Vitamin D supplementation effects on the clinical outcomes of patients with coronary artery disease: a systematic review and meta-analysis

In this systematic review and meta-analysis our aim was to assess the effect of vitamin D supplementation on cardiac outcomes in patients with coronary artery disease (CAD). The search terms were performed from January 2000 to January 2018, only randomized clinical trials (RCT) in human subjects were considered, with no language restrictions. The electronic databases used in this study were: PubMed; Cochran library; Embase; and Scopus. Two independent expert reviewers carried out data extraction according to Cochrane recommendations. Only four RCTs were found in relation to the effects of vitamin D supplementation on the coronary artery disease. In these 299 patients, vitamin D supplementation had significant favorable effects on Diastolic Blood Pressure (DBP) (− 2.96, p = 0.02) and Parathyroid hormone (PTH) (− 4.05, p < 0.001). However, it had no significant effects on hs-CRP mean difference (− 0.04, p = 0.25), total cholesterol (TC) (− 0.46, p = 0.83), triglyceride (TG) (0.68, p = 0.89), low-density lipoproteins (LDL) (2.08, p = 0.56), and high-density lipoproteins (HDL) (− 2.59, p = 0.16). In conclusion, the use of vitamin D was associated with improvements in some cardiac outcomes of CAD patients with vitamin D deficiency. Also, further research is needed to clarify these results.


Method
Research methods. The Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines (PRISMA) was considered appropriate for use in this systematic and meta-analysis review as it is covering a public health subject that requires transparent reporting 19 . Thus the effect of vitamin D intervention on clinical and biochemical outcomes in patients with CAD were evaluated by findings of randomized controlled trial (RCT) studies according to PRISMA (details are shown in Fig. 1) 19 .
Search strategy. We searched PubMed; Cochran library; Embase; and Scopus databases, studies were selected based on inclusion criteria by conducting a comprehensive search using the standard Mesh terms. Search items included vitamin D, vitamin D3, cholecalciferol, ergocalciferol, and calcitriol combined with coronary artery disease, blood pressure, hypertension, cardiovascular, heart disease, coronary disease, inflammation, inflammatory mediators, lipids, total cholesterol, triglycerides, high density lipoprotein (HDL), low density lipoprotein (LDL), high -sensitivity C reactive protein (hs-CRP), parathyroid hormone (PTH), blood pressure and RCTs. Studies search terms were regularly checked in the stated databases from January 2000 to January 2018 for randomized clinical trials in human subjects, with no language restrictions. Moreover, the reference list of each identified article was reviewed and eligible articles (only those reporting RCTs) were also included.
Study screening and inclusion criteria. By using the PICOS framework (Population, Intervention, Comparison, Outcomes, Study design) 20 , we determined the eligibility of studies and these eligibility criteria are reported in Table 1 Risk of bias and quality assessment. All RCTs were assessed by two independent reviewers according to Cochrane pre-specified criteria 21 . Using this strategy, each RCT was categorized and rated for bias as high, low and unclear risk of bias. The studies which had at least 3 items for risk of bias were categorized as good quality; studies were categorized as fair with two items for risk of bias; and studies with ≤ 1 item for risk of bias were categorized as poor (details are shown in Figs. 2,3,4,5,6,7,8,9). Also, the quality of each item was examined by using the method of Cochrane Collaboration risk of bias tools 21 . Reviewers graded quality score of studies by showing the risk of bias less than two a low quality score and higher than two an appropriate quality. Any disagreements were resolved by the third reviewer (AN) ( Table 1).  www.nature.com/scientificreports/ Statistical analysis. For meta-analysis, collected effect measure after supplementation period were pooled into weight mean difference (WMD) with 95% confidence intervals (CI) 22 . If the variables were non-random in terms of quantity we used them for a fixed model 22 , however if variables heterogeneity (het) existed, the random model was used. When there is heterogeneity that cannot be clarified, one statical approach is to combine it   www.nature.com/scientificreports/ into a random-effects model. This model involves an assumption that the effects being estimated in the different studies are not equal, but accordance some distribution. The center of this distribution explained the mean of the effects, while its width describes the degree of heterogeneity 23 . Heterogeneity was calculated by using the I 2 2 test with weighted Mantele-Haenszel method, in this regard, I 2 2 > 50% shows a notable heterogeneity 24 . According to the Egger and Begg statistical tests and visual symmetry of funnel plots, publication bias was determined 25,26 .

Results
Selection of studies and screening process are explained in PRISMA flow chart-diagram Fig. 1. A total of 1683 titles peer reviewed publications were retrieved; after scanning the titles, 822 were removed due to duplication and 690 were excluded as they lacked relevance. In the next step, 49 studies were eligible for full-text review. Finally, only four RCT studies met the eligibility criteria for pooled analysis as explained in Fig. 1 16- Intervention characteristics. Oral tablet of vitamin D was used in all four RCTs, with dose of 0.5 μg daily in one of the RCTs 27 and weekly dose of 50,000 IU in rest of RCT studies. Of these four RCTs, in two studies participants were supplemented with cholecalciferol 16,17 and in one study participants were supplemented with ergocalciferol 18 and in the other one participants were supplemented with oral calcitriol 27 . All of these studies were lasted for between 8 weeks to 6 months as reported in (Table 2).

