key: cord-0269628-kd3as3b9
authors: Varikasuvu, S. R.; Thangappazham, B.; Raj, H.
title: COVID-19 and Vitamin D (Co-VIVID Study): a systematic review and meta-analysis of randomized controlled trials
date: 2021-08-25
journal: nan
DOI: 10.1101/2021.08.22.21262216
sha: 4114053cc96b939698a45b79f710b438ce350439
doc_id: 269628
cord_uid: kd3as3b9

Background: Vitamin D levels have been reported to be associated with COVID-19 susceptibility, severity and mortality events.. We performed a meta-analysis of randomized controlled trials (RCTs) to evaluate the use of vitamin D intervention on COVID-19 outcomes. Methods: Literature search was conducted using PubMed, Cochrane library, and ClinicalTrials.gov databases (latest search on August 5, 2021). We included RCTs reporting the use of vitamin D intervention to control/placebo group in COVID-19. Two independent researchers did literature search, abstracted data, and the risk of bias assessment. Results: A total of 6 RCTs with 551 COVID-19 patients were included. The overall collective evidence pooling all the outcomes across all RCTs indicated the beneficial use of vitamin D intervention in COVID-19 (relative risk, RR = 0.60, 95% CI 0.40 to 0.92, Z=2.33, p=0.02, I2 = 48%). However, no statistical significance was observed for individual outcomes of ICU care (RR = 0.11, 95% CI 0.15 to 1.30, Z=1.48, p=0.14, I2 = 66%) and mortality (RR = 0.78, 95% CI 0.25 to 2.40, Z=0.66, p=0.02, I2 = 33%), though decreased rates were noted. The rates of RT-CR positivity was significantly decreased in the intervention group as compared to the non-vitamin D groups (RR = 0.46, 95% CI 0.24 to 0.89, Z=2.31, p=0.02, I2 = 0%). Conclusion: COVID-19 patients supplemented with vitamin D are more likely to demonstrate fewer rates of ICU admission, mortality events and RT-PCR positivity. However, no statistical significance has been achieved for individual outcomes of ICU and deaths. More RCTs and completion of ongoing trials largely needed to precisely establish the association between vitamin D use and COVID-19.

confirmed COVID-19 cases, including 4,255,892 deaths [1] . Multiple risk factors in the form of age, comorbidities, exaggerated immune response in the form of cytokine storm, oxidative stress, activation of pro-coagulation factors and severe inflammation contribute to the disease progression [2] .

It has been documented that vitamin D deficiency is associated with severity of viral infections such as influenza [3] . Recent evidence shows the potential of vitamin D to affect SARS-CoV-2 gene expression and alleviate infection upon binding to the vitamin D response element [4, 5] .

Vitamin D regulates the rennin-angiotensin system and expression of angiotensin converting enzyme 2 (ACE2), and its receptor that mediates SARS-CoV-2 infection. Further, vitamin D is 6 positivity in treated and control groups. We reported RR values with their 95% confidence intervals (CI) using Mantel-Haenszel analysis method and random-effects model. The overall effect size for RR was presented Z-score. A Z-score with a p value of <0.05 was considered statistically significant. The between-study heterogeneity was examined by the I 2 statistics and the values >50% were considered to indicate a high degree of heterogeneity [22] . We examined the funnel plot asymmetry for publication bias followed by Begg and Egger's tests.

We conducted sub-group analysis based on vitamin D form, vitamin D deficient studies, single or multi-centric trials, and double-blinded status. We also performed a one-study leave-out sensitivity analysis for individual outcomes by excluding one trial at a time and by repeating the analysis. The meta-regression analysis was not possible due to the small number of available trials.

We reviewed 755 articles for eligibility, 6 RCTs [14] [15] [16] [17] [18] [19] comprising 551 COVID-19 patients were selected for final analysis (Fig. 1) .

While all the studies enrolled participants aged >18 years with mean age in individual studies range from 36 to 56 years, the proportion of men varied from 44 to 59%. The symptoms of COVID-19 patients diagnosed by RT-PCR (viral RNA) or ELISA and/or radiographic testing varied across the individual studies (mild-moderate-severe). The criteria for inclusion and exclusion, varied study settings, participant characteristics, number of participants with preexisting comorbidities and treatment strategies, vitamin D form, dosage, reported outcomes and other study characteristics are presented in Table 1 .

