sssss
National Institute for Health and Care Excellence
NG187
Vitamin D for COVID-19
[A] Evidence reviews for the use of vitamin D supplementation as prevention and treatment of COVID-19
NICE guideline NG187
| Evidence reviews underpinning recommendations 1.1 to 1.3
and research recommendations in the NICE guideline |
| December 2020 |
|
Final |
|
These evidence reviews were developed by Centre for Guidelines Methods and Economics Team |
 |
|
| Disclaimer
The recommendations in this guideline represent the view of NICE, arrived at after careful consideration of the evidence available. When exercising their judgement, professionals are expected to take this guideline fully into account, alongside the individual needs, preferences and values of their patients or service users. The recommendations in this guideline are not mandatory and the guideline does not override the responsibility of healthcare professionals to make decisions appropriate to the circumstances of the individual patient, in consultation with the patient and/or their carer or guardian. Local commissioners and/or providers have a responsibility to enable the guideline to be applied when individual health professionals and their patients or service users wish to use it. They should do so in the context of local and national priorities for funding and developing services, and in light of their duties to have due regard to the need to eliminate unlawful discrimination, to advance equality of opportunity and to reduce health inequalities. Nothing in this guideline should be interpreted in a way that would be inconsistent with compliance with those duties. NICE guidelines cover health and care in England. Decisions on how they apply in other UK countries are made by ministers in the Welsh Government, Scottish Government, and Northern Ireland Executive. All NICE guidance is subject to regular review and may be updated or withdrawn. |
|
Copyright |
| © NICE 2020. All rights reserved. Subject to Notice of rights. |
| ISBN: 978-1-4731-3942-8 |
Contents
1 Vitamin D for COVID-19 prevention and treatment…………………………………………. 7
Review question…………………………………………………………………………………………………………….. 7
What is the clinical effectiveness and safety of vitamin D supplementation for
the treatment of COVID-19 in adults, young people and children?…………… 7
Introduction…………………………………………………………………………………………. 7
Summary of the protocol………………………………………………………………………. 7
Methods and process……………………………………………………………………………. 8
Effectiveness evidence………………………………………………………………………… 9
Summary of studies included in the effectiveness and prognostic
evidence…………………………………………………………………………………………… 9
Summary of the effectiveness evidence………………………………………………… 9
Economic evidence……………………………………………………………………………. 10
Review question…………………………………………………………………………………………………. 10
What is the clinical effectiveness and safety of vitamin D supplementation for the prevention of SARS CoV2 infection (and subsequent COVID-19) in
adults, young people and children?……………………………………………………. 10
Introduction……………………………………………………………………………………….. 10
Summary of the protocol…………………………………………………………………….. 10
Methods and process…………………………………………………………………………. 11
Effectiveness evidence………………………………………………………………………. 12
Summary of the effectiveness evidence………………………………………………. 12
Economic evidence……………………………………………………………………………. 12
Review question…………………………………………………………………………………………………. 12
Is vitamin D status independently associated with susceptibility to developing COVID-19, severity of COVID-19, and poorer outcomes from COVID-
19 in adults, young people and children?…………………………………………….. 12
Introduction……………………………………………………………………………………….. 12
Summary of the protocol…………………………………………………………………….. 13
Methods and process…………………………………………………………………………. 13
Association evidence………………………………………………………………………….. 15
Summary of studies included in the effectiveness and prognostic
evidence…………………………………………………………………………………………. 16
Summary of the association evidence………………………………………………….. 26
Economic evidence……………………………………………………………………………. 34
1.3.9 References - included studies…………………………………………………………….. 34
Appendices…………………………………………………………………………………………………………….. 37
Appendix A - Review protocols………………………………………………………………………… 37
Review question 1……………………………………………………………………………………. 37
Review question 2……………………………………………………………………………………. 44
Review question 3……………………………………………………………………………………. 51
Appendix B - Literature search strategies………………………………………………………… 58
Medline ALL………………………………………………………………………………………………………. 58
Embase…………………………………………………………………………………………………………….. 60
Cochrane Database of Systematic Reviews (CDSR) & CENTRAL…………………………. 62
MedRxiv & BioRxiv preprints……………………………………………………………………………….. 64
World Health Organization Global research on coronavirus disease (COVID-19)………. 65
Clinicaltrials.gov…………………………………………………………………………………………………. 66
Appendix C - Effectiveness & association evidence study selection…………………. 67
Review question 1……………………………………………………………………………………. 67
Review question 2……………………………………………………………………………………. 68
Review question 3……………………………………………………………………………………. 69
Appendix D - Effectiveness & association evidence………………………………………….. 70
Effectiveness evidence…………………………………………………………………………….. 70
Vitamin D as treatment…………………………………………………………………………….. 70
Entrenas Castillo 2020……………………………………………………………………………… 70
Association evidence……………………………………………………………………………….. 80
Annweiler 2020………………………………………………………………………………………… 80
Annweiler 2020a………………………………………………………………………………………. 92
Hastie 2020…………………………………………………………………………………………….. 101
Hernandez 2020……………………………………………………………………………………… 110
Karahan 2020…………………………………………………………………………………………. 121
Kaufman 2020………………………………………………………………………………………… 128
Macaya 2020…………………………………………………………………………………………… 134
Meltzer 2020…………………………………………………………………………………………… 141
Merzon 2020…………………………………………………………………………………………… 154
Radujkovic 2020……………………………………………………………………………………… 160
Raisi-Estabragh 2020……………………………………………………………………………… 168
D.2.1.12 Ye 2020…………………………………………………………………………………………………… 179
Appendix E - GRADE tables…………………………………………………………………………… 187
Effectiveness of vitamin D as a COVID-19 treatment……………………………….. 187
Effectiveness of vitamin D supplement as treatment for COVID-19…………. 187
Effectiveness of vitamin as a COVID-19 preventative measure………………… 187
Association between vitamin D status and COVID-19 outcomes…………….. 187
Association between vitamin D status and COVID-19 cases……………………. 187
Association between vitamin D status and severity of COVID-19……………. 189
Appendix F - Ongoing studies (clinicaltrials.gov)…………………………………………… 193
Prevention……………………………………………………………………………………………… 193
As monotherapy…………………………………………………………………………………….. 193
As combination therapy…………………………………………………………………………. 193
Of fundi ng
1 Vitamin D for COVID-19 prevention and treatment
What is the clinical effectiveness and safety of vitamin D supplementation for the treatment of COVID-19 in adults, young people and children?
Vitamin D is hypothesised to have a role in the immune response to respiratory viruses and can potentially mitigate the inflammatory response. During the COVID-19 pandemic, treatments are being explored as options for managing the disease. Therefore, it has been suggested vitamin D could improve outcomes in people diagnosed with confirmed COVID-
19. The 2 major forms of vitamin D, vitamin D3 (cholecalciferol) and vitamin D2 (ergocalciferol), are licensed for the prevention and treatment of vitamin D deficiency and are taken by many people who have vitamin D deficiency. The review aims to assess if vitamin D can be used in all people regardless of vitamin D status as a safe treatment option, alone or in combination with other therapies, to treat COVID-19.
Summary of the protocol
Table 1 PICO table for review question 1
| Field | Content |
| Population | Inclusion: Adults, young people and children with confirmed COVID-19.
Notes:
|
| Field | Content |
Exclusion: Adults, young people and children with other respiratory infections unrelated to coronavirus |
|
| Intervention | Vitamin D supplementation (all dosages, formulations and routes of administration)
Note: Vitamin D supplementation as an adjunctive treatment will be included if other treatments are balanced out in the control arm |
| Comparator | Placebo or standard care or no treatment
Note: for vitamin D supplementation as an adjunctive treatment, the comparator will be the index treatment(s). For example: Vitamin D + Treatment X versus Treatment X |
| Outcomes | Primary outcomes:
Secondary outcomes:
|
This evidence review was developed using the methods and process described in Developing NICE guidelines: the manual. Methods specific to this review question are described in the review protocol in appendix A.
Risk of bias in randomised controlled trials was assessed by using the Cochrane risk of bias tool (2.0). Results of the risk of bias assessments can be found alongside the evidence table for each study (Appendix D.1.1). Grading of Recommendations Assessment, Development and Evaluation (GRADE) was used to present results and to evaluate the quality of evidence by outcomes (see Appendix E). GRADE assessment domains include risk of bias, inconsistency, indirectness, and imprecision. Outcomes start at High, for example, for a randomised controlled trial, and can be marked down 1 or 2 levels for each domain through to Moderate, Low and Very Low evidence. Each of the evidence quality ratings are explained below:
High - Further research is very unlikely to change our confidence in the estimate of effect.
Moderate - Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low - Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low - Any estimate of effect is very uncertain.
Declarations of interest of the expert panel members who developed the guideline recommendations were recorded according to NICE’s conflicts of interest policy.
Included studies
One search was conducted for all 3 reviews questions, which returned 640 results. For review question 1, after screening, 3 were considered at full text and 1 study was included.
Excluded studies
Three studies were screened at full text; the 2 excluded studies can be found in Appendix G.
Summary of studies included in the effectiveness and prognostic evidence
Table 2 Summary of studies included in the evidence review
| Study | Population | Intervention | Analysis | Outcomes |
| Entrenas Castillo 2020
RCT
Spain |
N=76 admitted with confirmed COVID-19
randomised in a 2:1 ratio into intervention and comparator arms. n=50 in intervention arm, received calcifediol treatment plus standard care n=26 received standard care only |
Calcifediol (0.532 mg) on admission, then
0.266 mg on days 3 and 7, then weekly until discharge. |
Univariate and multivariable logistic regressions were used to estimate the probability of admission to intensive care unit (ICU).
Mortality was reported as number of event counts. |
ICU admission COVID-19
mortality |
See appendix D.1 for full evidence tables.
Summary of the effectiveness evidence
One study reported 2 outcomes concerning people who had been admitted to hospital with COVID-19: admission to intensive care and mortality. Evidence quality was graded as very
low because of a very serious risk of bias and the low number of participants in the study (n=76).
The number of people who were admitted to ICU in the calcifediol arm (n=50) was 1 (2%) and in the standard care only arm (n=26) was 13 (50%). An adjusted multivariable model reported that people who received calcifediol treatment plus standard care were less likely to be admitted to intensive care than people who received standard care only, OR 0.03 (95% CI
0.003 to 0.25) [adjusted for hypertension and diabetes as prevalence differed between arms].
Mortality was reported as number of deaths. There were 0 deaths in the calcifediol treatment plus standard care arm and 2 deaths in the standard care only arm, (OR 0.097, 95%CI 0.004 to 2.099).
See appendix E for full GRADE tables.
Economic evidence
Economic evidence was not considered for this review.
What is the clinical effectiveness and safety of vitamin D supplementation for the prevention of SARS CoV2 infection (and subsequent COVID-19) in adults, young people and children?
Introduction
The current COVID-19 pandemic is caused by transmission of SARS-CoV-2 between people. Lowering the infection rate is important in keeping hospital admissions manageable and preventing death and illness. A range of different approaches have been taken for lowering the infection rate. One option that has been suggested is vitamin D supplementation because it has been hypothesised to be involved in the body’s immune response and could mitigate the inflammatory response. This review aims to assess if vitamin D supplementation can prevent SARS CoV2 infection and subsequent COVID-19.
Table 3 PICO table for review question 2
| Field | Content |
| Population | Inclusion: Adults, young people and children who are not infected with SARS CoV2 (or SAR CoV1 and MERS).
Notes: Studies on specific sub-populations such as those identified as ‘vulnerable’, ‘extremely vulnerable’ or other comorbidities will be included.
|
| Field | Content |
Exclusion: Adults, young people and children who already have contracted SARS CoV2 at the point of study entry. |
|
| Intervention | Vitamin D supplementation (all dosages, formulations and routes of administration)
Note: Vitamin D supplementation as an adjunctive treatment will be included if other treatments are balanced out in the control arm |
| Comparator | Placebo or standard care or no treatment
Note: for vitamin D supplementation as an adjunctive treatment, the comparator will be the index treatment(s). For example: Vitamin D + Treatment X versus Treatment X |
| Outcomes | Primary outcomes:
Secondary outcomes:
|
This evidence review was developed using the methods and process described in Developing NICE guidelines: the manual. Methods specific to this review question are described in the review protocol in appendix A.
Risk of bias in randomised controlled trials was assessed by using the Cochrane risk of bias tool (2.0). Results of the risk of bias assessments can be found alongside the evidence table
for each study (Appendix D). Grading of Recommendations Assessment, Development and Evaluation (GRADE) was used to present results and to evaluate the quality of evidence by outcomes (see Appendix E). GRADE assessment domains include risk of bias, inconsistency, indirectness, and imprecision. Outcomes start at High, for example, from a randomised controlled trial, can be marked down 1 or 2 levels for each domain through to Moderate, Low and Very Low evidence. Each of the evidence quality ratings are explained below:
High - Further research is very unlikely to change our confidence in the estimate of effect.
Moderate - Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low - Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low - Any estimate of effect is very uncertain.
Declarations of interest of the expert panel members who developed the guideline recommendations were recorded according to NICE’s conflicts of interest policy.
One search was conducted for all 3 evidence reviews, which returned 640 results. For review question 2, after screening, 2 studies were at full text considered and no studies were included.
Excluded studies
Two studies were considered at full text. All were excluded and can be found with reasons in Appendix G.
Summary of the effectiveness evidence
No evidence relevant to the protocol was found for this question.
Economic evidence was not considered for this review.
Is vitamin D status independently associated with susceptibility to developing COVID-19, severity of COVID-19, and poorer outcomes from COVID-19 in adults, young people and children?
The COVID-19 pandemic has affected countries differently. Some have hypothesised that the differences are in part caused by people’s vitamin D status. It has also been suggested that the higher incidence of SARS CoV2 infection in older people and ethnic minorities could be partly explained by lower serum vitamin D, which is more common in these groups. The aim of this review is to assess if there is an association between vitamin D status and incidence of SARS CoV2 infection, poorer outcomes, and COVID-19 severity.
Table 4 protocol summary for review question 3
| Field | Content |
| Population | Inclusion: Adults, young people and children with confirmed COVID-19.
Note: Other coronavirus such as SARS CoV1 (symptomatic) and MERS will be included as indirect evidence
Exclusion: Adults, young people and children with other respiratory infections unrelated to coronavirus. |
| Dependent variables | Vitamin D status (as measured by serum/plasma 25- hydroxyvitamin D [25(OH)D concentration]).
