Hospital and laboratory outcomes of patients with COVID-19 who received vitamin D supplementation: a systematic review and meta-analysis of randomized controlled trials

Coronavirus disease 2019 (COVID-19) has a wide-ranging spectrum of clinical symptoms, from asymptomatic/mild to severe. Recent research indicates that, among several factors, a low vitamin D level is a modifiable risk factor for COVID-19 patients. This study aims to evaluate the effect of vitamin D on hospital and laboratory outcomes of patients with COVID-19. Five databases (PubMed, Embase, Scopus, Web of Science, and Cochrane Library) and clinicaltrials.gov were searched until July 2022, using relevant keywords/Mesh terms. Only randomized clinical trials (RCTs) that addressed the topic were included. The Cochrane tool was used to assess the studies’ risk of bias, and the data were analyzed using the review manager (RevMan 5.4). We included nine RCTs with 1586 confirmed COVID-19 patients. Vitamin D group showed a significant reduction of intensive care unit (ICU) admission (risk ratio = 0.59, 95% confidence interval (CI) [0.41, 0.84], P = 0.003), and higher change in vitamin D level (standardized mean difference = 2.27, 95% CI [2.08, 2.47], P < 0.00001) compared to the control group. Other studied hospital and laboratory outcomes showed non-significant difference between vitamin D and the control group (P ≥ 0.05). In conclusion, vitamin D reduced the risk of ICU admission and showed superiority in changing vitamin D level compared to the control group. However, other outcomes showed no difference between the two groups. More RCTs are needed to confirm these results. Supplementary Information The online version contains supplementary material available at 10.1007/s00210-022-02360-x.


Introduction
Coronavirus disease 2019  is generated by the novel beta coronavirus known as severe acute respiratory syndrome coronavirus 2 (SARSCoV2). The disease already had spread across the globe and leading the World Health Organization to declare it a pandemic. Since then, more than 508 million proven cases and 6 million mortalities have been reported through April 26, 2022 (Dong et al. 2020;Hu et al. 2021a;Cannata-Andía et al. 2022). COVID-19 has a wide spectrum of clinical symptoms, from asymptomatic or milder symptoms with fever, tiredness, and dry cough to severe and critical symptoms with dyspnea, intensive care unit (ICU) admission, acute respiratory distress syndrome, and multiorgan damage. Immunodeficiency, black ethnicity, older age, chronic kidney disease, obesity, and chronic metabolic disorders are risks related to COVID-19 severity (Xie et al. 2020;Hu et al. 2021b;Olumade and Uzairue 2021;Zhang et al. 2021;Guan et al. 2020).
Vitamin D deficiency was linked to the severity of viral diseases like influenza (Watkins et al. 2015). Recent research indicates that, among several factors, a low vitamin D level is a risk factor that can be modified for COVID-19 patients (Borna et al. 2022;Ilie et al. 2020;Grant et al. 2020;Meltzer et al. 2020). Vitamin D is known to have an antiinflammatory effect, modulate innate and adaptive immunological responses, and enhance the volume of antimicrobial proteins (AlSafar et al. 2021;Pinheiro et al. 2021; Malek Mahdavi 2020; Gois et al. 2017). According to new evidence, it possibly mitigates SARS-CoV-2 expression of the gene and reduces infection by binding to its receptor (Brito et al. 2021b;Glinsky 2020). However, there is no conclusive proof of vitamin D's preventative and therapeutic significance in COVID-19 (Brito et al. 2021a).

Materials and methods
We depended on the PRISMA-P statement and the guideline of the Cochrane handbook for systematic reviews during this systematic review and meta-analysis (Higgins et al. 2019;Page et al. 2021).