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 25, 2021. ; https://doi.org/10.1101/2021.08.22.21262216 doi: medRxiv preprint There were two multi-center [16, 18] and four single center RCTs [14, 15, 17, 19] , one doubleblinded [16] and four registered clinical trials [14] [15] [16] 18 ]. Vitamin D treatment was compared to placebo in two studies [16, 17] , non-vitamin control in three studies [14, 15, 19] , and standard treatment comparator group in one study [18] . While Castillo et al. [14] used calcifediol with an allocation ratio of 2:1; all other studies used cholecalciferol with an allocation ratio of 1:1. The baseline vitamin D statuses in three studies [15, 17, 18] were reported to be sub-optimal and one study reported a separate outcome analysis in vitamin D deficient participants [16] . The vitamin D sufficiency status, treatment doses, follow-up durations, adverse events and study limitations are detailed in Table 1 . The risk of bias assessment based on five domains and the overall bias of included RCTs is presented the supplementary appendix.

The collective evidence in Fig.2 shows that vitamin D treatment was significantly associated with reduced risk of COVID-19 severity when six observations on the number events for symptom severity, ICU care and mechanical ventilation were pooled (RR = 0.46, 95% CI 0.23 to 0.93, Z=2.16, p=0.03, I 2 = 52%). But the pooled estimate from four studies showed that the use of vitamin D was not significantly associated with ICU outcome alone (RR = 0.11, 95% CI 0.15 to 1.30, Z=1.48, p=0.14, I 2 = 66%).

The pooled estimate from two studies showed a statistically significant RR for COVID-19 RT-PCR positivity (RR = 0.46, 95% CI 0.24 to 0.89, Z=2.31, p=0.02, I 2 = 0%). Whereas the pooled evidence from four studies showed that the association of vitamin D with mortality outcome was not statistically significant (RR = 0.78, 95% CI 0.25 to 2.40, Z=0.66, p=0.02, I 2 = 33%).

However, when all the observations on all reported outcomes were pooled, there was statistically significant evidence on the use of vitamin D treatment in reducing overall COVID-19 related All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 25, 2021. ; https://doi.org/10.1101/2021.08.22.21262216 doi: medRxiv preprint outcomes (RR = 0.60, 95% CI 0.40 to 0.92, Z=2.33, p=0.02, I 2 = 48%). The test for subgroup differences was not statistically significant (I 2 = 49%, p = 0.12).

The results of sub-group analysis were presented in Table 2 . None of the outcomes in different categories of subgroups showed statistically significant RR values. No statistically significant difference was observed for the pooled estimate of outcomes from studies with vitamin D suboptimal status. The sensitivity analysis performed leaving-out any one of the included trials at a time and repeating the analysis showed statistically non-significant RR values for individual outcomes. Whereas, for all studied outcomes together, the pooled RR remained statistically significant after leaving our any particular study/observation. The I 2 value significantly changed from 48% to 5% after leaving-out a study by Castillo et al. (ICU and mortality observations) and

repeating the analysis suggestive of major source of heterogeneity. The funnel plot analysis ( Fig.3) with Begg's (p = 0.17) and Egger tests (p = 0.14) on all the outcomes across all the RCTs indicated no significant publication bias.

This meta-analysis of RCTs showed that COVID-19 patients supplemented with vitamin D had reduced overall risk for all outcomes. The collective overall evidence on severity, ICU care, mortality, sero and RT-PCR positivity events reported in all trials indicated that COVID-19 patients treated with vitamin D showed lower rates of these outcomes relative to patients receiving no-vitamin D/standard/placebo. Though there were no statistically significant differences in the individual outcomes of ICU admission and mortality, the respective RRs indicated a decrease in the rates of these outcomes in vitamin D treated groups. However, there All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. This study reports more mortality events in the intervention arm (9/119) than the placebo (6/118) group.