Deficiency is defined as serum 25(OH)D concentration <25nmol/L |
| Independent variables, confounders |
|
| Outcomes |
|
This evidence review was developed using the methods and process described in Developing NICE guidelines: the manual. Methods specific to this review question are described in the review protocol in appendix A.
Studies included in this review were anticipated to be cohort, case-control and cross- sectional designs analysing associations between vitamin D using regression adjusting for confounding variables. Regression analyses reveal relationships among variables and adjustment attempts to eliminate the effect of other factors to assess if the variable in question is independently associated with an outcome. In the example of vitamin D, older people have low vitamin D and poorer COVID-19 outcomes. A regression analysis without any adjustment will not infer whether age or vitamin D, or both, are associated with poorer COVID-19 outcomes. A regression model adjusted for age will allow us to see associations, including vitamin D, that are not due to age. This can be done with numerous measurable variables but does not mean that all factors have been accounted for. Many other factors are associated both with COVID-19 outcomes and with vitamin D status, for example obesity, ethnicity, diabetes, socioeconomic status, household crowding and urban place of residence. For this reason, only studies that reported multivariable (adjusted) models for outcomes of interest were considered because at least some confounding variables are considered in these models.
However, it is noteworthy that associations demonstrated in an adjusted model do not imply that the relationships are causal. There are other factors that could be influencing the association that were not adjusted for. Association should not be confused with causality.
This is especially important when many variables are studied in a complex public health scenario. In this scenario spurious associations can arise because the large number of factors studies makes it possible that an association could be discovered by chance or because of multicollinearity, explored below.
Studies on associations can be used to form the basis for hypothesis testing for causality. Therefore, association studies are usually published to inform future research. It was not expected that many published RCTs would be available for vitamin D as treatment or prevention at this stage of the pandemic. This is because association studies can generate hypotheses and guide timely and costly RCTs. However, there are a number of methodological and statistical considerations that need to be taken into account when interpreting findings from association studies. Firstly, to assess whether there is a confirmed or scientifically proven biological plausibility between the dependent variable and outcome of interest. Caution is required if the hypothesis of the relationship is based on indirect association, that is if A is caused by B, A is similar to C, therefore C must also be caused by B.
Another consideration that needs careful interpretation in any multivariable analysis is the issue of multicollinearity, meaning 2 or more variables are associated with each other. The mediating and moderating effect from other variables (for example confounding variables or other independent variables in the model) need to be explored before concluding the association between the dependent variable and outcome of interest. This is because the model assumption that a change in the independent variable will lead to a change in the dependent variable while all other variables are held constant will be violated. Without these considerations, estimates for this association will be imprecise and small changes to the raw data may generate large changes to the model results.
There are also several statistical principles in multivariable analysis that are either not commonly conducted by researchers or poorly reported. For example, the lack of goodness- of-fit testing, lack of residual analysis, the family-wise error rate (due to multiple comparisons in a multivariable model) is not adjusted, violation of assumption of normality, lack of cross- validation, and others. All data was extracted from the studies in relation to these principles, where available.
All the elements above need careful assessment in order to decide the quality of the study and whether the concluded association is valid. As the studies that were included were looking at association, the risk of bias checklist used was the QUIPS checklist. This checklist has domains that considers the above issues. A modified version of Grading of
Recommendations Assessment, Development and Evaluation (GRADE) was used to present results and to evaluate the quality of evidence by outcomes (see Appendix E). Only outcomes that had been adjusted for covariates were presented, all conducted through multivariable analysis. When unadjusted (univariable) values were presented, they were included in the modified GRADE table for reference. In modified GRADE, the assessment domains include risk of bias, inconsistency, indirectness, and imprecision. For observational studies, outcomes start at Low (for example, cohort study and case control study) can be marked down 1 or 2 levels for each domain through to Very Low evidence. Evidence quality can be raised 1 or levels in specific circumstances: if the effect estimate is large or if the effect shows a dose-response curve. Each of the evidence quality ratings are explained below:
High - Further research is very unlikely to change our confidence in the estimate of effect.
Moderate - Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low - Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low - Any estimate of effect is very uncertain.
Declarations of interest of the expert panel members who developed the guideline recommendations were recorded according to NICE’s conflicts of interest policy.
Included studies
One search was conducted for all 3 evidence reviews, which returned 640 results. For review question 3, after screening, 69 studies were considered at full text and 12 studies were included.
Twelve studies were identified as pre-prints and listed as references but data was not extracted from them for this review as they have not yet been peer reviewed. One extra study in preprint was identified through stakeholder consultation and included in the list of preprints. This made a total of 13 preprint studies.
Excluded studies
Sixty-nine studies were considered at full text. Forty-five were excluded; reasons for exclusion can be found in Appendix G.
Summary of studies included in the effectiveness and prognostic evidence
Table 5 Summary of studies included in the evidence review
| Study | Population | Analysis | Vitamin D status | Dependent variable/ confounders | Outcomes |
| Annweiler 2020
Retrospective quasi- experimental study
France |
N=66
Nursing home residents diagnosed with COVID-19 Residents received chronic vitamin D supplementation with regular maintenance boluses (single oral dose of 80,000 IU vitamin D3 every 2 to 3 months). When residents last received supplementation dictated which group they were in: n=57 received vitamin D bolus |
Associations between predictor variables, such as vitamin D3 supplements, and the likelihood of COVID-19
mortality at a specific time. Comparing time to death between intervention and comparator groups. Associations between bolus vitamin D3 supplements and World Health Organization Ordinal Scale |
Determined by whether the participant had received the vitamin D bolus or not | Dependent variable:
COVID-19 confirmed with RT-PCR Confounders:
|
COVID-19 mortality
World Health Organization Ordinal Scale for Clinical Improvement (OSCI) score for COVID-19 in acute phase |
| Study | Population | Analysis | Vitamin D status | Dependent variable/ confounders | Outcomes |
| within 1 month of or a week after COVID-19 diagnosis
n=9 did not receive vitamin D bolus |
for Clinical Improvement (OSCI) score, taking into account factors that may affect the result. | ||||
| Annweiler 2020a
Retrospective cohort study
France |
N=77 admitted to hospital with COVID-19
n=29 received vitamin D bolus over the preceding year n=16 received a vitamin D supplement after COVID-19 diagnosis n=32 received no vitamin D supplement |
Comparisons between groups for the reported outcomes.
The association between each group and 14- day mortality at a specific time, adjusting for confounders. Comparison of survival between the groups. Association between vitamin D status and severe COVID- 19, adjusted for confounding variables. |
Determined by whether the participant had received the vitamin D bolus or not | Dependent variable:
COVID-19 confirmed with RT-PCR Confounders:
|
14-day COVID-19
mortality OSCI score for COVID- 19 in acute phase |
| Hastie 2020 | N = 348,598 | Association between vitamin | Measurements taken when participants were |
|
SARS-CoV2 infection |
| Study | Population | Analysis | Vitamin D status | Dependent variable/ confounders | Outcomes |
| Retrospective cohort study
UK |
n=348,139 no COVID
n=449 COVID- positive Participants in the UK Biobank study with complete data |
D and SARS- CoV-2 infection explored using multivariable logistic regression adjusted for confounding variables.
Interaction between vitamin D and ethnicity and its association with SARS-CoV-2 infection was explored with multivariable analysis. |
first recruited between 2006 to 2010 | CoV-2 infection as recorded by Public Health England.
Confounders:
|
|
| Hernandez 2020 | N=413
n=216 cases were aged 18 or |
4 models were conducted: | Measurements were taken at hospital admission (cases) or during | Dependent variable:
SARS-CoV-2 infection confirmed by RT-PCR |
|
| Study | Population | Analysis | Vitamin D status | Dependent variable/ confounders | Outcomes |
| Case-control study
Spain |
over admitted to hospital with confirmed COVID-19,
n=19 were taking vitamin D supplements. n=197 controls were recruited from the Camargo study cohort and were sex-matched with the non- vitamin D supplemented cases. |
multivariable general linear model comparing 25(OH)D levels between COVID-19 patients and controls that took into account confounding factors
regression |
recruitment into study (controls) |
|
unit (ICU), requirement for mechanical ventilation, or in- hospital mortality.
|
| Study | Population | Analysis | Vitamin D status | Dependent variable/ confounders | Outcomes |
between 25 (OH)D and disease severity
|
|||||
| Karahan 2020
Case-control study
Turkey |
N=149 COVID-19
patients who were admitted into the hospital with confirmed COVID-19. n=47, moderate COVID-19 n=102, severe- critical COVID- 19 |
Compared patients in the moderate arm to the severe/critical arm; and then compared patients who survived and patients who died.
Pearson’s correlation evaluated the bivariate |
Measurements taken by electrochemiluminescence at admission.
Patients were stratified according to their serum 25(OH) vitamin D levels:
|
Dependent variable:
SARS-CoV-2 infection confirmed by PCR Confounders:
|
In-hospital mortality |
| Study | Population | Analysis | Vitamin D status | Dependent variable/ confounders | Outcomes |
| correlation between the serum 25(OH) vitamin D level and inflammatory marker.
Univariable and multivariable logistic regression analyses were used to determine the independent associates of mortality. |
|||||
| Kaufman 2020
Retrospective cohort study
US |
N=191779
Participant data was collected from a Quest Diagnostics database that processed SARS-CoV-2 tests and matched it to data held on individual’s vitamin D results from the |
Polynomial regression was fitted between vitamin D values and infection.
Multivariable logistic regression was performed adjusting for confounders. |
Measurements taken by electrochemiluminescence or liquid chromatograph/tandem mass spectrometry.
Patients were stratified according to their serum 25(OH) vitamin D levels:
|
SARS-CoV-2 infection confirmed by PCR.
Confounders:
|
SARS-CoV2 infection |
| Study | Population | Analysis | Vitamin D status | Dependent variable/ confounders | Outcomes |
| preceding 12 months. | |||||
| Macaya 2020 Case series Spain | admitted to hospital with COVID-19
n=49 had non- severe COVID- 19 n=31 had severe COVID- 19 |
Association between the composite COVID-19
outcome and vitamin D level was conducted by multivariable logistic regression adjusting for confounders. |
Measured by chemiluminescent immunoassay at admission or within the previous 3 months. | Dependent variable: COVID-19 severity Confounders:
|
Composite COVID-19 outcome: death, admission to ICU, need for higher oxygen flow than provided by a nasal cannula. |
| Meltzer 2020
Retrospective cohort study
US |
N=4313 patients tested for
SARS-CoV-2 infection at the university. |
Association between vitamin D status and SARS-CoV-2
infection multivariable generalised linear model with binomial residuals and log-link function was estimated adjusting for confounders |
Vitamin D measurements were from the preceding 12 months. To account for changes to vitamin D status, status was estimated by taking into account changes to supplements taken.
Participants were grouped as follows:
|
Dependent variable:
SARS-CoV-2 infection confirmed by PCR. Confounders:
|
SARS-CoV2 infection |
| Study | Population | Analysis | Vitamin D status | Dependent variable/ confounders | Outcomes |
| Merzon 2020
Case-control study
Israel |
N=14,000
People of the Leumit Health Services who were tested for SARS-CoV-2. n=782, COVID- 19 positive n=7025, COVID-19 negative |
Univariate analyses assessed the association between baseline characteristics and COVID-19
infection and hospitalisation. Multivariable analyses assessed the association between 25(OH)D levels and COVID-19 infection and hospitalisation, adjusting for confounders. |
Measured by chemiluminescence assay.
Low vitamin D level was considered <75 nmol/L. |
Dependent variable:
SARS-CoV-2 infection confirmed by RT-PCR Confounders:
|
SARS-CoV2 infection Hospitalisation |
| Radujkovic 2020
Retrospective cohort study
Germany |
N=185
Consecutive symptomatic SARS-CoV2- positive patients admitted to hospital |
Cox regression was used to assess association between vitamin D status and composite endpoint, adjusted for confounders. | Samples taken at admission retrospectively measured by immunoassay.
Low vitamin D level was considered <30 nmol/L |
Dependent variable:
SARS-CoV-2 infection measured by RT-PCR Confounders:
|
Composite endpoint: invasive mechanical ventilation and/or death.
All-cause mortality. |
| Study | Population | Analysis | Vitamin D status | Dependent variable/ confounders | Outcomes |
| Survival analysis was conducted by Kaplan-Meier curves. | |||||
| Raisi- Estabragh 2020
Nested case- control study
UK |
N=4510
People aged 40 to 69 in the UK Biobank study who had taken a SARS-CoV-2 test. n=1326, COVID-19 positive n=3184, COVID-19 negative |
Univariate logistic regression was performed for every variable individually to assess the association between them and SARS- CoV-2 infection.
Multivariable logistic regression models were run to associated groups of variables with COVID-19 infection, one of which included vitamin D levels, adjusted for confounders. |
Measured at central laboratory at time of recruitment, 2006 to 2010. | Dependent variable:
SARS-CoV-2 infection as recorded by Public Health England Confounders:
|
SARS-CoV2 infection |
| Study | Population | Analysis | Vitamin D status | Dependent variable/ confounders | Outcomes |
| Ye 2020
Case-control study
China |
N=142
n=62 cases with COVID-19 treated in hospital n=80 controls recruited from an examination centre with no medical condition, matched to cases by sex and age. |
Unconditional logistic regression assessing association between all measured risk factors and severe/critical disease, and another for the association between all measured risk factors and cases versus controls. | Electrochemiluminescence immunoassay on samples taken at admission.
Patients were stratified according to their serum 25(OH) vitamin D levels:
|
Dependent variable:
SARS-CoV-2 infection diagnosed by RNA from throat swab samples. Confounders:
|
COVID-19 severity SARS-CoV2 infection |
See appendix D.2 for full evidence tables.
Summary of the association evidence
Twelve studies reported on how vitamin D status associated with COVID-19 outcomes. Outcomes included COVID-19 cases, COVID-19 severity including hospitalisation and mechanical ventilation, and mortality. Studies reported vitamin D status differently. Some measured associations between outcomes and linear vitamin D concentrations. Whereas others used cut-offs to define deficiency and assessed the risk of an outcome by comparing people considered vitamin D deficient and vitamin D sufficient. All outcomes were rated as very low quality of evidence. Evidence was downgraded for high risk of bias, indirectness, and imprecision due to the number of factors adjusted and low participant numbers.
Association between vitamin D status and COVID-19 cases
Six studies explored the association between vitamin D status and COVID-19 incidence. Some studies found a negative association between vitamin D status and COVID-19 incidence. Whether studies measured vitamin D as a linear scale or compared people with vitamin D deficiency to people without did not have an impact on whether a significant association was found.