Eligibility criteria and study selection
All RCTs (S) that reported on COVID-19 patients (P) who received vitamin D supplementation (any type) (I) and compared their hospital and laboratory outcomes (O) with similar patients who received no intervention/placebo (C). Two types of outcomes were of this review focus as the following: • Primary outcomes (hospital): The need for ICU admission, ventilation and oxygen therapy, the risk of death, and the length of hospital stay (days). • Secondary outcomes (laboratory): The level of C-reactive protein (mg/dL), interleukin-6 (pg/mL), vitamin D concentration, lactate dehydrogenase (LDH), calcium concentration, creatinine, d-dimer, neutrophil count, lymphocyte count, platelet count, and leucocytes (no./μL).
Studies of other designs were excluded, including case reports, case series, reviews, editorials, in vitro, postmortem, conference abstract, letters to the editor, and author opinion papers. Titles and abstracts of potentially included studies were screened to include relevant ones, and then the fulltexts were reviewed thoroughly to confirm the eligibility to be finally included. Four independent authors conducted the previous two steps, but in cases of indecision, a supervisor was involved to confirm the decision.

Data extraction and risk of bias assessment
Three authors extracted the following baseline items from the included trials: (a) general data: study arms, sample size, sex, age, and body mass index (BMI) (kg/m 2 ) of participants; (b) comorbidities outcomes: diabetes, chronic obstructive pulmonary disease, hypertension, cardiovascular disease, and asthma; and (c) common COVID-19 symptoms: fever, cough, weakness, and diarrhea. Another three reviewers extracted the following summary data from the included trials, including NCT, vitamin D administration, follow-up period, and study's primary outcomes and main findings. Six authors extracted the outcomes mentioned above.
The quality of the RCTs was appraised independently by five co-authors using the Cochrane tool to assess the risk of bias reported in the Cochrane Handbook for Systematic Reviews (part 2, chapter 8.5), which categorized the evaluated studies into three categories: high, low, or unclear risk. Indecisions, if any, were resolved through discussion and consensus with six co-authors.

Statistical analysis
We conducted this meta-analysis using Review Manager Software 5.4. Continuous outcomes were pooled as mean difference (MD) and 95% confidence intervals (CIs). In case of different assessment tools/devices, the data were pooled as standardized mean difference (SMD). Dichotomous outcomes were pooled as risk ratio (RR) and 95% CI. We pooled the data under the fixed-effect model and tested the heterogeneity between pooled studies by X 2 and I 2 tests. Once the heterogeneity between studies was detected (P-value < 0.1 and I 2 > 50%), a random-effect model was used. We tried to solve the heterogeneity by sensitivity analysis using the leave-one-out method. The data were considered statistically significant if P-value < 0.05. Since the number of the included studies (n = 9) is less than 10, the publication bias could not be evaluated, according to Egger et al. (1997).

Literature search
We retrieved 1244 records through an extended literature search on different search engines and excluded 571 papers by duplicate removal. The title and abstract screening excluded 641 articles. Thirty-two articles underwent full-text screening, and nine RCTs matched our criteria and entered all steps of meta-analysis to get the evidence (Cannata-Andía et al. 2022;Entrenas Castillo et al. 2020;Fernandes et al. 2022;Maghbooli et al. 2021;Murai et al. 2021b;Rastogi et al. 2022;Mariani et al. 2022;Karonova et al. 2022;Soliman et al. 2021). See the PRISMA chart in Fig. 1.

Characteristics of included studies
We included all RCTs that studied the effect of vitamin D on hospital and laboratory outcomes of 1586 confirmed COVID-19 patients with a mean (SD) age of 56.41 (11.69). The intervention and control groups sample ranged from 16 to 274 participants, and males were more prominent than females. COVID-19 symptoms varied among patients, including fever, cough, general weakness, and diarrhea. Most patients received oral administration regarding vitamin D supplementation, but a single group received an intramuscular injection. Most participants suffered from comorbidities such as hypertension, diabetes, or asthma. Researchers specified the follow-up duration by either period ranged from 7 days to 4 months or an event such as intensive care unit admission, hospital discharge, or death. The summary and the baseline features of included RCTs are shown in Tables 1 and 2.

Risk of bias
The quality of the selected RCTs ranged from moderate to high quality. Eight and six studies were low-biased in random sequence generation and allocation concealment domains, respectively. Participants' blinding occurred in four trials, while the blinding of outcome assessors occurred in five. Seven studies contained no attrition bias. Reporting bias domain was low-biased in all the included trials. Five studies were judged as high biased regarding other sources of bias domain. The risk of bias graph is presented in Fig. 2.

