Effect of vitamin D supplementation on COVID-19 patients: A systematic review and meta-analysis

Objective To systematically evaluate the impact of vitamin D supplementation on mortality, ICU admission, and the rates of mechanical ventilation or intubation among COVID-19 patients. Data sources and study selection The PubMed, Embase, Cochrane Library, CBM, CNKI, VIP, and WanFang databases were searched from 1 December 2019 to 31 December 2022. The authors sought to identify randomized controlled trials and cohort studies that examined the relationship between vitamin D supplementation and mortality, ICU admission, and mechanical ventilation or intubation rates among COVID-19 patients. Data extraction and synthesis Two investigators independently searched the literature, extracted the data, and assessed the quality of the included studies. The Grading of Recommendation, Assessment, Development, and Evaluation approach was used to evaluate the quality of the evidence. Meta-analysis was conducted using RevMan 5.3, STATA 15.1, and R 4.1.3 software. Results Eight randomized controlled trials (RCTs) and eight cohort studies were included, involving 3359 COVID-19 patients. The pooled analysis of randomized controlled trials showed that vitamin D supplementation did not have a significant effect on reducing mortality (Relative Risk, RR = 0.94, 95% CI 0.69–1.29, P = 0.7), while the results of cohort studies indicated that vitamin D supplementation had a positive impact on reducing mortality among COVID-19 patients (RR = 0.33, 95% CI 0.23–0.47, P < 0.001). There was no statistically significant difference in the rates of ICU admission (RCTs: RR = 0.64, 95%CI 0.38–1.08, P = 0.10; cohort studies: RR = 0.32, 95% CI 0.08–1.29, P = 0.109) or rates of mechanical ventilation or intubation (RCTs: RR = 0.77, 95% CI 0.58–1.02, P = 0.07; cohort studies: RR = 0.93, 95% CI 0.55–1.58, P = 0.789). Conclusion The results of this systematic review and meta-analysis suggest that vitamin D supplementation does not have a significant impact on reducing mortality, ICU admission, and the rates of mechanical ventilation or intubation among COVID-19 patients. However, due to the limited number and quality of the studies included, further high-quality studies are needed to confirm these findings. Systematic review registration www.crd.york.ac.uk, identifier CRD42021299521.


Introduction
The global outbreak of coronavirus disease 2019 (COVID- 19) has caused a major health crisis with 655,689,115 confirmed cases and 6,671,624 confirmed deaths as of 3 January 2023 (1). The infection caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) leads to a wide range of symptoms, and patients with comorbidities such as diabetes, cardiovascular disease, and hypertension may face adverse outcomes (2), including ICU admission, mechanical ventilation or intubation, and death.
While vaccines and antiviral drugs have demonstrated efficacy against , additional measures, such as vitamin D supplementation, continue to play an important role in managing the disease. Low serum 25-hydroxycholecalciferol [25(OH)D] levels have been linked to increased susceptibility to novel coronavirus infection and greater severity of COVID-19 symptoms (4). Some studies have suggested that vitamin D supplementation may reduce mortality in COVID-19 patients (5,6), but a previous meta-analyze published in the year 2022 has failed to reach a definitive conclusion due to limited studies and inconsistent study design (7).
With the ongoing spread of COVID-19, the number of clinical studies on the effect of vitamin D supplementation on COVID-19 outcomes has increased (5, 6, 8-13) but the results remain conflicting. Thus, it is necessary to conduct an updated metaanalysis of randomized controlled trials and cohort studies to determine the impact of vitamin D supplementation on mortality, ICU admission, and mechanical ventilation or intubation rates in COVID-19 patients.

Materials and methods
The present meta-analysis was conducted following the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analysis) statement (14) and has been registered on the international database of prospectively registered systematic reviews, PROSPERO (Registration number: CRD42021299521).

Study selection and data extraction
Two investigators independently searched the literature, extracted the data, cross-checked the data, and consulted a third party to resolve any disagreements. The titles and abstracts of the literature were initially screened, followed by a full-text review to determine final inclusion based on the established inclusion and exclusion criteria. The extracted data included (1) the first author, year of publication, location, and date of the study; (2) baseline characteristics and interventions of subjects; and (3)outcome indicators and data, including mortality, ICU admission rates, and mechanical ventilation or intubation rats in COVID-19 patients.

Risk of bias assessment
The assessment of the risk of bias in the included literature was carried out independently by two investigators, and the results were verified through cross-checked. The risk of bias in cohort studies was evaluated using the Robin-I tool by the Cochrane guidelines for non-randomized studies (15), and RCTs were evaluated by the Cochrane Collaborations Tool For Assessing Risk of Bias recommended by the Cochrane Manual 5.1.0 (16).