Three studies assessed the linear association between vitamin D concentration measured and the risk of COVID-19 cases and 1 found a significant association, OR 0.984 (95% CI 0.983 to 0.986) N=191,779 (Kaufman 2020). Conversely, another study did not find an association, OR 1.00 (0.998 to 1.01) N=349,017 (Hastie 2020). This study also demonstrated that ethnicity did not impact on the association between vitamin D status and COVID-19 cases. However, it is difficult to provide a definite effect estimate due to lack of power in this analysis. Raisi-Esrabragh 2020 used the same UK Biobank database for their analysis as Hastie 2020. There was no association found between vitamin D status COVID-19 cases, OR 1.00 (1.00 to 1.00), N=4510. However, the UK Biobank study measured serum vitamin D during initial recruitment in 2006-2010. Serum vitamin D fluctuates and a lag of at least 10 years means the recorded concentrations may not be reflective of current concentrations.
Hastie 2020 and Kaufman 2020 used data from everyone that was available in their databases, but Raisi-Esrabragh 2020 only used data from people who had taken a SARS CoV2 test. The difference between the results of the studies that used the UK Biobank data and the Kaufman study may have arisen between the two populations that are included in the analyses. The UK Biobank data has a lower proportion of ethnic minorities and includes a more people from higher socioeconomic groups.
Within the same study, Hastie 2020 also assessed association between receiving a positive COVID-19 test in 2 groups, people who were vitamin D deficient (<25nmol/L) and people who were insufficient (<50nmol/L) were compared with people with serum concentrations above the thresholds. Both analyses showed no difference in COVID-19 cases between people above and below the thresholds, <25nmol/L OR 0.92 (95% CI 0.71 to 1.21),
<50nmol/L OR 0.88 (95% CI 0.72 to 1.08). Meltzer 2020 used the same threshold for deficiency, <50nmol/L and demonstrated an association between vitamin D deficiency and COVID-19 cases, OR 1.77 (95% CI 1.12 to 2.81). However, this study did not adjust for demographic factors, such as sex, gender and ethnicity, that are known to have an impact on COVID-19 case rates. Merzon 2020 did control for demographic variables and found that people with suboptimal serum vitamin D (<75nmol/L) were more likely to contract COVID-19 than people above the threshold OR 1.5 (95% CI 1.13 to 1.98).
Effects are presented as OR/HR (95% CI) in Table 6 unless otherwise stated.
Table 6 Summary of evidence for the association between vitamin D status and COVID-19 cases
| Vitamin D | Association | Study | N | Adjusted for | Quality |
| Vitamin D level (nmol/L) | OR 1.00
(0.998 to 1.01) |
Hastie 2020 | Cases n=449
Control n=348,59 8 |
Ethnicity, sex, month of assessment, Townsend deprivation quintile, household income, self-reported health rating, smoking status, body mass index (BMI) category, age at assessment, diabetes, SBP, DBP, and long- standing illness, disability or infirmity | Very low |
| Vitamin D level (nmol/L) | OR 1 (1 to 1) | Raisi- Esrabragh 2020 | Cases n=1326
Control n=3184 |
Sex, age and ethnicity | Very low |
| Vitamin D level (ng/ml) | MD: -9.3; p<0.001 | Hernandez 2020 | Cases n=197
Control n=197 |
Age, smoking, hypertension, diabetes mellitus, history of cardiovascular events, immunosuppressio n, body mass index (BMI), serum corrected calcium, glomerular filtration rate and the month of vitamin D determination | Very low |
| Vitamin D level (ng/ml) | OR 0.984
(0.983 to 0.986) |
Kaufman 2020 | Cohort N=191,77 9 | Gender, age, latitudes, ethnicity | Very low |
| Vitamin D level (nmol/L) by
ethnicity |
OR 0.90
(0.66 to 1.23) |
Hastie 2020 | Cases n=449
Control n=348,59 8 |
Ethnicity, sex, month of assessment, Townsend deprivation quintile, household income, self-reported health rating, smoking status, BMI category, age at assessment, diabetes, SBP, DBP, and long- standing illness, disability or infirmity | Very low |
| Vitamin D deficiency
<25 nmol/L |
OR 0.92
(0.71 to 1.21) |
Hastie 2020 | Cases n=449
Control n=348,59 8 |
Ethnicity, sex, month of assessment, Townsend deprivation quintile, household income, self-reported health rating, smoking status, BMI category, age at assessment, diabetes, SBP, DBP, and long- standing illness, disability or infirmity | Very low |
| Vitamin D insufficiency
<50 nmol/L |
OR 0.88
(0.72 to 1.08) |
Hastie 2020 | Cases n=449
Control n=348,59 8 |
Ethnicity, sex, month of assessment, Townsend deprivation quintile, household income, self-reported health rating, smoking status, BMI category, age at assessment, diabetes, SBP, DBP, and long- standing illness, disability or infirmity | Very low |
| Vitamin D deficient
<50 nmol/L |
OR 1.77
(1.12 to 2.81) |
Meltzer 2020 | Positive n=71
Negative n=418 |
Hypertension, diabetes, chronic pulmonary disease, pulmonary circulation disorders, depression, immunosuppressio n, liver disease, and chronic kidney disease. | Very low |
| Vitamin D suboptimal
<75 nmol/L |
OR 1.5 (1.13
to 1.98) |
Merzon 2020 | Cases n=782
Control n=7025 |
Age, gender, ethnicity, smoking, depression/anxiety, schizophrenia, dementia, diabetes, hypertension, cardiovascular disease, chronic lung disease,
obesity, BMI and |
Very low |
| socioeconomic status. |
Association between vitamin D status and COVID-19 severity
Seven studies reported on the association between vitamin D status and COVID-19 severity. Measures for severity included the WHO Ordinal Scale for Clinical Improvement (OSCI) score, composite outcomes, mortality, and hospitalisation.
Two studies report on composite outcomes. Hernandez 2020 did not find a significant association between vitamin D level in nmol/L and the composite outcome of admission to the intensive care unit (ICU), requirement for mechanical ventilation, or in-hospital mortality, OR 1.13 (95% CI 0.27 to 4.77), n=197. Macaya 2020 also did not find an association between vitamin D deficiency (<50nmol/L) and the composite outcome death, admission to ICU, and/or need for higher oxygen flow than that provided by a nasal cannula, OR 3.2 (95% CI 0.99 to 11.4). Radujkovic 2020 did find a significant association between suboptimal vitamin D (<30 nmol/L) and the composite outcome mechanical ventilation and death, HR
6.12 (95% CI 2.79 to 13.42), n=185. The difference between these studies may lie in which factors were adjusted for in the multivariable analysis.
Results associating vitamin D status with COVID-19 severity scores were mixed. Ye 2020 used the guidelines of the National Health Commission of China to define severe/critical cases. The study found an association between vitamin D deficiency and severe/critical COVID-19, OR 15.18 (95% CI 1.23 to 187.45). Annweiler 2020a used OSCI scores ≥5 to define severe COVID-19 and included 3 groups: people classed as vitamin D sufficient if they received supplements for a year before COVID-19 diagnosis, people who received vitamin D bolus when diagnosed with COVID-19, and people who had not received any supplementation. Compared with no supplementation, supplementation for a year was significantly negatively associated with having severe COVID-19, OR 0.08 (95% CI 0.01 to 0.81), but there was no difference when only receiving a bolus when diagnosed, OR 0.46 (95% CI 0.07 to 2.85).
Four studies reported mortality as a standalone outcome. 3 studies found that a higher measured vitamin D status (Karahan 2020, linear vitamin D measurement) or supplementation before diagnosis (Annweiler 2020, within a month before or up to a week after diagnosis; Annweiler 2020a, supplemented for a year before diagnosis) were negatively associated with death post COVID-19 diagnosis, OR 0.92 (95% CI 0.88 to 0.98), HR 0.11
(95% CI 0.03 to 0.48) and HR 0.07 (95% CI 0.01 to 0.61, respectively. Radujkovic 2020 also found that suboptimal serum vitamin D (<30nmol/L) was associated with higher mortality rate, HR 14.73 (95% CI 4.16 to 52.19). However, receiving a vitamin D bolus when diagnosed with COVID-19 was as associated with death as no supplementation (HR 0.37 (95% CI 0.06 to 2.21), Annweiler 2020a).
Hospitalisation for COVID-19 symptoms was reported as a standalone outcome in 1 study (Merzon 2020). Suboptimal vitamin D levels (<75 nmol/L) were as associated with hospitalisation as higher vitamin D, OR 1.95 (95% CI 0.99 to 4.78).
Effects are presented as OR/HR (95% CI) in table 7.
.
Table 7 Summary of evidence for the association between vitamin D status and COVID-19 severity
| Outcome | Effect | Study | Vitamin D status | Number of people in study | Adjusted for | Quality |
| Composite endpoint: (admission to the intensive care unit (ICU), requirement for mechanical ventilation, or in-hospital mortality) | OR 1.13
(0.27 to 4.77) |
Hernandez 2020 | Vitamin D level, nmol/L | 197 | Smoking, hypertension, diabetes mellitus, history of cardiovascular events, immunosuppression, body mass index, serum corrected calcium, glomerular filtration rate and the month of vitamin D determination. | Very low |
| Composite endpoint including mechanical ventilation and death | HR 6.12
(2.79 to 13.42) |
Radujkovic 2020 | Vitamin D suboptimal
<30 nmol/L |
185 | Age, gender and comorbidities. Includes values for whole cohort not for inpatients only | Very low |
| Severe/critical cases defined as having one of the following: breathing rate
>30/min, O2 saturation ≤93% at rest, PaO2/FiO2 ≤ mmHg or lung imaging shows significant progression, respiratory failure (PaO2 <60mmHg), shock, organ failures that requires ICU care |
OR 15.18
(1.23 to 187.45) |
Ye 2020 | Vitamin D deficiency
<50 nmol/L |
60 | Age, sex and comorbidities | Very low |
| Mortality | OR 0.92 | Karahan 2020 | Vitamin D level (ng/ml), | 149 | Age, smoking, hyperlipidaemia, diabetes
mellitus, chronic kidney disease, chronic atrial fibrillation, congestive heart failure, |
Very low |
| Outcome | Effect | Study | Vitamin D status | Number of people in study | Adjusted for | Quality |
| (0.88 to
0.98) |
continuous measure | acute kidney injury, estimated glomerular filtration rate, haemoglobin, neutrophil count | ||||
| Composite outcome defining severity of COVID-19 included death, admission to ICU, and/or need for higher oxygen flow than that provided by a nasal cannula | OR 3.2
(0.99 to 11.4) |
Macaya 2020 | Vitamin D deficiency
<50 nmol/L |
80 | Age, gender, obesity, severe chronic kidney (CKD) disease, cardiac disease | Very low |
| Mortality, follow-up to 2 months | HR 0.11
(0.03 to 0.48) |
Annweiler 2020 | Vitamin D sufficient if received a vitamin D booster supplement within a month of COVID-19
diagnosis |
66 | Age, gender, number of drugs daily taken, functional abilities, nutritional status,
COVID-19 treatment with corticosteroids and/or hydroxychloroquine and/or dedicated antibiotics, and hospitalisation for COVID-19 |
Very low |
| COVID-19 mortality, 14-day follow-up | HR 0.07
(0.01 to 0.61) |
Annweiler 2020a | Vitamin D sufficient if received vitamin D supplements for a year before | 77 | Age, gender, GIR score, severe undernutrition, history of cancer, history of hypertension, history of cardiomyopathy, glycated haemoglobin, number of acute health problems, use of antibiotics, use of systemic corticosteroids, use of treatments of respiratory disorders | Very low |
| Outcome | Effect | Study | Vitamin D status | Number of people in study | Adjusted for | Quality |
| COVID-19
diagnosis |
||||||
| COVID-19 mortality, 14-day follow-up | HR 0.37
(0.06 to 2.21) |
Annweiler 2020a | Vitamin D sufficient if received vitamin D supplement when diagnosed with COVID- 19 | 77 | Age, gender, GIR score, severe undernutrition, history of cancer, history of hypertension, history of cardiomyopathy, glycated haemoglobin, number of acute health problems, use of antibiotics, use of systemic corticosteroids, use of treatments of respiratory disorders | Very low |
| Severe COVID-19 - World Health Organization Ordinal Scale for Clinical Improvement (OSCI) score ≥ 5 | OR 0.08
(0.01 to 0.81) |
Annweiler 2020 | Vitamin D sufficient if received vitamin D supplements for a year before COVID-19
diagnosis |
77 | Age, gender, GIR score, severe undernutrition, history of cancer, history of hypertension, history of cardiomyopathy, glycated haemoglobin, number of acute health problems, use of antibiotics, use of systemic corticosteroids, use of treatments of respiratory disorders | Very low |
| Severe COVID-19 - OSCI
score ≥ 5 |
OR 0.46
(0.07 to 2.85) |
Annweiler 2020a | Vitamin D sufficient if received vitamin D supplement
when |
77 | Age, gender, GIR score, severe undernutrition, history of cancer, history of hypertension, history of cardiomyopathy, glycated haemoglobin, number of acute health problems, use of antibiotics, use of | Very low |
| Outcome | Effect | Study | Vitamin D status | Number of people in study | Adjusted for | Quality |
| diagnosed with COVID- 19 | systemic corticosteroids, use of treatments of respiratory disorders | |||||
| Hospitalisation for COVID-19 symptoms | OR 1.95
(0.99 to 4.78) |
Merzon 2020 | Vitamin D suboptimal
<75 nmol/L |
7,807 | Multiple conditions and demographic variables | Very low |
| Mortality | HR 14.73
(4.16 to 52.19) |
Radujkovic 2020 | Vitamin D suboptimal (<30 nmol/L) | 185 | Age, gender, comorbidity | Very low |
See appendix E for full GRADE tables.
No economic evidence was considered for this review.