Change in vitamin D
Under random-effect model the results were still significant (SMD = 2.62, 95% CI [0.95, 4.29], P = 0.002) and the heterogeneity could not be solved by sensitivity analysis (Suppl. Figure 3b).  Figure 5).

Change in LDH (U/L)
NR not reported   Figure 6b).

Discussion
This systematic review and meta-analysis of nine RCTs aimed to find a definitive role of vitamin D on hospital and laboratory outcomes of COVID-19 patients. The analysis showed a significantly reduced risk of ICU admission. Also, vitamin D3 levels significantly affect its level positively. However, administration of vitamin D showed no significant difference compared to placebo regarding most hospital-related outcomes of the COVID-19 disease, including requiring ventilation, requiring oxygen therapy, death rate, and length of hospital stay. As for laboratory outcomes, a non-significant difference was also detected in the change in levels of interleukin-6, C-reactive protein, LDH, serum calcium, serum creatinine, d-dimer, neutrophil count, lymphocyte count, platelet count, and leucocytic count. As for the ICU admission, our results showed a significant reduction in COVID-19 patients who received vitamin D. However, after solving the heterogeneity, the results turned non-significant. Our results were supported by another meta-analysis that concluded the positive effect of vitamin D on ICU admission; however, this study included observational studies, which may affect the results . Another meta-analysis of six studies suggested the influential role of vitamin D in ICU admission (Tentolouris et al. 2022). In another RCT, a significantly lower likelihood of ICU admission was maintained even after correcting for comorbidities such as hypertension and diabetes (Entrenas Castillo et al. 2020).
The previously mentioned results differ from Rawat et al., which excluded the retrospective study and found a nonsignificant effect on ICU admission (Rawat et al. 2021). The first multicenter, double-blind RCT in moderate-severe COVID-19 patients concluded that receiving a single high dosage of vitamin D3 (200,000 IU orally) did not lower the ICU admission, length of hospital stay, or rates of mechanical ventilation compared to peanut oil (Murai et al. 2021b). In another multicenter RCT on mild-moderate COVID-19 patients, insignificant changes in ICU or mortality events were observed even though the vitamin D arm had a considerably quicker recovery time to symptoms (even after controlling for age, gender, BMI, and d-dimer) (Sabico et al. 2021). The variations between the abovementioned studies may be due to the different comorbidities, the standard of care, severity of COVID-19, and vitamin D levels at the beginning of each trial. Regarding ventilation, previous studies reported inconsistent results with ours (Rawat et al. 2021;Bassatne et al. 2021;Murai et al. 2021a, b;Maghbooli et al. 2021). However, Maghbooli et al. concluded that vitamin D would benefit COVID-19 patients despite the insignificant results (Maghbooli et al. 2021).
Previous studies supported our results regarding death from COVID-19 and found that vitamin D did not reduce mortality (Tentolouris et al. 2021;Bassatne et al. 2021;Shah et al. 2021;Rawat et al. 2021;Cannata-Andía et al. 2022;Hernández et al. 2021;Murai et al. 2021b;Sabico et al. 2021;Beran et al. 2022). In contrast, Varikasuvu et al. reported that vitamin D significantly reduces mortality (Varikasuvu et al. 2022a). Other studies also reported a significant reduction in mortality favoring vitamin D over placebo (Nikniaz et al. 2021;Pal et al. 2022). Furthermore, a positive association between vitamin D insufficiency and the increased mortality from COVID-19 was detected, especially in the elderly . This was explained by lower exposure to the sun, lower levels of 7-dehydrocholesterol in the skin, higher risk of severe COVID-19 due to comorbidities, and interference of vitamin D levels by the drugs used to treat these comorbidities (Adami et al. 2009;Pimenta et al. 2015;Grant et al. 2020;Jin et al. 2020). Also, Drame et al., in their systematic review, suggested an association between vitamin D deficiency and increased risk of COVID-19 positivity, unfavorable disease course, bad outcomes regarding mortality, disease severity, oxygen therapy requirements, and ventilation need (Dramé et al. 2021). Elamir et al. reported that the intervention group did not affect the length of hospital stay and intubation need, which supports our results; however, they reported a significant reduction in oxygen therapy requirements, which is inconsistent with ours (Elamir et al. 2022). They explained this by the small number of participants in the trial but suggested a beneficial role of vitamin D on respiration (Elamir et al. 2022). A recent metaanalysis reported that vitamin D benefits both length of hospital stay and intubation requirements, which contrasts with our results (Beran et al. 2022). Another cohort analysis of the length of hospital stay and the death rate showed superiority in the highest serum calcidiol group (> 25 ng/mL) (Cannata-Andía et al. 2022). However, Bassatne et al. reported insignificant results (Bassatne et al. 2021). So, determination of any vitamin D deficiency in any patients is mandatory as the baseline vitamin D level would influence the benefits of its supplementation and the COVID-19 outcomes (Griffin et al. 2020).
