis are chronically inflamed, and its presence is associated with poor outcomes including CV events [54]. In this population, low serum levels of 25(OH)D have been associated with elevated levels of CRP and IL-6 [55]. Vitamin D has been identified as a potential modifier of inflammation where it has been proposed that 1.25(OH) 2 D, synthesized in monocytes, may inhibit the production of pro-inflammatory cytokines [56]. Recent meta-analyses have demonstrated that vitamin D supplementation improved levels of CRP in the general population [57], in patients with diabetes [58], as well as in patients with diabetic kidney disease [59]. In our review, two studies demonstrated a significant difference in CRP levels with supplementation; however, there was significant loss to follow-up in both studies [30,34]. The remaining randomized trials did not show any difference in CRP levels despite substantial improvement in 25(OH)D levels, a longer duration of follow-up (6-12 months), and much larger sample sizes [15,28]. Overall, there is minimal evidence from randomized trials to support a role for vitamin D supplementation in reducing levels of CRP in patients receiving dialysis.
Low serum testosterone levels are associated with mortality in male patients receiving HD, and in early stages of CKD, testosterone levels have been linked to muscle strength [60,61]. Hypogonadism has been linked to vitamin D deficiency in several patient cohorts, but whether vitamin D replacement modifies testosterone levels in patients receiving HD is not known [62,63]. The one trial that evaluated testosterone levels did not show any change over 12 weeks; however, the sample size was very small [41]. Neither trial that evaluated measures of muscle strength demonstrated any benefit of vitamin D [33,40]. Although these trials do not support the use of vitamin D in hemodialysis patients for muscle and strength outcomes, the sample sizes are small, and the duration of follow-up may not be long enough to find a significant effect of treatment, should one exist.
Osteoporosis is a key component of bone disease in this population where fractures occur frequently. Vitamin D supplementation has been shown to decrease the incidence of bone fracture in people above the age of 60 but not in younger individuals [50,64]. The impact of vitamin D on bone mineral density is not proven in the general population but may benefit those individuals above the age of 50 [50,65]. In our review, BMD results were presented in one abstract, one uncontrolled study of 19 participants, and a placebocontrolled study of 60 participants who were also receiving cinacalcet and a vitamin D analog [21,23,25]. No conclusions can be reasonably drawn from these limited data, and specifically designed RCTs would be necessary to confirm or refute the observational data from other elderly patient groups.
HRQOL is significantly reduced in patients with ESKD on HD and has been associated with numerous adverse outcomes including hospitalization and mortality [66]. Low levels of 25(OH)D have been associated with reduced HRQOL in patients receiving dialysis [67]. In the trial conducted by Hewitt et al., baseline HRQOL scores were not associated with 25(OH)D levels, and following six months of vitamin D supplementation, there were no between-group differences in HRQOL domains [33]. In the 12-month trial, supplementation resulted in substantial differences in achieved 25(OH)D levels between groups but did not translate into differences in HRQOL measures [40]. One study of 746 patients randomized hemodialysis patients with depression [15]. Vascular depression describes depressive disorders in later life in patients with either clinical or imaging evidence of CV disease. In the sub-group of patients with vascular depression, there was some benefit in favor of cholecalciferol, and the authors attributed this as being dependent on improvements in unmeasured CV factors [15].
A number of trials evaluated the impact of vitamin D on outcomes that are unique to the ESKD population. Many patients with ESKD experience pruritis, which can have significant impact on sleep quality and mood [68,69]. Ultraviolet light, which increases the cutaneous production of calcitriol via vitamin D precursors, has been shown to improve symptoms in patients with refractory pruritis [70,71]. One trial sought to determine whether vitamin D might have anti-pruritic effects. However, no significant difference in pruritis severity scores between the placebo and treated groups was noted at any time point during the 12-week study [39], including in those supplemented patients that converted from vitamin D insufficiency to repletion. Both groups experienced a reduction in pruritis scores, suggesting either abatement of itching symptoms over time or regression to the mean [39]. Although increasing 1.25(OH) 2 D production in the skin is an attractive hypothesis to modify cutaneous immunity and down-regulate cutaneous inflammation, this small trial did not support this.
Given a putative role for vitamin D in suppressing neointimal hyperplasia, it was hypothesized that vitamin D supplementation might improve arteriovenous fistula (AVF) outcomes. This pilot study was conducted to determine the feasibility of a trial to determine whether high-dose cholecalciferol, compared to placebo, modified the maturation of AVFs [24]. At follow-up, there was no difference in successful cannulation between cholecalciferol-and placebo-treated groups [24]. Conclusive results might require a longer period of treatment prior to AVF creation, a larger sample size, and a study population at risk for non-maturation.
Anemia has been associated with poor quality of life and increased CV complications and mortality in patients with ESKD [72]. The majority of patients receive erythropoietin for anemia correction; however, a fraction of these patients are hypo-responsive and require large doses. Vitamin D-mediated suppression of inflammatory cytokines is one postulated mechanism by which vitamin D may promote erythropoeisis [73]. One randomized trial, adequately powered to evaluate whether 6 months of vitamin D supplementation reduced erythropoietin dose requirements in patients receiving HD, was negative [28]. The authors also conducted a sensitivity analysis that did not demonstrate any effect on erythropoetin dose in participants who achieved vitamin D sufficiency. Two much smaller trials suggested that administration of vitamin D might reduce erythropoietin doses; however, these trials studied far fewer participants over a shorter duration and performed poorly on several measures of bias [20,36]. The study by Miskulin et al. was adequately powered to address the primary outcome of epoetin usage; therefore, we determine that vitamin D supplementation does not play a significant role in the management of anemia in ESKD [28].
The results of this systematic review confirm that nutritional vitamin D insufficiency and deficiency can be safely corrected in patients with ESKD. It is therefore unlikely that the mostly negative results were on the basis of not achieving adequate vitamin D levels or that different results would have been obtained if higher doses were used. Overall, hypercalcemia was infrequent and typically not higher in patients randomized to vitamin D. Only three studies [22,31,40] specifically addressed the upper limit of vitamin D, and in these studies only one patient exceeded the threshold. Although the results do not indicate a vitamin D-treatment benefit on a variety of pleiotropic outcomes, it is unlikely that this was based on unsuccessful correction of vitamin D deficiency. Further, in those studies that conducted a sensitivity analysis based on adequate correction of vitamin D, the negative conclusions remained unchanged. Limitations of this systematic review relate to overall study quality and the small sample sizes with limited power to address the outcomes. Due to the underlying heterogeneity in terms of interventions and captured outcomes, the data are not amenable to meta-analysis. From the majority of trials we can exclude large effects of vitamin D treatment on the various outcomes; however, smaller effect sizes may still be possible. This review is the first systematic review in the literature that examines randomized trials of vitamin D treatment and pleiotropic outcomes in patients receiving dialysis where a high prevalence of vitamin D deficiency exists. There is a lack of adequately powered randomized trial evidence to support a beneficial role for vitamin D in outcomes relating to its potential pleiotropic effects. At present, it is not clear from clinical trials conducted in humans with ESKD that a causal pathway exists between 25(OH)D and clinical outcomes that operates through correction of vitamin D deficiency and is responsive to vitamin D treatment.