References - included studies
Entrenas Castillo, Marta, Entrenas Costa, Luis Manuel, Vaquero Barrios, José Manuel et al. (2020) “Effect of calcifediol treatment and best available therapy versus best available therapy on intensive care unit admission and mortality among patients hospitalized for COVID-19: A pilot randomized clinical study”;. J Steroid Biochem Mol Biol 203: 105751- 105751
Annweiler, Cedric, Hanotte, Berangere, de l’Eprevier, Claire Grandin et al. (2020) Vitamin D and survival in COVID-19 patients: A quasi-experimental study. The Journal of steroid biochemistry and molecular biology: 105771
Annweiler, Gaelle, Corvaisier, Mathieu, Gautier, Jennifer et al. (2020) Vitamin D Supplementation Associated to Better Survival in Hospitalized Frail Elderly COVID-19 Patients: The GERIA-COVID Quasi-Experimental Study. Nutrients 12(11)
Hastie, Claire E, Mackay, Daniel F, Ho, Frederick et al. (2020) Vitamin D concentrations and COVID-19 infection in UK Biobank. Diabetes & metabolic syndrome 14(4): 561-565
Hernandez, JL, Nan, D, Fernandez-Ayala, M et al. (2020) Vitamin D Status in Hospitalized Patients With SARS-CoV-2 Infection. The Journal of Clinical Endocrinology & Metabolism
Karahan, S. and Katkat, F. (2020) Impact of Serum 25(OH) Vitamin D Level on Mortality in Patients with COVID-19 in Turkey. Journal of Nutrition, Health and Aging
Kaufman, Harvey W, Niles, Justin K, Kroll, Martin H et al. (2020) SARS-CoV-2 positivity rates associated with circulating 25-hydroxyvitamin D levels. PloS one 15(9): e0239252
Macaya, Fernando, Espejo Paeres, Carolina, Valls, Adrian et al. (2020) Interaction between age and vitamin D deficiency in severe COVID-19 infection. Nutricion hospitalaria 37(5): 1039-1042
Meltzer, David O, Best, Thomas J, Zhang, Hui et al. (2020) Association of Vitamin D Status and Other Clinical Characteristics With COVID-19 Test Results. JAMA network open 3(9): e2019722
Merzon, Eugene, Tworowski, Dmitry, Gorohovski, Alessandro et al. (2020) Low plasma 25(OH) vitamin D level is associated with increased risk of COVID-19 infection: an Israeli population-based study. The FEBS journal
Radujkovic, Aleksandar, Hippchen, Theresa, Tiwari-Heckler, Shilpa et al. (2020) Vitamin D Deficiency and Outcome of COVID-19 Patients. Nutrients 12(9)
Raisi-Estabragh, Zahra, McCracken, Celeste, Bethell, Mae S et al. (2020) Greater risk of severe COVID-19 in Black, Asian and Minority Ethnic populations is not explained by cardiometabolic, socioeconomic or behavioural factors, or by 25(OH)-vitamin D status: study of 1326 cases from the UK Biobank. Journal of public health (Oxford, England) 42(3): 451- 460
Ye, Kun, Tang, Fen, Liao, Xin et al. (2020) Does Serum Vitamin D Level Affect COVID-19 Infection and Its Severity?-A Case-Control Study. Journal of the American College of Nutrition: 1-8
Other
Hastie, C.E., Mackay, D.F., Ho, F. et al. (2020) Corrigendum to “Vitamin D concentrations and COVID-19 infection in UK Biobank” [Diabetes Metabol Syndr: Clin Res Rev 2020 14 (4) 561-5] (Diabetes & Metabolic Syndrome: Clinical Research & Reviews (2020) 14(4) (561- 565), (S1871402120301156), (10.1016/j.dsx.2020.04.050)). Diabetes and Metabolic Syndrome: Clinical Research and Reviews 14(5): 1315-1316
Pre-prints
Abu Z M Dayem, Ullah, Lavanya, Sivapalan, Claude, Chelala et al. COVID-19 in patients with hepatobiliary and pancreatic diseases in East London: A single-centre cohort study.
Aduragbemi A, Faniyi, Sebastian T, Lugg, Sian E, Faustini et al. Vitamin D status and seroconversion for COVID-19 in UK healthcare workers who isolated for COVID-19 like symptoms during the 2020 pandemic.
Ariel, Israel, Assi Albert, Cicurel, Ilan, Feldhamer et al. The link between vitamin D deficiency and Covid-19 in a large population.
Chang, Timothy S, Ding, Yi, Freund, Malika K et al. (2020) Prior diagnoses and medications as risk factors for COVID-19 in a Los Angeles Health System. medRxiv : the preprint server for health sciences
Claire E, Hastie; Jill P, Pell; Naveed, Sattar Short Communication: Vitamin D and COVID-19 infection and mortality in UK Biobank.
Grigorios, Panagiotou, Su Ann, Tee, Yasir, Ihsan et al. Low serum 25-hydroxyvitamin D (25D) levels in patients hospitalised with COVID-19 are associated with greater disease severity: results of a local audit of practice.
Isaac Z, Pugach and Sofya, Pugach Strong Correlation Between Prevalence of Severe Vitamin D Deficiency and Population Mortality Rate from COVID-19 in Europe.
Jie, Chen, Lixia, Xie, Xing, Yuan et al. Low serum vitamin D level and COVID-19 infection and outcomes, a multivariate meta-analysis.
Li, Mengyuan, Zhang, Zhilan, Cao, Wenxiu et al. (2020) Identifying novel factors associated with COVID-19 transmission and fatality using the machine learning approach. The Science of the total environment: 142810
Li X, van Geffen J, van Weele M et al. (2020) Genetically-predicted vitamin D status, ambient UVB during the pandemic and COVID-19 risk in UK Biobank: Mendelian Randomisation study.
Mendy, Angelico, Apewokin, Senu, Wells, Anjanette A et al. (2020) Factors Associated with Hospitalization and Disease Severity in a Racially and Ethnically Diverse Population of COVID-19 Patients. medRxiv: the preprint server for health sciences
Panagiotou, G., Tee, S.A., Ihsan, Y. et al. (2020) Original publication: Low serum 25- hydroxyvitamin D (25[OH]D) levels in patients hospitalized with COVID-19 are associated with greater disease severity. Clinical Endocrinology
Raharusun, Prabowo, Priambada, Sadiah, Budiarti, Cahni et al. (2020) Patterns of COVID-19 Mortality and Vitamin D: An Indonesian Study.
Review question 1
Table 1 Review protocol for review question 1
| Field | Content |
| PROSPERO registration number | Not registered. |
| Review title | Vitamin D supplementation for treating COVID-19 |
| Review question | What is the clinical effectiveness and safety of vitamin D supplementation for the treatment of COVID-19 in adults, young people and children? |
| Objective | To investigate the clinical efficacy, effectiveness and safety of vitamin D supplementation as a treatment for people with confirmed COVID-19. |
| Searches | The following databases will be searched:
Searches will be restricted by:
|
| Field | Content |
Other searches:
There will be no re-run of searches.
The full search strategies for MEDLINE, Embase and CENTRAL database will be published in the final review. |
|
| Condition or domain being studied | COVID-19 |
| Population | Inclusion: Adults, young people and children with confirmed COVID-19. Notes:
included
Exclusion: Adults, young people and children with other respiratory infections unrelated to coronavirus. |
| Intervention | Vitamin D supplementation (all dosages, formulations and routes of administration).
Note: Vitamin D supplementation as an adjunctive treatment will be included if other treatments are balanced out in the control arm. |
| Field | Content |
| Comparator | Placebo or standard care or no treatment
Note: for vitamin D supplementation as an adjunctive treatment, the comparator will be the index treatment(s). For example: Vitamin D + Treatment X versus Treatment X. |
| Types of study to be included | Inclusion:
If there is insufficient RCT and CCT, prospective and retrospective cohort studies with control arm will be included.
Exclusion:
|
| Other exclusion criteria |
|
| Field | Content |
Decisions based on data from unknown authenticity may be harmful to patients. |
|
| Context | It has been hypothesised that vitamin D may have a role in the body’s immune response to respiratory viruses. The 2 major forms of vitamin D, vitamin D3 (colecalciferol) and vitamin D2 (ergocalciferol), are licensed for the prevention and treatment of vitamin D deficiency. Vitamin D supplements are not specifically licensed for preventing or treating any infection, including the novel coronavirus infection that causes COVID-19. |
| Primary outcomes (critical outcomes) |
|
| Secondary outcomes (important outcomes) |
|
| Data extraction (selection and coding) | All references identified by the searches and from other sources will be uploaded into EPPI reviewer and de-duplicated. 10% of the abstracts will be reviewed by two reviewers, with any disagreements resolved by discussion or, if necessary, a third independent reviewer.
The full text of potentially eligible studies will be retrieved and will be assessed in line with the criteria outlined above. A standardised form will be used to extract |
| Field | Content |
| data (for example, baseline vitamin D status, dosage) from studies (see Developing NICE guidelines: the manual, Appendix L). | |
| Risk of bias (quality) assessment | Risk of bias will be assessed using the appropriate checklist as described in Developing NICE guidelines: the manual.
For systematic review, ROBIS will be used For RCT and CCT, Cochrane RoB 2 tool will be used For prospective and retrospective cohort study, Cochrane ROBINS-I will be used. |
| Strategy for data synthesis | Where appropriate, pairwise meta-analysis will be conducted based on Cochrane Handbook for Systematic Reviews of Interventions (version 6.1, 2020)
|
| Analysis of sub-groups | Subgroup effects will be explored for:
|
| Field | Content |
|
|
| Type and method of review | Intervention |
| Language | English |
| Country | Review conducted by NICE, England |
| Anticipated or actual start date | 26/10/2020 |
| Anticipated completion date | 23/11/2020 |
| Named contact | 5a. Named contact
Toni Tan Catrin Austin
5b. Organisational affiliation of the review National Institute for Health and Care Excellence (NICE) and Public Health England (PHE). |
| Review team members | From NICE:
|
| Funding sources/sponsor | This systematic review is being completed by NICE which receives funding from DH&SC, NHSE and PHE. |
| Field | Content |
| Conflicts of interest | All guideline committee members and anyone who has direct input into NICE guidelines (including the evidence review team and expert witnesses) must declare any potential conflicts of interest in line with NICE’s code of practice for declaring and dealing with conflicts of interest. Any relevant interests, or changes to interests, will also be declared publicly at the start of each guideline committee meeting. Before each meeting, any potential conflicts of interest will be considered by the guideline committee Chair and a senior member of the development team. Any decisions to exclude a person from all or part of a meeting will be documented. Any changes to a member’s declaration of interests will be recorded in the minutes of the meeting. Declarations of interests will be published with the final guideline. |
| Collaborators | Development of this systematic review will be overseen by an advisory committee who will use the review to inform the development of evidence-based recommendations in line with section 3 of appendix L of Developing NICE guidelines: the manual. Members of the guideline committee are available on the NICE website. |
| Other registration details | N/A |
| Reference/URL for published protocol | N/A |
| Dissemination plans | NICE may use a range of different methods to raise awareness of the guideline. These include standard approaches such as:
|
| Keywords | Vitamin D, COVID-19 |
| Details of existing review of same topic by same authors | N/A |
| Additional information | None |
| Field | Content |
| Details of final publication | www.nice.org.uk |
Table 2 Review protocol for review question 2
| Field | Content |
| PROSPERO registration number | Not registered |
| Review title | Vitamin D supplementation for preventing SARS CoV 2 infection (and subsequent COVID-19) |
| Review question | What is the clinical effectiveness and safety of vitamin D supplementation for the prevention of SARS CoV2 infection (and subsequent COVID-19) in adults, young people and children? |
| Objective | To investigate the clinical efficacy, effectiveness and safety of vitamin D supplementation to prevent SARS CoV2 infection (and subsequent COVID-19). |
| Searches | The following databases will be searched:
Searches will be restricted by:
|
| Field | Content |
Other searches:
There will be no re-run of searches.
The full search strategies for MEDLINE, Embase and CENTRAL database will be published in the final review. |
|
| Condition or domain being studied | COVID-19 |
| Population | Inclusion:
Notes: Studies on specific sub-populations such as those identified as ‘vulnerable’, ‘extremely vulnerable’ or other comorbidities will be included. Subgroups will be explored, see (section 17). Exclusion: Adults, young people and children who already have contracted SARS CoV2 at the point of study entry. |
| Intervention | Vitamin D supplementation (all dosages, formulations and routes of administration). |
| Field | Content |
| Note: Vitamin D supplementation as a combination preventative strategy with other preventative interventions will be included if other preventative interventions are balanced out in the control arm. | |
| Comparator | Placebo or no preventative intervention
Note: for vitamin D supplementation as a combination preventative strategy with other preventative interventions, the comparator will be the index preventative intervention(s). For example: Vitamin D + Preventative treatment X versus Preventative treatment X. |
| Types of study to be included | Inclusion:
If there is insufficient RCT and CCT, prospective and retrospective cohort studies with control arm will be included.
Exclusion:
|
| Other exclusion criteria |
|
| Field | Content |
Decisions based on data from unknown authenticity may be harmful to patients. |
|
| Context | It has been hypothesised that vitamin D may have a role in the body’s immune response to respiratory viruses. The 2 major forms of vitamin D, vitamin D3 (colecalciferol) and vitamin D2 (ergocalciferol), are licensed for the prevention and treatment of vitamin D deficiency. Vitamin D supplements are not specifically licensed for preventing or treating any infection, including the novel coronavirus infection that causes COVID-19. |
| Primary outcomes (critical outcomes) |
|
| Secondary outcomes (important outcomes) |
|
| Data extraction (selection and coding) | All references identified by the searches and from other sources will be uploaded into EPPI reviewer and de-duplicated. 10% of the abstracts will be reviewed by |
| Field | Content |
| two reviewers, with any disagreements resolved by discussion or, if necessary, a third independent reviewer.
The full text of potentially eligible studies will be retrieved and will be assessed in line with the criteria outlined above. A standardised form will be used to extract data (for example, baseline vitamin D status, dosage) from studies (see Developing NICE guidelines: the manual, Appendix L). |
|
| Risk of bias (quality) assessment | Risk of bias will be assessed using the appropriate checklist as described in Developing NICE guidelines: the manual.