In another multicenter RCT [18] randomizing 73 mild-moderate COVID-19 patients with suboptimal vitamin D status into experimental (n=36) and standard-comparator (n=33) groups receiving 5000 IU and 1000 IU of oral cholecalciferol daily for two weeks. This study though reports a significantly shorter recovery time to symptoms (even after adjusting for age, sex, BMI and D-dimer) in the intervention arm, no significant differences in ICU, mortality events and days to discharge were reported between groups. This study differs from that of Murai et al. [16] as it excludes severe COVID-19 cases, vitamin D dosage and duration, using standard comparator group in place of placebo, and in defining the suboptimal vitamin D status (<50 nmol/L). Further, this study also differs from all other trials as 47% of randomized participants had also received vitamin C supplements. The significant increase in vitamin D levels reported in treatment arm (5000 IU) post intervention along with other study findings are to be interpreted All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 25, 2021. and placebo arms of 20 cases each. The intervention arm received 60000 IU of cholecalciferol daily for 7 days and continued for another 7 days in six cases (who did not achieve a therapeutic target of >50 ng/mL on day 7) and distilled water was supplied to placebo group.

There were two open label RCTs [14, 15] . Lakkireddy et al.

[15] randomized 130 mild-moderate COVID-19 cases, of which 87 cases who completed the study were analyzed in the intervention (n=44) and comparator (n=43) groups. The intervention arm received 60000 IU of oral vitamin D3 daily for 8-10 days and the outcomes were recorded till 21 days. Supplementation resulted in a significant increase in vitamin D levels with a lower rate of ICU and mortalities in the intervention arm as compared to the comparator group. In the only trial using calcifediol, Castillo All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 25, 2021. ; 1 1 et al. [14] randomized 76 patients into intervention (n=50) and comparator (n=26) groups depending on whether or not supplemented with calcifediol. The oral calcifediol supplemented varied at admission (0.532 mg), on days 3 and 7 (0.266 mg) and weekly until ICU/discharge (0.266 mg). This study concludes that vitamin D treatment resulted in significantly less probability of ICU admission and the statistical significance retained even after adjusting for comorbidities like diabetes and hypertension. However, there is no information available on the baseline and post-treatment vitamin D levels.

In general, it has been demonstrated that vitamin D induce antimicrobial peptides and mediates antiviral, apoptotic and autophagic activities [12, 23] . [13] , the results of this study strongly suggests the need for future/ongoing RCTs to consider better designs, large sample sizes adequate enough to assess the effect of vitamin D supplementation on the individual COVID-19 related outcomes. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. However, this study has some limitations. First, the heterogeneity observed in the meta-analysis could be due to methodological, participant and treatment variations of the included trials. While the single center RCTs have mainly contributed to the heterogeneity, leaving-out a study by Castillo et al. [14] decreased the I 2 values from 48 to 5%, 66 to 0%, and 33 to 13% for the overall In conclusion, vitamin D use was associated with significant decrease in rates of COVID-19 related events when all the outcomes were pooled across all RCTs. However, there was no significant difference observed for the relative risk for ICU admission and mortality outcomes upon vitamin D supplementation. The overall pooled results in addition to a significant decrease All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 25, 2021. 

The costs involved in the conduction of this systematic review and meta-analysis were borne by the authors themselves.

The authors declare no conflict of interest. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 25, 2021. ; 1 8 (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted August 25, 2021.

;

COVID-19) Dashboard | WHO Coronavirus (COVID-19) Dashboard With Vaccination Data n

Aging & COVID-19 susceptibility, disease severity, and All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity

Vitamin D Status Is Associated With In-Hospital Mortality and Mechanical Ventilation: A Cohort of COVID-19 Hospitalized Patients

Low Circulating Vitamin D in Intensive Care Unit-Admitted COVID-19 Patients as a Predictor of Negative Outcomes

Evidence Regarding Vitamin D and Risk of COVID-19 and Its Severity

Therapeutic and prognostic role of vitamin D for COVID-19 infection: A systematic review and meta-analysis of 43 observational studies

Dr. S.R. Varikasuvu specially acknowledge Bhairavi Sisters (Sahasra and Aagneya) for the time I could not give them during this work, as it was performed before 8 am and after 6 pm from its initiation to completion.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted August 25, 2021.