Statistical analysis
RevMan (version 5.3) software (Cochrane Collaboration, UK), Stata (version 15.1) software (Stata Corporation, Lakeway, TX, USA) and R software (version 4.1.3) were used for meta-analysis. The effect size was analyzed using relative risk (RR) and a 95% confidence interval (CI). Hazard ratio (HR) was considered as RR in the study, and the following formula was used to convert odds ratio (OR) into RR: RR = OR/[(1 − Po) + (Po × OR)], where Po represents the incidence of the outcome of interest in the non-exposed group (17). The standard error of the resulting converted RR was calculated using the formula: SElog(RR) = SElog(OR) × log(RR)/log(OR). The adjusted HR or RR and 95% CI were utilized to reduce the impact of confounding factors if available. Otherwise, unadjusted HR or RR was adopted. The heterogeneity of the included studies was analyzed using the Q test, and if I 2 < 50% and P > 0.1, all studies were considered homogenous and the data were analyzed by a fixed-effect model. In case of I 2 ≥ 50% and P ≤ 0.1, indicating the presence of heterogeneity, data were analyzed using a random effects model. Potential publication bias was evaluated through funnel plots and Egger's test.
Stratified analyses were performed based on the type of study design, and sensitivity analyses were conducted to test the reliability of the combined analysis of adjusted/unadjusted RR.

Quality of evidence
The quality of the evidence was evaluated using the Grading of Recommendation, Assessment, Development, and Evaluation (GRADE) approach (18, 19) and was classified as high, moderate, low, or very low based on the following domains: study design, risk of bias, inconsistency, indirectness, imprecision, and other considerations (such as evidence of publication bias). The results are presented in Table 2.    Table 1 presents the characteristics of the included studies. The RCTs included 1,318 subjects, with 677 in the vitamin D supplementation group and 641 in the control group. The cohort studies included 2,041 subjects, with 976 in the vitamin D supplementation group and 1,065 in the control group. All the studies were carried out in hospitals, except for one which was conducted in a nursing home in France (6). The sample sizes of RCTs ranged from 43 to 543, with mean or median ages ranging from 10.7 to 69 years and follow-up from 7 days to 4 months (8-10, 20-24). Cholecalciferol was administered in the intervention arm of six RCTs (9, 10, 20-22, 24), while calcifediol (23) and calcitriol (8) were used in the remaining two RCTs. The sample sizes of the eight cohort studies ranged from 48 to 785, with mean ages ranging from 45.5 to 87.7 years, and follow-up from 5 days to 3 months. Cholecalciferol was administered in the intervention arm of six cohort studies (5,6,11,13,25,27), and calcifediol was administered in the remaining two studies (12,26). Out of the 16 included studies, only 10 reported the mean baseline levels of serum 25(OH)D, which ranged from 6.65 to 32.5 ng/ml in the intervention groups and 7.14 to 30.5 ng/ml in the control groups ( Table 1).

Study characteristics and risk of bias of the included literature
Four RCTs had a low risk of bias (10,20,22,24), one was at a high risk of bias (21) and the rest three studies had an uncertain risk of bias (8, 9, 23) (Supplementary Figures 1, 2). Six cohort studies had a moderate risk of bias (5,12,13,(25)(26)(27), and the other two had a serious risk of bias (6,11) (Supplementary Figure 3).

GRADE assessment
The quality of evidence was assessed using the GRADE methods, as presented in Table 2. The certainty of the evidence for mortality (RCTs were very low, cohort studies were low), ICU admission (both RCTs and cohort studies were very low), and mechanical ventilation or intubation (both RCTs and cohort studies were very low) were rated as low to very low due to the heterogeneity in drug type and dosing, population characteristic, and the quality of the included studies.

Outcomes of meta-analyses Effect of vitamin D supplementation on mortality
All eight RCTs (n = 1,318) and eight cohort studies (n = 2,041) reported the effect of vitamin D supplementation on mortality in COVID-19 patients. The meta-analysis of RCTs indicated no significant difference in mortality between the intervention group and control group (RR = 0.94, 95% CI 0.69-1.29, P = 0.7; fixed effect model; very low-certainty evidence; Figure 2). For the Some do not concern with the method of randomization used/allocation concealment/blinding of participants/blinding of outcome assessment/selective reporting. b There were differences in vitamin D dosages and duration.
The confidence interval was not narrow enough for us to be confident that this is the effect or does it reduce and has no effect.  Forest plot of RCTs for vitamin D supplementation on mortality. Forest plot of cohort studies for vitamin D supplementation on mortality (All cohort studies).