For systematic review, ROBIS will be used For RCT and CCT, Cochrane RoB 2 tool will be used For prospective and retrospective cohort study, Cochrane ROBINS-I will be used. |
| Strategy for data synthesis | Where appropriate, pairwise meta-analysis will be conducted based on Cochrane Handbook for Systematic Reviews of Interventions (version 6.1, 2020)
|
| Analysis of sub-groups | Subgroup effects will be explored for:
|
| Field | Content |
|
|
| Type and method of review | Intervention |
| Language | English |
| Country | Review conducted by NICE, England |
| Anticipated or actual start date | 26/10/2020 |
| Anticipated completion date | 23/11/2020 |
| Named contact | 5a. Named contact
Toni Tan Catrin Austin
5b. Organisational affiliation of the review National Institute for Health and Care Excellence (NICE) and Public Health England (PHE) |
| Review team members | From NICE:
|
| Field | Content |
| Funding sources/sponsor | This systematic review is being completed by NICE which receives funding from DH&SC, NHSE and PHE. |
| Conflicts of interest | All guideline committee members and anyone who has direct input into NICE guidelines (including the evidence review team and expert witnesses) must declare any potential conflicts of interest in line with NICE’s code of practice for declaring and dealing with conflicts of interest. Any relevant interests, or changes to interests, will also be declared publicly at the start of each guideline committee meeting. Before each meeting, any potential conflicts of interest will be considered by the guideline committee Chair and a senior member of the development team. Any decisions to exclude a person from all or part of a meeting will be documented. Any changes to a member’s declaration of interests will be recorded in the minutes of the meeting. Declarations of interests will be published with the final guideline. |
| Collaborators | Development of this systematic review will be overseen by an advisory committee who will use the review to inform the development of evidence-based recommendations in line with section 3 of appendix L of Developing NICE guidelines: the manual. Members of the guideline committee are available on the NICE website. |
| Other registration details | N/A |
| Reference/URL for published protocol | N/A |
| Dissemination plans | NICE may use a range of different methods to raise awareness of the guideline. These include standard approaches such as:
|
| Keywords | Vitamin D, COVID-19. |
| Field | Content |
| Details of existing review of same topic by same authors | N/A |
| Additional information | None |
| Details of final publication | www.nice.org.uk |
Table 3 Review protocol for review question 3
| Field | Content |
| PROSPERO registration number | Not registered |
| Review title | The associations of vitamin D status with COVID-19, severity and poorer outcomes. |
| Review question | Is vitamin D status independently associated with susceptibility to developing COVID-19, severity of COVID-19, and poorer outcomes from COVID-19 in adults, young people and children? |
| Objective | To investigate whether vitamin D status alone is independently associated with higher risk of contracting COVID-19, higher risk of having more severe COVID- 19 infection, and more likely to have poorer outcomes. |
| Searches | The following databases will be searched:
|
| Field | Content |
|
Searches will be restricted by:
Other searches:
There will be no re-run of searches.
The full search strategies for MEDLINE, Embase and CENTRAL database will be published in the final review. |
|
| Condition or domain being studied | COVID-19 |
| Population | Inclusion: Adults, young people and children with confirmed COVID-19.
Note: Other coronavirus such as SARS CoV1 (symptomatic) and MERS will be included as indirect evidence
Exclusion: Adults, young people and children with other respiratory infections unrelated to coronavirus. |
| Dependent variable | Vitamin D status (as measured by serum/plasma 25(OH)D concentration) |
| Independent variables/confounders/effect moderators/effect mediators | Statistical methods including adjustments will be assessed in included studies for the following (but not limited to): |
| Field | Content |
|
|
| Types of study to be included | Inclusion:
Exclusion:
|
| Field | Content |
|
|
| Other exclusion criteria |
Decisions based on data from unknown authenticity may be harmful to patients. |
| Context | It has been hypothesised that vitamin D may have a role in the body’s immune response to respiratory viruses. It has also been hypothesised that vitamin D status could be a contributing risk factor for COVID-19. |
| Outcomes of interest |
|
| Data extraction (selection and coding) | All references identified by the searches and from other sources will be uploaded into EPPI reviewer and de-duplicated. 10% of the abstracts will be reviewed by |
| Field | Content |
| two reviewers, with any disagreements resolved by discussion or, if necessary, a third independent reviewer.
The full text of potentially eligible studies will be retrieved and will be assessed in line with the criteria outlined above. A standardised form will be used to extract data (for example, baseline vitamin D status, dosage) from studies (see Developing NICE guidelines: the manual, Appendix L). |
|
| Risk of bias (quality) assessment | Risk of bias will be assessed using the appropriate checklist as described in Developing NICE guidelines: the manual.
For systematic review, ROBIS will be used For other study designs, QUIPS (for univariate analysis) and PROBAST (for multivariate analysis) checklist will be used. |
| Strategy for data synthesis | Where appropriate, for example, 2 or more studies that have adjusted for exactly the same confounders, moderators or mediators, pairwise meta-analysis will be conducted based on Cochrane Handbook for Systematic Reviews of Interventions (version 6.1, 2020)
|
| Analysis of sub-groups | See section 8 |
| Type and method of review | Intervention |
| Language | English |
| Country | Review conducted by NICE, England |
| Field | Content |
| Anticipated or actual start date | 26/10/2020 |
| Anticipated completion date | 23/11/2020 |
| Named contact | 5a. Named contact
Toni Tan Catrin Austin
5b. Organisational affiliation of the review National Institute for Health and Care Excellence (NICE) and Public Health England (PHE) |
| Review team members | From NICE:
|
| Funding sources/sponsor | This systematic review is being completed by NICE which receives funding from DH&SC, NHSE and PHE. |
| Conflicts of interest | All guideline committee members and anyone who has direct input into NICE guidelines (including the evidence review team and expert witnesses) must declare any potential conflicts of interest in line with NICE’s code of practice for declaring and dealing with conflicts of interest. Any relevant interests, or changes to interests, will also be declared publicly at the start of each guideline committee meeting. Before each meeting, any potential conflicts of interest will be considered by the guideline committee Chair and a senior member of the development team. Any decisions to exclude a person from all or part of a meeting will be documented. Any changes to a member’s declaration of interests will be recorded in the minutes of the meeting. Declarations of interests will be published with the final guideline. |
| Field | Content |
| Collaborators | Development of this systematic review will be overseen by an advisory committee who will use the review to inform the development of evidence-based recommendations in line with section 3 of appendix L of Developing NICE guidelines: the manual. Members of the guideline committee are available on the NICE website. |
| Other registration details | N/A |
| Reference/URL for published protocol | N/A |
| Dissemination plans | NICE may use a range of different methods to raise awareness of the guideline. These include standard approaches such as:
|
| Keywords | Vitamin D, COVID-19. |
| Details of existing review of same topic by same authors | N/A |
| Additional information | None |
| Details of final publication | www.nice.org.uk |
Appendix B - Literature search strategies
Deduplication against the previous review was used to draw out any new references, rather than date limiting in any individual source.
Search looked at 2002 to October 27th 2020
1 exp Vitamin D/ (59709)
2 exp Vitamin D Deficiency/ (28155)
3 ((vitamin* adj5 D*2) or vitaminD*2).af. (99557)
4 (ergocalciferol* or calciferol* or vs041h42xc or dihydrotachysterol* or dihydrotachysterin* or calcamine or 67-96-9 or r5lm3h112r or Hydroxyvitamin D*2 or 25Hydroxyvitamin D*2 or HydroxyvitaminD*2 or 25HydroxyvitaminD*2 or hydroxycalciferol* or 25hydroxycalciferol* or hydroxyergocalciferol* or 25hydroxyergocalciferol* or ercalcidiol or “25(OH)D” or 21343-40-8 or alfacalcidol*).af. (24782)
5 (cholecalciferol* or colecalciferol* or calciol or 67-97-0 or 1c6v77qf41 or hydroxycholecalciferol* or hydroxycolecalciferol* or 25hydroxycholecalciferol* or 25hydroxycolecalciferol* or calcifediol* or calcidiol* or “19356-17-3″ or p6yz13c99q or t0wxw8f54e or dihydroxycholecalciferol* or dihydroxycolecalciferol* or 25dihydroxycholecalciferol* or 25dihydroxycolecalciferol* or dihydroxyvitamin D*2 or 25dihydroxyvitamin* or dihydroxyvitaminD*2 or calcitriol* or 32222-06-3 or 40013-87-4 or 55721-11-4).af. (38720)
6 or/1-5 (120807)
7 exp coronavirus/ (39116)
8 exp Coronavirus Infections/ (42273)
9 ((corona* or corono*) adj1 (virus* or viral* or virinae*)).ti,ab,kw,kf. (2176)
10 (coronavirus* or coronovirus* or coronavirinae* or CoV or HCoV* or Betacoronavirus* or Betacoronovirus*).ti,ab,kw,kf. (49098)
11 (”2019-nCoV*” or 2019nCoV* or “19-nCoV*” or 19nCoV* or nCoV2019* or “nCoV-2019*” or nCoV19* or “nCoV-19*” or “COVID-19*” or COVID19* or “COVID-2019*” or COVID2019* or “HCoV- 19*” or HCoV19* or “HCoV-2019*” or HCoV2019* or “2019 novel*” or Ncov* or “n-cov” or “SARS-CoV- 2*” or “SARSCoV-2*” or “SARSCoV2*” or “SARS-CoV2*” or SARSCov19* or “SARS-Cov19*” or “SARSCov-19*” or “SARS-Cov-19*” or SARSCov2019* or “SARS-Cov2019*” or “SARSCov-2019*” or “SARS-Cov-2019*” or SARS2* or “SARS-2*” or SARScoronavirus2* or “SARS-coronavirus-2*” or “SARScoronavirus 2*” or “SARS coronavirus2*” or SARScoronovirus2* or “SARS-coronovirus-2*” or “SARScoronovirus 2*” or “SARS coronovirus2*” or covid).ti,ab,kw,kf. (65326)
12 (respiratory* adj2 (symptom* or disease* or illness* or condition*) adj5 (Wuhan* or Hubei* or China* or Chinese* or Huanan*)).ti,ab,kw,kf. (301)
13 ((”seafood market*” or “food market*”) adj10 (Wuhan* or Hubei* or China* or Chinese* or Huanan*)).ti,ab,kw,kf. (85)
14 (pneumonia* adj3 (Wuhan* or Hubei* or China* or Chinese* or Huanan*)).ti,ab,kw,kf. (527)
15 ((outbreak* or wildlife* or pandemic* or epidemic*) adj1 (Wuhan* or Hubei* or China* or Chinese* or Huanan*)).ti,ab,kw,kf. (316)
16 Middle East Respiratory Syndrome Coronavirus/ (1371)
17 (”middle east respiratory syndrome*” or “middle eastern respiratory syndrome*” or MERSCoV* or “MERS-CoV*” or MERS).ti,ab,kw,kf. (5948)
18 (”severe acute respiratory syndrome*” or SARS).ti,ab,kw,kf. (32888)
19 (”SARS-CoV-1*” or “SARSCoV-1*” or “SARSCoV1*” or “SARS-CoV1*” or SARSCoV or “SARS- CoV” or SARS1* or “SARS-1*” or SARScoronavirus1* or “SARS-coronavirus-1*” or “SARScoronavirus 1*” or “SARS coronavirus1*” or SARScoronovirus1* or “SARS-coronovirus-1*” or “SARScoronovirus 1*” or “SARS coronovirus1*”).ti,ab,kw,kf. (23577)
20 or/7-19 (95256)
21 6 and 20 (290)
22 21 (290)
23 limit 22 to (english language and yr=”2002 -Current”) (288)
24 animals/ not (humans/ and animals/) (4716381) 25 23 not 24 (283)
2002 to October 27th 2020
1 exp vitamin D/ (144538)
2 vitamin D deficiency/ (30650)
3 ((vitamin* adj5 D*2) or vitaminD*2).af. (152453)
4 (ergocalciferol* or calciferol* or vs041h42xc or dihydrotachysterol* or dihydrotachysterin* or calcamine or 67-96-9 or r5lm3h112r or Hydroxyvitamin D*2 or 25Hydroxyvitamin D*2 or HydroxyvitaminD*2 or 25HydroxyvitaminD*2 or hydroxycalciferol* or 25hydroxycalciferol* or hydroxyergocalciferol* or 25hydroxyergocalciferol* or ercalcidiol or “25(OH)D” or 21343-40-8 or alfacalcidol*).af. (46690)
5 (cholecalciferol* or colecalciferol* or calciol or 67-97-0 or 1c6v77qf41 or hydroxycholecalciferol* or hydroxycolecalciferol* or 25hydroxycholecalciferol* or 25hydroxycolecalciferol* or calcifediol* or calcidiol* or “19356-17-3″ or p6yz13c99q or t0wxw8f54e or dihydroxycholecalciferol* or dihydroxycolecalciferol* or 25dihydroxycholecalciferol* or 25dihydroxycolecalciferol* or dihydroxyvitamin D*2 or 25dihydroxyvitamin* or dihydroxyvitaminD*2 or calcitriol* or 32222-06-3 or 40013-87-4 or 55721-11-4).af. (62783)