Hypercalcemia was not observed in our included studies either in intervention or control groups, which means no difference between groups and proves the safety of vitamin D on the calcium level (Elamir et al. 2022;Rastogi et al. 2022). Previous research found a significant increase from baseline in vitamin D levels after vitamin D3 supplementation, consistent with our results (Fernandes et al. 2022;Rastogi et al. 2022;Murai et al. 2021b;Soliman et al. 2021). In similarity to our results, other studies reported insignificant results regarding d-dimer, CRP, IL-6, and LDH levels (Rawat et al. 2021;Rastogi et al. 2022;Fernandes et al. 2022;Maghbooli et al. 2021). It is known that COVID-19 raises inflammatory markers like d-dimer, fibrinogen, IL-6, and CRP, especially in severe cases, which are considered good indicators for severity and recovery of COVID-19 (Velavan and Meyer 2020).
Previous studies reported consistent results with ours regarding serum creatinine levels (Cannata-Andía et al. 2022;Maghbooli et al. 2021;Murai et al. 2021b). Furthermore, similar to our results regarding the change in the count of platelets, lymphocytes, and leucocytes, some researchers reported insignificant results, but others reported significant results regarding lymphocytic count (Maghbooli et al. 2021;Murai et al. 2021b). In response to inflammation such as COVID-19 events, leukocytes provide innate immunity, and lymphocytes provide adaptive immunity, so body defense occurs (Denman 1979).
An acute illness such as COVID-19 reduces the circulation of vitamin D binding protein and interferes with the effective production of the body's active form of vitamin D (Zehnder et al. 2001;Waldron et al. 2013). These phenomena may help explain the conflict between studies regarding the effectiveness of vitamin D on COVID-19.
Our study has several strengths which support the quality of the evidence. For example, we applied a comprehensive search strategy and literature search on different databases without language or time restrictions. We included only relevant RCTs that studied clinical and laboratory outcomes and excluded any other design. The included trials are considered low-biased regarding many quality assessment domains, which is supportive.
However, we found high heterogeneity between the included studies, such as different populations' characteristics, including age, sex, race, body mass index, general status, the severity of COVID-19 symptoms, treatment protocol of the patients, and associated comorbidities. The regimens of vitamin D supplementation also varied across the studies regarding the form, the dose, the timing of administration, and the baseline levels of vitamin D. Patients received variable amounts of vitamin D, ranging from low to high doses and from single to daily doses. Previous research found that the daily doses of vitamin D prevent and treat certain diseases such as acute respiratory infections, rickets, and tuberculosis better than the intermittent doses . Most of the studies included a low sample size, which also affected the quality of the evidence. During acute illness, vitamin D binding protein and albumin tend to decrease by the negative acute phase response, which affects vitamin D levels bound to them . Time of vitamin D administration also would impact its effect as most patients have received it after being infected and diagnosed with COVID-19.

Conclusions
Our study suggested that vitamin D supplementation benefits COVID-19 patients by reducing ICU admission and increasing changes in vitamin D levels. However, it produces no difference in other outcomes compared to no vitamin D intake. The definite role of vitamin D on COVID-19 outcomes strongly needs further well-conducted and high-quality research, especially after its known effect on the body's immune system and defense mechanisms and the previously collected data on its benefits on certain respiratory diseases, including COVID-19. Data availability The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations
Ethics approval Not applicable.

Competing interests The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.