Discussion
Summary of findings. The pooled outcome of this study demonstrated that groups with vitamin D administration showed favorable impacts in diastolic blood pressure and parathyroid hormone levels as compared to placebo groups. However, there was not a significant difference between vitamin D and control groups with regards to levels of hs-CRP, total cholesterol, triglyceride, LDL, HDL and SBP. Our findings suggest that vitamin D supplementation may have a modest clinical effects in CAD patients. Previously, human cross-sectional studies, demonstrated a reverse association between serum vitamin D concentrations and inflammation in heart failure patients. In our meta-analysis with four clinical trials in CAD patients, vitamin D supplements did not reduce the circulating hs-CRP concentrations. While, current study findings are contradictory to a previous meta-analysis by Jiang et al. 28 , where they reported lower concentrations of hs-CRP in treatment group compared to placebo group. However, in support of our study, Rodriguez et al. 29 , reported that pooled outcomes from three studies with 231 heart failure patients illustrated that vitamin D supplementation had no effect on CRP concentrations. The possible mechanism associated with Vitamin D in the regulation of lipid profile levels, could be due to the high lipoprotein lipase activity, increase in calcium absorption rate and decrease in fatty acid absorption levels and LDL formation 30,31 . In the current review, we demonstrated that vitamin D supplementation had no significant effect on the LDL levels in CAD patients. In contrast to our results, according to a meta-analysis study by Mirhosseini et al. 32 , vitamin D treatment in obese subjects improved their lipid profile. In addition, Jafari et al. 33 , also presented a significant decline in the levels of total serum cholesterol, TG and LDL in diabetic patients. This inconsistency between other reported results and the results of our study are probably due to the heterogeneity of the population in Jafari et al. 33 , study and the inclusion of healthy individuals in the study by Mirhosseini et al. 32 .
Notably, antihypertensive function of cholecalciferol is proposed through suppression of the renin angiotensin pathway with its anti-endothelial stiffness effect, followed by secondary hyperparathyroidism prevention. Our meta-analysis demonstrated a significant reduction in DBP levels by cholecalciferol supplementation in CAD patients, which was consistent with previous findings in a study conducted by Mirhosseini 32 , while inconsistent with Beveridge et al. 34 , findings. Moreover, in the current meta-analysis the pooled analysis indicated no effect of SBP reduction in these patients. The moderate heterogeneity for weighted SBP results propose that a clinically significant reduction in blood pressure is unlikely, based on the selected dose of vitamin D in this analysis. Moreover, the moderate heterogeneity for weighted systolic blood pressure suggests no effects of vitamin D supplementation on systolic blood pressure. These outcomes are in accordance with several previous metaanalyses [35][36][37] . While, it is important to note that different categories of patients, numbers, dose of vitamin D and duration of interventions were pooled in those studies. Since vitamin D deficiency results in parathyroid gland hyperactivity, the PTH concentrations increases 38 . Sudden increase in PTH concentrations, leads to transport of large amount of calcium into the cardiocytes, where heart muscles become hardened 35 . Moreover, the change in the calcium concentrations in the smooth muscle of vessels may lead to muscle contraction and therefore increases the levels of blood pressure in CAD patients. In the present study, vitamin D intervention in CAD patients could suppress the production of PTH levels, in comparison with control group. This result was in accordance with findings from a study conducted by Mirhosseini et al. 32 .
Limitations of previous studies included. All of the 4 included studies had a small sample size and short duration of interventions. Moreover, smoking status was not reported in most of these studies, which may influence the outcomes of patients with CAD treated with vitamin D supplements.
Current study limitations and strengths. Due to limited number of studies no meta-regression or subgroup analysis were conducted on the effect of confounding factors on the results of current study. Moreover, different types, doses and durations of vitamin D supplements were used (ergocalciferol and calcitriol), which may lead to some limitations to our analysis. The strengths of the current study include: use of only randomized clinical trial studies with low risk of bias which are considered as the gold standard. A comprehensive search on electronic databases with no language restrictions and no publication bias were conducted in this systematic review and meta-analysis.

Conclusions
In conclusion, our results indicated that vitamin D supplementation in vitamin D deficient subjects had a favorable effect on diastolic blood pressure levels and parathyroid hormone concentrations in comparison with control group. Therefore, vitamin D may be recommended to be used as an adjunct therapy to routine treatment in coronary artery disease patients with vitamin D deficiency. However, further well-designed clinical trials with on a larger scale and of longer duration are required to determine the actual impact of vitamin D supplementation on clinical outcomes of patients with CAD.