6 or/1-5 (184996)
7 exp Coronavirinae/ (20645)
8 exp Coronavirus infection/ (22005)
9 (”coronavirus disease 2019″ or “severe acute respiratory syndrome coronavirus 2″).sh,dj. (59354)
10 ((corona* or corono*) adj1 (virus* or viral* or virinae*)).ti,ab,kw. (1675)
11 (coronavirus* or coronovirus* or coronavirinae* or CoV or HCoV* or Betacoronavirus* or Betacoronovirus*).ti,ab,kw. (49400)
12 (”2019-nCoV*” or 2019nCoV* or “19-nCoV*” or 19nCoV* or nCoV2019* or “nCoV-2019*” or nCoV19* or “nCoV-19*” or “COVID-19*” or COVID19* or “COVID-2019*” or COVID2019* or “HCoV- 19*” or HCoV19* or “HCoV-2019*” or HCoV2019* or “2019 novel*” or Ncov* or “n-cov” or “SARS-CoV- 2*” or “SARSCoV-2*” or “SARSCoV2*” or “SARS-CoV2*” or SARSCov19* or “SARS-Cov19*” or “SARSCov-19*” or “SARS-Cov-19*” or SARSCov2019* or “SARS-Cov2019*” or “SARSCov-2019*” or “SARS-Cov-2019*” or SARS2* or “SARS-2*” or SARScoronavirus2* or “SARS-coronavirus-2*” or “SARScoronavirus 2*” or “SARS coronavirus2*” or SARScoronovirus2* or “SARS-coronovirus-2*” or “SARScoronovirus 2*” or “SARS coronovirus2*” or covid).ti,ab,kw. (62449)
13 (respiratory* adj2 (symptom* or disease* or illness* or condition*) adj5 (Wuhan* or Hubei* or China* or Chinese* or Huanan*)).ti,ab,kw. (372)
14 ((”seafood market*” or “food market*”) adj10 (Wuhan* or Hubei* or China* or Chinese* or Huanan*)).ti,ab,kw. (93)
15 (pneumonia* adj3 (Wuhan* or Hubei* or China* or Chinese* or Huanan*)).ti,ab,kw. (583)
16 ((outbreak* or wildlife* or pandemic* or epidemic*) adj1 (Wuhan* or Hubei* or China* or Chinese* or Huanan*)).ti,ab,kw. (146)
17 Middle East respiratory syndrome/ (1633)
18 (”middle east respiratory syndrome*” or “middle eastern respiratory syndrome*” or MERSCoV* or “MERS-CoV*” or MERS).ti,ab,kw. (6425)
19 (”severe acute respiratory syndrome*” or SARS).ti,ab,kw. (32952)
20 (”SARS-CoV-1*” or “SARSCoV-1*” or “SARSCoV1*” or “SARS-CoV1*” or SARSCoV or “SARS- CoV” or SARS1* or “SARS-1*” or SARScoronavirus1* or “SARS-coronavirus-1*” or “SARScoronavirus 1*” or “SARS coronavirus1*” or SARScoronovirus1* or “SARS-coronovirus-1*” or “SARScoronovirus 1*” or “SARS coronovirus1*”).ti,ab,kw. (22626)
21 or/7-20 (99755)
22 limit 21 to medline (25815) 23 21 not 22 (73940)
24 6 and 23 (413)
25 nonhuman/ not (human/ and nonhuman/) (4724889) 26 24 not 25 (398)
27 limit 26 to (english language and yr=”2002 -Current”) (388)
Cochrane Database of Systematic Reviews (CDSR) & CENTRAL
Issue 10 of 12, 2020
ID Search
#1 MeSH descriptor: [Vitamin D] explode all trees
#2 MeSH descriptor: [Vitamin D Deficiency] explode all trees #3 ((vitamin* near/5 D*) or vitaminD*)
#4 (ergocalciferol* or calciferol* or vs041h42xc or dihydrotachysterol* or dihydrotachysterin* or calcamine or “67-96-9″ or “r5lm3h112r” or “Hydroxyvitamin D*” or “25Hydroxyvitamin D*” or “HydroxyvitaminD*” or “25HydroxyvitaminD*” or hydroxycalciferol* or 25hydroxycalciferol* or hydroxyergocalciferol* or 25hydroxyergocalciferol* or ercalcidiol or “25(OH)D” or “21343-40-8″ or alfacalcidol*)
#5 (cholecalciferol* or colecalciferol* or calciol or “67-97-0″ or 1c6v77qf41 or hydroxycholecalciferol* or hydroxycolecalciferol* or 25hydroxycholecalciferol* or 25hydroxycolecalciferol* or calcifediol* or calcidiol* or “19356-17-3″ or p6yz13c99q or t0wxw8f54e or dihydroxycholecalciferol* or dihydroxycolecalciferol* or 25dihydroxycholecalciferol* or 25dihydroxycolecalciferol* or dihydroxyvitamin D* or 25dihydroxyvitamin* or dihydroxyvitaminD* or calcitriol* or “32222-06-3″ or “40013-87-4″ or “55721-11-4″)
#6 #1 or #2 or #3 or #4 or #5
#7 MeSH descriptor: [Coronavirus] explode all trees
#8 MeSH descriptor: [Coronavirus Infections] explode all trees
#9 ((corona* or corono*) near/1 (virus* or viral* or virinae*)):ti,ab,kw
#10 (coronavirus* or coronovirus* or coronavirinae* or CoV or HCoV* or Betacoronavirus* or Betacoronovirus*):ti,ab,kw
#11 (”2019 nCoV” or 2019nCoV* or “19 nCoV” or 19nCoV* or nCoV2019* or “nCoV 2019″ or nCoV19* or “nCoV 19″ or “COVID 19″ or COVID19* or “COVID 2019″ or COVID2019* or “HCoV 19″ or
HCoV19* or “HCoV 2019″ or HCoV2019* or “2019 novel” or Ncov* or “n cov” or “SARS CoV 2″ or “SARSCoV 2″ or “SARSCoV2″ or “SARS CoV2″ or SARSCov19* or “SARS Cov19″ or “SARSCov 19″ or “SARS Cov 19″ or SARSCov2019* or “SARS Cov2019″ or “SARSCov 2019″ or “SARS Cov 2019″
or SARS2* or “SARS 2″ or SARScoronavirus2* or “SARS coronavirus 2″ or “SARScoronavirus 2″ or “SARS coronavirus2″ or SARScoronovirus2* or “SARS coronovirus 2″ or “SARScoronovirus 2″ or “SARS coronovirus2″ or covid):ti,ab,kw
#12 (respiratory* near/2 (symptom* or disease* or illness* or condition*) near/5 (Wuhan* or Hubei* or China* or Chinese* or Huanan*)):ti,ab,kw
#13 ((”seafood market” or “seafood markets” or “food market” or “food markets”) near/10 (Wuhan* or Hubei* or China* or Chinese* or Huanan*)):ti,ab,kw
#14 (pneumonia* near/3 (Wuhan* or Hubei* or China* or Chinese* or Huanan*)):ti,ab,kw
#15 ((outbreak* or wildlife* or pandemic* or epidemic*) near/1 (Wuhan* or Hubei* or China* or Chinese* or Huanan*)):ti,ab,kw
#16 MeSH descriptor: [Middle East Respiratory Syndrome Coronavirus] explode all trees
#17 (”middle east respiratory syndrome” or “middle eastern respiratory syndrome” or “middle east respiratory syndromes” or “middle eastern respiratory syndromes” or MERSCoV* or “MERS CoV” or MERS):ti,ab,kw
#18 (”severe acute respiratory syndrome” or “severe acute respiratory syndromes” or SARS):ti,ab,kw
#19 (”SARS CoV 1″ or “SARSCoV 1″ or “SARSCoV1″ or “SARS CoV1″ or SARSCoV or SARS
CoV or SARS1 or “SARS 1″ or SARScoronavirus1 or “SARS coronavirus 1″ or “SARScoronavirus 1″ or “SARS coronavirus1″ or SARScoronovirus1 or “SARS coronovirus 1″ or “SARScoronovirus 1″ or “SARS coronovirus1″):ti,ab,kw
#20 {or #7-#19}
#21 #6 and #20
#22 (clinicaltrials or trialsearch):so #23 #21 not #22
This database comprises a full copy of the medRxiv and bioRxiv Covid-19/SARS-CoV-2 collection, which was downloaded via a custom feed. The EPPI database feed was run at 8am on the 23rd October 2020 before being sifted for relevance to the vitamin D topic. We did not use the native interface in MedRxiv or BioRxiv due to problems with the search functionality and data extraction options.
1 additional result was identified by searching for references from the medRxiv and bioRxiv Covid- 19/SARS-CoV-2 collection that had appeared in our custom feed between 8am on the 23rd October and 10am on the 29th October 2020. This set of references have not previously been assessed by NICE for relevance to the vitamin D topic. The following terms were searched for in the title and abstract fields of these records in EPPI reviewer: vitamin; hydroxyvitamin; dihydroxyvitamin, ergocalciferol; calciferol; cholecalciferol; colecalciferol (combined with the Boolean ‘OR’).
Note that EPPI Reviewer 5 automatically truncates search terms.
World Health Organization Global research on coronavirus disease (COVID-19)
Search conducted 29/10/2020.
tw:(”vitamin D” OR “vitamin D1″ OR “vitamin D2″ OR “vitamin D3″ OR “vitamin D4″ OR “vitamin D5″ OR hydroxyvitamin OR dihydroxyvitamin OR ergocalciferol OR calciferol OR cholecalciferol OR colecalciferol)
348 results
Search conducted 29/10/2020.
Condition: covid OR ncov OR 2019nCoV OR covid19 OR SARS OR (Severe Acute Respiratory Syndrome) OR Coronavirus OR corona OR orthocoronavirinae OR MERS OR (middle respiratory syndrome)
Other terms: (Vitamin (d OR d1 OR d2 OR d3 OR d4 OR d5)) OR hydroxyvitamin OR dihydroxyvitamin OR ergocalciferol OR calciferol OR cholecalciferol OR colecalciferol
Appendix C - Effectiveness & association evidence study selection
|
 |
|
Figure 1 Review question 1
  |
|||||
 |
|||||
  |
|||||
 |
|||||
Figure 2 Review question 2
 |
|||||||||
 |
|||||||||
 |
|||||||||
  |
|||||||||
 |
|||||||||
 |
|||||||||
|
|
Figure 3 Review question 3
Appendix D - Effectiveness & association evidence
Effectiveness evidence
Table 1 Entrenas Castillo 2020
| Bibliographic reference | - | Entrenas Castillo, Marta; Entrenas Costa, Luis Manuel; Vaquero Barrios, José Manuel; Alcalá Díaz, Juan Francisco; López Miranda, José; Bouillon, Roger; Quesada Gomez, José Manuel; Effect of calcifediol treatment and best available therapy versus best available therapy on intensive care unit admission and mortality among patients hospitalized for COVID-19: A pilot randomized clinical study.; J Steroid Biochem Mol Biol; 2020; vol. 203; 105751-105751 |
| Study details | Trial registration number and/or trial name | NCT04366908 in NIH clinical trials database |
| Study details | Study type | Randomised controlled trial (RCT) |
| Study details | Study location | Córdoba, Spain. |
| Study details | Study setting | Hospital, Reina Sofía University Hospital. |
| Study details | Study dates | Not reported. Paper received by journal 6th July 2020. |
| Study details | Sources of funding | Clinical Research Program at COVID-19 “Progreso y Salud” Foundation and Foundation for Biomedical Research of Córdoba (FIBICO). |
| Study details | Inclusion criteria | COVID-19 confirmed by a radiographic pattern of viral pneumonia scored by CURB65 and by a positive SARS-CoV-2 PCR.
Clinical samples for SARS-CoV-2 diagnostic testing were obtained according to WHO guidelines. For each patient, a sampling strategy was implemented in which samples were obtained on admission. Upper respiratory tract samples were obtained by nasopharyngeal |
| exudate sampling. Procedures for RNA extraction and real-time RT-PCR (rtRT-PCR) were undertaken in the local Central Microbiology Laboratory (Code 202 MagCore® Viral Nucleic Acid Extraction Kit and AllplexTM 2019-nCoV Assay by Seegene or VIASURE SARS-CoV-2 Real Time PCR Detection Kit).Respiratory function was assessed by PaO2/FiO2 index. A chest X-ray was taken in all patients on admission All X-ray tests were evaluated by an expert team of chest radiologist. | ||
| Study details | Exclusion criteria | <18 years of age
Pregnant women |
| Study details | Intervention(s) | Eligible patients were allocated at a 2 calcifediol:1 no calcifediol ratio through electronic randomisation performed by hospital statisticians. Participants allocated to the intervention arm took oral Calcifediol in soft capsules (0.532 mg) on the day of admission. They took another
0.266 mg of Calcifediol on days 3 and 7 and then weekly until discharge or intensive care unit (ICU) admission. They also received standard care per hospital protocol: a combination of hydroxychloroquine (400 mg every 12 hours on the first day, and 200 mg every 12 hours for the following 5 days), azithromycin (500 mg orally for 5 days) and for patients with pneumonia and NEWS score≥5, a broad spectrum antibiotic (ceftriaxone 2 g intravenously every 24 hours for 5 days) was added to hydroxychloroquine and azithromycin. |
| Study details | Comparator | The participants randomised to the comparator arm received standard care per hospital protocol only and no calcifediol: a combination of hydroxychloroquine (400 mg every 12 hours on the first day, and 200 mg every 12 hours for the following 5 days), azithromycin (500 mg orally for 5 days) and for patients with pneumonia and NEWS score ≥ 5, a broad spectrum antibiotic (ceftriaxone 2 g intravenously every 24 hours for 5 days) was added to hydroxychloroquine and azithromycin. |
| Study details | Outcome measures | COVID-19 mortality |
| Study details | Number of participants | N=76
n=50 randomised to intervention arm n=26 randomised to comparator arm |
| Study details | Duration of follow-up | Until ICU admission, death or discharge from hospital. |
| Study details | Loss to follow-up | No loss to follow-up. |
| Study details | Methods of analysis | Descriptive statistics were used for demographic, laboratory, and clinical prognostic factors related to COVID-19 for each treatment arm. The comparison between groups of quantitative |
| variables were performed by using t-test for qualitative variables, Chi squared tests and Fisher’s exact tests (with frequencies < 5) were used. Univariate and multivariable logistic regressions were used to estimate odds ratio and 95 % CIs for the probability of admission to ICU. Significant p-value was considered when p < 0.05.
All the analysis has been done using IBM SPSS. The pilot trial was reported according to the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline. |
||
| Study details | Study limitations (reviewer) | There is a possible interaction between vitamin D and azithromycin.