FIGURE 4
Forest plot of cohort studies for vitamin D supplementation on mortality (studies with adjusted RR values only).
Frontiers in Nutrition 07 frontiersin.org Subgroup analyses of mortality.
eight cohort studies, three reported adjusted HRs, another three reported adjusted ORs, and the remaining two studies reported the number of deaths. Subjects with vitamin D supplementation had significantly lower mortality than the control group (RR = 0.33, 95% CI 0.23-0.47, P < 0.001; fixed effect model; low-certainty evidence; Figure 3). The results remained consistent even after excluding studies that reported unadjusted RRs or numbers of deaths (RR = 0.31, 95% CI 0.21-0.44, P < 0.001; fixed effect model; Figure 4). We performed subgroup analyses to investigate the association between the average daily vitamin D supplement dose and serum 25(OH)D levels with mortality. The results revealed no significant differences in mortality between individuals with baseline 25OHD levels below 20 ng/ml (RR = 0.93, 95% CI 0.66-1.32, P = 0.68) (9, FIGURE 6 Forest plot of RCTs for vitamin D supplementation on ICU admission. Forest plot of cohort studies for vitamin D supplementation on ICU admission. Forest plot of RCTs for vitamin D supplementation on mechanical ventilation or intubation. Forest plot of cohort studies for vitamin D supplementation on mechanical ventilation or intubation.
Frontiers in Nutrition 09 frontiersin.org Funnel plot of RCTs.

FIGURE 11
Funnel plot of cohort studies.

The effect of vitamin D supplementation on ICU admission
Six RCTs and two cohort studies reported the effect of vitamin D supplementation on ICU admission. Meta-analyses showed that there was no difference in ICU admission between the vitamin D supplementation and control groups in either RCTs (RR = 0.64, 95%CI 0.38-1.08, P = 0.10; random effect model; very low-certainty evidence; Figure 6) or cohort studies (RR = 0.32, 95% CI 0.08-1.29, P = 0.109; random effect model; very low-certainty evidence; Figure 7).

The effect of vitamin D supplementation on mechanical ventilation or intubation
Five RCTs and three cohort studies reported the effect of vitamin D supplementation on mechanical ventilation or intubation. Meta-analyses of RCTs (RR = 0.77, 95% CI 0.58-1.02, P = 0.07; fixed effect model; very low-certainty evidence; Figure 8) and cohorts (RR = 0.93, 95% CI 0.55-1.58, P = 0.789; fixed effect model; very low-certainty evidence; Figure 9) showed that there was no difference in mechanical ventilation or intubation rate in COVID-19 patients with or without vitamin D supplementation.

Publication bias
No evidence of publication bias was identified through the analysis of the funnel plots (RCTs' Egger's test P = 0.266, Figure 10; cohort's Egger's test P = 0.604, Figure 11).

Discussion
This present meta-analysis included eight RCTs (8-10, 20-24) and eight cohort studies (5, 6, 11-13, 25-27) involving a total of 3,359 subjects. The results of pooled data indicated that vitamin D supplementation did not significantly reduce mortality, ICU admission, or rates of mechanical ventilation and intubation in COVID-19 patients. The conclusion should be interpreted with caution due to the low quality of the studies included, their small sample sizes, and significant baseline heterogeneity in baseline factors, including drug type and dosing, and population characteristics.
It is widely recognized that vitamin D can regulate the immune system, and its deficiency has been linked to an increased risk of developing the "cytokine storm" associated with COVID-19 (28). Recent reviews of the literature have also suggested that optimizing vitamin D levels in the general population may have served as a protective measure against COVID-19 infection (29,30). Our study is not the first meta-analysis of vitamin D supplementation in COVID-19 patients. A previous meta-analysis published in 2021 (31) comprising 3 RCTs (20, 23, 32) and 2 cohort studies (6,11) found that vitamin D supplementation did not result in a significant reduction in mortality, ICU admission rates, or mechanical ventilation (31). Another meta-analysis published in 2021 (33) involving 2 RCTs (20, 23) and 1 case-control study (34) showed that vitamin D supplementation resulted in comparable mortality but lower intensive care unit needs in patients with COVID-19. These two meta-analyses pooled studies with different study types and had much smaller sample sizes than our study. Our meta-analysis was based on a comprehensive search strategy and use established scales to assess the quality of research and strength of evidence. Furthermore, adjusted ORs were used to minimize bias in cohort studies. As a result, our conclusions are more robust and reliable compared to previous metaanalyses.
The pooled analysis found an inconsistent effect of vitamin D supplementation on mortality in cohort studies and RCTs. Although evidence showed that patients receiving higher cumulative doses and average daily doses had a greater decrease in COVID-19 infection rates compared to those receiving lower doses (35), subgroup analysis indicated that there were no significant differences in mortality between individuals with lower or higher baseline 25OHD levels, as well as those receiving small or larger vitamin D supplementation doses in RCTs. Nevertheless, the results from RCTs were more reliable due to the superior methodology.
There are some limitations in this meta-analysis, including the small sample sizes and low quality of the included RCTs and cohort studies, as well as the lack of complete information regarding the study population, such as race, sex, and 25(OH)D level before and after vitamin D supplementation. There was also significant heterogeneity among the included studies in terms of drug type and dosing, population features, and COVID-19 severity and treatment strategies.
In conclusion, while the results of this meta-analysis suggest that vitamin D supplementation may not significantly reduce mortality, ICU admission, and rates of mechanical ventilation intubation in COVID-19 patients, additional well-designed RCTs with large sample sizes are needed to further explore the potential benefit of vitamin D supplementation in this population.

Data availability statement
The original contributions presented in this study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.