The standard care given to participants is not standard for the UK, limiting study applicability. Follow-up length was not noted in the study. This study was conducted during the first wave of infections. There is a difference in people being admitted to ICU between the first wave population and the second wave population. For example, in this study during the first wave patients were well enough to take capsules on admission. Blinding was incomplete and participants may have been put into ICU earlier in the comparator group than in the intervention group. |
| Study arms | Calcifediol (N = 50) | Participants who were randomised to receive calcifediol, the intervention. |
| Study arms | No calcifediol (N = 26) | Participants who were randomised to receive no calcifediol, the comparator. |
| Study-level characteristics | Ethnicity | (N = 76) Custom value N/A |
| Study-level characteristics | Body mass index | (N = 76) Custom value N/A |
| Study-level characteristics | Socioeconomic status | (N = 76) Custom value N/A |
| Study-level characteristics | Previous history of COVID-19 | (N = 76) Custom value N/A |
| Study-level characteristics | Other supplement use | (N = 76) Custom value N/A |
| Study-level characteristics | Timing of vitamin D measurements | (N = 76) Custom value N/A |
| Study-level characteristics | Shielding status | (N = 76) Custom value N/A |
| Study-level characteristics | Living in care homes | (N = 76) Custom value N/A |
| Arm-level characteristics | Age | Calcifediol (N = 50) Mean/SD 53.14 (10.77)
No calcifediol (N = 26) Mean/SD 52.77 (9.35) |
| Arm-level characteristics | Males | Calcifediol (N = 50) Mean/SD 56.3 (8.29)
No calcifediol (N = 26) Mean/SD 52.13 (10.05) |
| Arm-level characteristics | Females | Calcifediol (N = 50) Mean/SD 49.43 (12.28)
No calcifediol (N = 26) Mean/SD 54.13 (7.99) |
| Arm-level characteristics | % Female | Calcifediol (N = 50) Sample Size n = 23; % = 46
No calcifediol (N = 26) Sample Size n = 8; % = 31 |
| Arm-level characteristics | Comorbidities | Calcifediol (N = 50)
No calcifediol (N = 26) |
| Arm-level characteristics | ≥60 years | Calcifediol (N = 50) Sample Size n = 14; % = 28
No calcifediol (N = 26) Sample Size n = 5; % = 19.23 |
| Arm-level characteristics | Previous lung disease | Calcifediol (N = 50) Sample Size n = 4; % = 28
No calcifediol (N = 26) Sample Size n = 2; % = 7.69 |
| Arm-level characteristics | Previous chronic kidney disease | Calcifediol (N = 50) Sample Size n = 0; % = 0
No calcifediol (N = 26) Sample Size n = 0; % = 0 |
| Arm-level characteristics | Previous diabetes | Calcifediol (N = 50) Sample Size n = 3; % = 6
No calcifediol (N = 26) Sample Size n = 5; % = 19.23 |
| Arm-level characteristics | Previous high blood pressure | Calcifediol (N = 50) Sample Size n = 11; % = 24.19
No calcifediol (N = 26) Sample Size n = 15; % = 57.69 |
| Arm-level characteristics | Previous cardiovascular disease | Calcifediol (N = 50) Sample Size n = 2; % = 4
No calcifediol (N = 26) Sample Size n = 1; % = 3.85 |
| Arm-level characteristics | At least one prognostic risk factor | Calcifediol (N = 50) Sample Size n = 24; % = 48
No calcifediol (N = 26) Sample Size n = 16; % = 61.54 |
| Arm-level characteristics | PaO2/FiO2 | Calcifediol (N = 50) Mean/SD 346.57 (73.38)
No calcifediol (N = 26) Mean/SD 334.62 (66.33) |
| Arm-level characteristics | C-reactive protein | Calcifediol (N = 50) mg/L Mean/SD 82.93 (62.74)
No calcifediol (N = 26) mg/L Mean/SD 94.71 (63.64) |
| Arm-level characteristics | LDH (U/L) | Calcifediol (N = 50) Mean/SD 308.12 (83.83)
No calcifediol (N = 26) Mean/SD 345.81 (108.57) |
| Arm-level characteristics | D-dimer (ng/mL) | Calcifediol (N = 50) Mean/SD 650.92 (405.61)
No calcifediol (N = 26) Mean/SD 1333.54 (2570.5) |
| Arm-level characteristics | Ferritin (ng/mL) | Calcifediol (N = 50) Mean/SD 691.04 (603.54)
No calcifediol (N = 26) Mean/SD 825.16 (19.54) |
| Arm-level characteristics | IL-6 (22/48) pg/mL | Calcifediol (N = 50) Mean/SD 28.88 (75.05)
No calcifediol (N = 26) Mean/SD 19.54 (19.45) |
| Arm-level characteristics | Use of immune suppressing treatments
Immunosuppressed and transplanted |
Calcifediol (N = 50) Sample size n = 6; % = 12
No calcifediol (N = 26) Sample size n = 1; % = 3.85 |
| Outcomes | ICU admission
Polarity: Lower values are better |
Calcifediol (N = 50) Sample size n = 1; % = 2
No calcifediol (N = 26) Sample size n = 13; % = 50 |
| Outcomes | Mortality
Polarity: Lower values are better |
Calcifediol (N = 50) Sample size n = 0; % = 0
No calcifediol (N = 26) Sample size n = 2; % = 7.69 |
Risk of ICU admission depending on treatment
Comparison of calcifediol vs no calcifediol on ICU admission. A statistically significant difference was identified for the variable hypertension (26 had a history of hypertension of which 11 (42 %) received Calcifediol and 15 (58 %) not (CI: - 0.58 to - 0.13; p: 0.002) and close to statistical significance for diabetes 3 (6%)
versus 5 (19 %). Therefore, a multivariable logistic regression analysis was performed to adjust the model by possible confounding variables such as hypertension and type 2 diabetes mellitus for the probability of the admission to the Intensive Care Unit.
| Calcifediol vs No calcifediol | |
|
N1 = 50, N2 = 26 |
|
|
ICU admission Polarity: Lower values are better |
|
|
Unadjusted Univariate analysis without taking into account other variables |
|
|
Odds ratio/95% CI |
0.02 (0.002 to 0.17) |
|
Adjusted Multivariable analysis taking into account other variables |
|
|
Odds ratio/95% CI |
0.03 (0.003 to 0.25) |
|
Section |
Question |
Answer |
|
Domain 1: Bias arising from the randomisation process |
1. 1. Was the allocation sequence random? |
Yes |
|
1. 2. Was the allocation sequence concealed until participants were enrolled and assigned to interventions? |
Probably yes |
|
|
1.3 Did baseline differences between intervention groups suggest a problem with the randomisation process? |
Yes |
|
Section |
Question |
Answer |
|
Risk of bias judgement for the randomisation process |
High
(2:1 [intervention:comparator] ratio not justified. Two comorbidities were unbalanced and were found to be more prevalent in the comparator group |
|
|
Domain 2a: Risk of bias due to deviations from the intended interventions (effect of assignment to intervention) |
2.1. Were participants aware of their assigned intervention during the trial? |
Yes |
|
2.2. Were carers and people delivering the interventions aware of participants’ assigned intervention during the trial? |
Yes |
|
|
2.3. If Y/PY/NI to 2.1 or 2.2: Were there deviations from the intended intervention that arose because of the experimental context? |
No/Probably no |
|
|
2.4. If Y/PY to 2.3: Were these deviations from intended intervention balanced between groups? |
Not applicable |
|
|
2.5 If N/PN/NI to 2.4: Were these deviations likely to have affected the outcome? |
Not applicable |
|
|
2.6 Was an appropriate analysis used to estimate the effect of assignment to intervention? |
Yes |
|
|
2.7 If N/PN/NI to 2.6: Was there potential for a substantial impact (on the result) of the failure to analyse participants in the group to which they were randomized? |
Not applicable |
|
Section |
Question |
Answer |
|
Risk of bias for deviations from the intended interventions (effect of assignment to intervention) |
Some concerns
(Difficult or impossible for participants in the comparator group to receive vitamin D, but missing doses could have occurred for the intervention group. ITT analysis conducted.) |
|
|
Domain 2b: Risk of bias due to deviations from the intended interventions (effect of adhering to intervention) |
2.1. Were participants aware of their assigned intervention during the trial? |
Yes |
|
2.2. Were carers and people delivering the interventions aware of participants’ assigned intervention during the trial? |
Yes |
|
|
2.3. If Y/PY/NI to 2.1 or 2.2: Were important co-interventions balanced across intervention groups? |
Yes |
|
|
2.4. Could failures in implementing the intervention have affected the outcome? |
Yes |
|
|
2.5. Did study participants adhere to the assigned intervention regimen? |
Yes |
|
|
2.6. If N/PN/NI to 2.3 or 2.5 or Y/PY/NI to 2.4: Was an appropriate analysis used to estimate the effect of adhering to the intervention? |
Yes |
|
|
Risk of bias judgement for deviations from the intended interventions (effect of adhering to intervention) |
Some concerns
(Difficult or impossible for participants in the comparator group to receive vitamin D, but missing doses could have occurred for the intervention group.) |
|
|
Domain 3. Bias due to missing outcome data |
3.1 Were data for this outcome available for all, or nearly all, participants randomised? |
Yes |
|
Section |
Question |
Answer |
|
3.2 If N/PN/NI to 3.1: Is there evidence that result was not biased by missing outcome data? |
Not applicable |
|
|
3.3 If N/PN to 3.2: Could missingness in the outcome depend on its true value? |
Not applicable |
|
|
3.4 If Y/PY/NI to 3.3: Do the proportions of missing outcome data differ between intervention groups? |
Not applicable |
|
|
3.5 If Y/PY/NI to 3.3: Is it likely that missingness in the outcome depended on its true value? |
Not applicable |
|
|
Risk-of-bias judgement for missing outcome data |
Low
(No missing data.) |
|
|
Domain 4. Bias in measurement of the outcome |
4.1 Was the method of measuring the outcome inappropriate? |
Probably no |
|
4.2 Could measurement or ascertainment of the outcome have differed between intervention groups ? |
No |
|
|
4.3 If N/PN/NI to 4.1 and 4.2: Were outcome assessors aware of the intervention received by study participants ? |
Not applicable |
|
|
4.4 If Y/PY/NI to 4.3: Could assessment of the outcome have been influenced by knowledge of intervention received? |
Not applicable |
|
Section |
Question |
Answer |
|
4.5 If Y/PY/NI to 4.4: Is it likely that assessment of the outcome was influenced by knowledge of intervention received? |
Not applicable |
|
|
Risk-of-bias judgement for measurement of the outcome |
Low
(Objective measures of outcome conducted at one site.) |
|
|
Domain 5. Bias in selection of the reported result |
5.1 Was the trial analysed in accordance with a pre-specified plan that was finalised before unblinded outcome data were available for analysis ? |
No |
|
5.2 Is the numerical result being assessed likely to have been selected, on the basis of the results, from multiple outcome measurements (e.g. scales, definitions, time points) within the outcome domain? |
No/Probably no |
|
|
5.3 Is the numerical result being assessed likely to have been selected, on the basis of the results, from multiple analyses of the data? |
Yes/Probably yes |
|
|
Risk-of-bias judgement for selection of the reported result |
High
(Baseline characteristics suggest problem with randomisation; Reported outcome, mortality, not analysed in multivariable analysis. Only ICU was reported in this way, even though they are both listed on the clinical trials register as outcomes. Adjustment for multivariable analysis not fully explored or reported, only hypertension and diabetes are reported as definitively included in the model but does include “others”.) |
|
|
Overall bias and Directness |
Risk of bias judgement |
High
(Randomisation; Selection of the reported result.) |
|
Section |
Question |
Answer |
|
Overall Directness |
Directly applicable (There could be differences in the clinical decisions made before hospitalisation and ICU admission due to this study not being in the UK and changes over the course of the pandemic) |
|
Annweiler, 2020 |
| Bibliographic
Reference |
Annweiler, Cedric; Hanotte, Berangere; de l’Eprevier, Claire Grandin; Sabatier, Jean-Marc; Lafaie, Ludovic; Celarier, Thomas; Vitamin D
and survival in COVID-19 patients: A quasi-experimental study.; The Journal of steroid biochemistry and molecular biology; 2020; 105771 |
Study details
|
Trial registration number and/or trial name |
Not reported. |
|
Study type |
Non-randomised controlled trial Quasi-experimental intervention study |
|
Study location |
Rhône, France |
|
Study setting |
Nursing home |
|
Study dates |
March - 15th May 2020 |
|
Sources of funding |
No sources of funding declared. |
|
Inclusion criteria |
Clinically obvious or diagnosed COVID-19 with RT-PCR in March-April 2020 Data available on the treatments received, including vitamin D supplementation, since the diagnosis of COVID-19 and at least during the previous month. Data available on the vital status and COVID-19 evolution as of May 15, 2020. No objection from the resident and/or relatives to the use of anonymized clinical and biological data for research purposes. |
|
Exclusion criteria |
Not reported |
|
Intervention(s) |
Bolus vitamin D3 supplementation during or just before COVID-19.
All residents in the nursing-home receive chronic vitamin D supplementation with regular maintenance boluses (single oral dose of 80,000 IU vitamin D3 every 2-3 months), without systematically performing serum control test as recommended in French nursing- homes due to the very high prevalence of hypovitaminosis D reaching 90-100 % in this population. The “Intervention group” was defined as all COVID-19 residents who received an oral bolus of 80,000 IU vitamin D3 either in the week following the suspicion or diagnosis of COVID-19, or during the previous month. None received D2 or intramuscular supplements. All medications were dispensed and supervised by a nurse. |
|
Comparator |
The “Comparator group” corresponded to all other COVID-19 residents who did not receive any recent vitamin D supplementation. None received D2 or intramuscular supplements. All medications were dispensed and supervised by a nurse. |
|
Outcome measures |
COVID-19 mortality Measured during follow-up period. Follow-up started from the day of COVID-19 diagnosis for each patient, and continued until May 15, 2020, or until death if applicable.
OSCI score for COVID-19 in acute phase The secondary outcome was the score on the World Health Organisation’s Ordinal Scale for Clinical Improvement (OSCI) for COVID-19. The score was calculated by the geriatrician of the nursing-home during the most severe acute phase of COVID-19 for each patient. The OSCI distinguishes between several levels of COVID-19 clinical severity according to the outcomes and dedicated treatments required, with a score ranging from 0 (benign) to 8 (death). A score of 4 corresponds to the introduction of oxygen (nasal oxygen catheter or oral nasal mask), and a score of 6 to intubation and invasive ventilation. |
|
Number of participants |
N=66
Intervention group, n=57 Comparator, n=9 |
|
Duration of follow-up |
From initial diagnosis in March/April to death or 15th May. |
|
Loss to follow-up |
None reported. |
|
Methods of analysis |
Participants’ characteristics were summarized using means and standard deviations (SD) or frequencies and percentages, as appropriate. The study reported that the number of observations was higher than 40, comparisons were not affected by the shape of the error distribution and no transform was applied. Comparisons between participants separated into Intervention and Comparator groups were performed using Mann-Whitney U test or the Chi-square test or Fisher test, as appropriate, and then according to mortality.
3 models were conducted: 1) associations between predictor variables, such as vitamin D3 supplements, and the likelihood of COVID- 19 mortality at a specific time; 2) comparing time to death between intervention and comparator groups; 3) associations between bolus vitamin D3 supplements and OSCI score, taking into account factors that may affect the result.
1) A full-adjusted Cox regression was used to examine the associations of mortality (dependent variable) with bolus vitamin D3 supplements and covariables (independent variables). The model produces a survival function that provides the probability of death at a given time for the characteristics supplied for the independent variables.
2) The elapsed time to death was studied by survival curves computed according to Kaplan-Meier method and compared by log-rank test.
3) Univariate and multiple linear regressions were used to examine the association of bolus vitamin D3 supplementation (independent variable) with OSCI score (dependent variable), while adjusting for potential confounders.
P-values<0.05 were considered significant.
All statistics were performed using SPSS (v23.0, IBM Corporation, Chicago, IL) and SAS® version 9.4 software (Sas Institute Inc). |
|
Study limitations (authors) |
The study cohort was restricted to a limited number of nursing-home residents who may be unrepresentative of all older adults.
The study aimed to control for important characteristics that could modify the association, residual potential confounders might still be present such as the serum concentration of 25(OH)D at baseline. As this analysis was not planned, no concerted efforts were made to systematically measure the serum 25(OH)D concentration before and after supplementation.
The quasi-experimental design is less robust than an RCT. Participants in the comparator group did not receive vitamin D placebo, and there was no randomization. However, the authors noted that the characteristics of the two groups did not differ at baseline, which, they suggest, links the survival difference to vitamin D3 supplementation. |
|
Study limitations (reviewer) |
Even though there were no differences in measured baseline characteristics between groups, there may be other unmeasured differences that could bias the result.
It is unknown how much contact the participants had with the nurse who dispensed the medication. The nurse may have given other health protective advice that provided a short-term benefit of people who had more recent supplementation over people who did not.
The timing of vitamin D supplementation relative to the timing of diagnosis meant that this study looked at both prevention and intervention, which reduces clarity on the mechanism in which vitamin D works. It is not possible to discern whether people who had supplementation before COVID-19 diagnosis experienced better outcomes and people who had supplementation after. Subgroup or sensitivity analyses would not have helped here either because of the small sample size. |
Study arms
|
Intervention group (N = 57)
Participants who received bolus vitamin D3 supplement within a month before or up to a week after COVID-19 diagnosis or suspicion of diagnosis. |
|
Comparator (N = 9)
Participants who did not receive bolus vitamin D3 supplement within a month before or up to a week after COVID-19 diagnosis or suspicion of diagnosis. |
Characteristics
Study-level characteristics
|
Study (N = 66) |
|
|
Age |
|
|
Mean/SD |
87.7 (9) |
|
Intervention |
|
Study (N = 66) |
|
|
Mean/SD |
87.7 (9.3) |
|
Comparator |
|
|
Mean/SD |
87.4 (7.2) |
|
% Female |
|
|
Sample Size |
n = 51 ; % = 77.3 |
|
Intervention |
|
|
Sample Size |
n = 45 ; % = 78.9 |
|
Comparator |
|
|
Sample Size |
n = 6 ; % = 66.7 |
|
Ethnicity |
|
|
Custom value |
NA |
|
Comorbidities |
|
|
Custom value |
NA |
|
BMI |
|
|
Custom value |
NA |
|
Use of immune suppressing treatments |
|
|
Custom value |
NA |
|
Study (N = 66) |
|
|
Socioeconomic status |
|
|
Custom value |
NA |
|
Previous history of COVID-19 |
|
|
Custom value |
NA |
|
Other supplement use |
|
|
Custom value |
NA |
|
Timing of vitamin D measurements |
|
|
Custom value |
NA |
|
Shielding status |
|
|
Custom value |
NA |
|
Living in care homes |
|
|
Sample Size |
n = 66 ; % = 100 |
|
Vitamin D status |
|
|
Custom value |
na, assumed that most, if not all, of the residents were vitamin D deficient but were supplemented. |
|
Use of corticosteroids |
|
|
Sample Size |
n = 4 ; % = 6.1 |
|
Intervention |
|
Study (N = 66) |
|
|
Sample Size |
n = 3 ; % = 5.3 |
|
Comparator |
|
|
Sample Size |
n = 1 ; % = 11.1 |
|
Use of hydrocychloroquine |
|
|
Sample Size |
n = 2 ; % = 3 |
|
Intervention |
|
|
Sample Size |
n = 2 ; % = 3.5 |
|
Comparator |
|
|
Sample Size |
n = 0 ; % = 0 |
|
Use of dedicated antibiotics |
|
|
Sample Size |
n = 34 ; % = 51.5 |
|
Intervention |
|
|
Sample Size |
n = 21 ; % = 54.4 |
|
Comparator |
|
|
Sample Size |
n = 3 ; % = 3.33 |
|
Hospitalisation for COVID-19 |
|
|
Sample Size |
n = 4 ; % = 6.1 |
|
Study (N = 66) |
|
|
Intervention |
|
|
Sample Size |
n = 4 ; % = 7 |
|
Comparator |
|
|
Sample Size |
n = 0 ; % = 0 |
Outcomes
Comparison of study outcomes according to the study arm.
Includes mortality and OSCI scores for both arms.
|
Intervention group |
Comparator |
|
|
N = 57 |
N = 9 |
|
|
Mortality Polarity: Lower values are better |
||
|
Sample Size |
n = 10 ; % = 17.5 |
n = 5 ; % = 55.6 |
|
OSCI score Measurements taken in COVID-19 acute phase Polarity: Lower values are better |
||
|
Zero |
||
|
Sample Size |
n = 1 ; % = 1.8 |
n = 0 ; % = 0 |
| Intervention group | Comparator | |
|
N = 57 |
N = 9 |
|
|
One |
||
|
Sample Size |
n = 21 ; % = 37.5 |
n = 1 ; % = 11.1 |
|
Two |
||
|
Sample Size |
n = 18 ; % = 32.1 |
n = 1 ; % = 1.11 |
|
Three |
||
|
Sample Size |
n = 1 ; % = 1.8 |
n = 0 ; % = 0 |
|
Four |
||
|
Sample Size |
n = 4 ; % = 7.1 |
n = 1 ; % = 11.1 |
|
Five |
||
|
Sample Size |
n = 2 ; % = 3.6 |
n = 1 ; % = 11.1 |
|
Six |
||
|
Sample Size |
n = 0 ; % = 0 |
n = 0 ; % = 0 |
|
Seven |
||
|
Sample Size |
n = 0 ; % = 0 |
n = 0 ; % = 0 |
|
Eight |
||
|
Sample Size |
n = 10 ; % = 17.5 |
n = 5 ; % = 55.6 |
| Intervention group | Comparator | |
|
N = 57 |
N = 9 |
|
|
Follow-up after COVID-19 diagnosis Polarity: Not set |
||
|
Mean/SD |
38.9 (15.6) |
20.9 (14.3) |
Hazard ratio for COVID-19 mortality according to the use of bolus vitamin D3 supplements
Values shown are adjusted for potential confounders unless stated otherwise: age, gender, number of drugs daily taken, functional abilities, nutritional status, COVID-19 treatment with corticosteroids and/or hydroxychloroquine and/or dedicated antibiotics, and hospitalization for COVID-19.
| Intervention group vs Comparator | |
|
N1 = 57, N2 = 9 |
|
|
Recent bolus vitamin D3 supplementation Polarity: Lower values are better |
|
|
Adjusted |
|
|
Hazard ratio/95% CI |
0.11 (0.03 to 0.48) |
|
Unadjusted |
|
|
Hazard ratio/95% CI |
0.21 (0.07 to 0.63) |
|
Age Polarity: Lower values are better |
|
|
Hazard ratio/95% CI |
1.06 (0.98 to 1.15) |
| Intervention group vs Comparator | |
|
N1 = 57, N2 = 9 |
|
|
Female gender Polarity: Lower values are better |
|
|
Hazard ratio/95% CI |
1.03 (0.3 to 3.54) |
|
Number of drugs usually taken per day Polarity: Lower values are better |
|
|
Hazard ratio/95% CI |
0.73 (0.52 to 1.02) |
|
Use of corticosteroids Polarity: Lower values are better |
|
|
Hazard ratio/95% CI |
6.64 (0.46 to 95.24) |
|
Use of hydroxychloroquine Polarity: Lower values are better |
|
|
Hazard ratio/95% CI |
15.07 (0.75 to 302.53) |
|
Use of dedicated antibiotics Polarity: Lower values are better |
|
|
Hazard ratio/95% CI |
0.36 (0.07 to 1.95) |
|
Hospitalisation for COVID-19 Polarity: Lower values are better |
| Intervention group vs Comparator | |
|
N1 = 57, N2 = 9 |
|
|
Hazard ratio/95% CI |
0.38 (0.02 to 7.06) |
| Section | Question | Answer |
| Study participation | Summary Study participation | Moderate risk of bias (Important baseline characteristics, such as BMI, ethnicity, use of other supplements and socioeconomic status not included) |
| Study Attrition | Study Attrition Summary | Low risk of bias (no attrition reported) |
| Prognostic factor measurement | Prognostic factor Measurement Summary |
Moderate risk of bias (Two groups who had vitamin D at separate times were not split in analyses) |
| Outcome Measurement | Outcome Measurement Summary | Low risk of bias (outcomes were objective and/or a valid, recognised tool for measuring COVID-19 severity, completed by geriatrician) |
| Study Confounding | Study Confounding Summary | High risk of bias (Important confounders, such as BMI, ethnicity, use of other supplements and socioeconomic status not included) |
| Statistical Analysis and Reporting | Statistical Analysis and Presentation Summary | High risk of bias (Small sample size and event rate for large number of adjustments made. Important confounders, such as BMI, ethnicity, use of other supplements and socioeconomic status not accounted for in analyses) |
| Overall risk of bias and directness | Risk of Bias | High |
|
Directness |
Partially applicable (Analysing outcomes from people who had supplementation before and after diagnosis does not make association as clear as a study that would split these ways of supplementing people. There could be differences in the clinical decisions made before hospitalisation due to this study not being in the UK and changes over the course of the pandemic) |
|
Annweiler, 2020a |
| Bibliographic Reference | Annweiler, Gaelle; Corvaisier, Mathieu; Gautier, Jennifer; Dubee, Vincent; Legrand, Erick; Sacco, Guillaume; Annweiler, Cedric; Vitamin D
Supplementation Associated to Better Survival in Hospitalized Frail Elderly COVID-19 Patients: The GERIA-COVID Quasi-Experimental Study.; Nutrients; 2020; vol. 12 (no. 11) |
Study details
|
Trial registration number and/or trial name |
Not reported. |
|
Study type |
Retrospective cohort study |
|
Study location |
France |
|
Study setting |
Hospital, Angers University Hospital. |
|
Study dates |
March-May 2020 |
|
Sources of funding |
Funding reported as “not applicable”. |
|
Inclusion criteria |
Patients hospitalized in the geriatric acute care unit of the hospital No objection from the patient and/or relatives to the use of anonymized clinical and biological data for research purpose |
|
COVID-19 diagnosed with RT-PCR and/or chest CT-scan Data available on the treatments received, including vitamin D supplementation, since the diagnosis of COVID-19 and over the preceding year at least Data available on the vital status 14 days after the diagnosis of COVID-19 |
|
|
Exclusion criteria |
Not reported |
|
Intervention(s) |
The regular intake of bolus vitamin D supplements over the preceding year was systematically noted from the primary care physicians’ prescriptions and sought by questioning the patients and their relatives.
“Group 1″ was defined as all COVID-19 patients who had received oral boluses of vitamin D supplements over the preceding year. Bolus included the doses of 50,000 IU vitamin D3 per month, or the doses of 80,000 IU or 100,000 IU vitamin D3 every 2-3 months. None received D2 or intramuscular supplements, and no patient in Group 1 received additional supplements following the diagnosis of COVID-19.
“Group 2″ was defined as the COVID-19 patients usually not supplemented with vitamin D, but who received an oral supplement of 80,000 IU vitamin D3 within a few hours of the diagnosis of COVID-19. |
|
Comparator |
“Group 3″ was all COVID-19 patients who had received no vitamin D supplements, neither over the preceding year nor after the diagnosis of COVID-19. The absence of vitamin D treatment being mostly explained by the patients’ refusal to be supplemented, since vitamin D supplementation is recommended with no biological testing in all patients over 65 years of age in France. |
|
Outcome measures |
COVID-19 mortality Within 14 days of COVID-19 diagnosis. Follow-up continued for 14 days or until death.
OSCI score for COVID-19 in acute phase The score on the 9-pointWorld Health Organization’s ordinal scale for clinical improvement (OSCI) for COVID-19. The OSCI distinguishes between several levels of COVID-19 clinical severity according to the outcomes and dedicated treatments required, with a score ranging from 0 (no clinical or virological sign of infection) to 8 (death). The score was determined by the geriatrician of the hospital unit on admission, then revised regularly according to the clinical course of the patients. The highest score during hospitalization was used for the present analysis, corresponding to the most severe acute phase of COVID-19 for each patient. A score of 3 corresponds to a degree of severity requiring hospitalization (i.e., all GERIA-COVID participants had an OSCI score ≥3 here), a score of 5 corresponds to the introduction of non-invasive ventilation, and a score of 6 to intubation and invasive ventilation. Severe COVID-19 was defined here as a score of 5 or more. |
|
Number of participants |
N=77
Group 1, n=29 |
| Group 2, n=16
Group 3, n=32 |
|
|
Duration of follow-up |
14 days or until death. |
|
Loss to follow-up |
None reported. |
|
Methods of analysis |
The participants’ characteristics were summarized using means and standard deviations (SD) or frequencies and percentages, as appropriate. As the number of observations was higher than 40, comparisons were not affected by the shape of the error distribution and no transformation was applied.
4 models were made: 1) comparisons between groups for the reported outcomes; 2) the association between each group and 14-day mortality at a specific time, adjusting for confounders; 3) comparison of survival between the groups; 4) association between vitamin D status and severe COVID-19, adjusted for confounding variables.
1) Comparisons between participants separated into three groups according to the intervention (i.e., regular supplementation versus supplementation initiated after COVID-19 diagnosis versus no supplementation) were performed using analysis of variance (ANOVA) or Mann-Whitney-U and Kruskal-Wallis tests for quantitative variables as appropriate, and using Chi-square test or Fisher exact test for qualitative variables as appropriate. To address the issue of multiple comparisons, analyses were completed by a post hoc Fisher’s least significant difference (LSD) test.
2) A fully adjusted Cox regression was used to examine the associations of 14-day mortality (dependent variable) with vitamin D supplementation and covariables (independent variables). The model produces a survival function that provides the probability of death at a given time for the characteristics supplied for the independent variables.
3) The elapsed time to death was studied by survival curves computed according to the Kaplan-Meier method and compared by log- rank test.
4) A multiple logistic regression was used to examine the association of vitamin D supplementation (independent variable) with severe COVID-19 defined as an OSCI score ≥5 (dependent variable), while adjusting for potential confounders.
p-values <0.05 were considered significant. All statistics were performed using SPSS and SAS. |
|
Study limitations (authors) |
The study participants were restricted to a limited number of hospitalized frail elderly patients who might be unrepresentative of all older adults. It is also possible that the limited sample size in each group had resulted in a lack of power with increased beta risk. |
| The study aimed to control for the important characteristics that could modify the association, but residual potential confounders might still be present such as the serum concentration of 25(OH)D at baseline, a low level classically ensuring the efficacy of the supplementation, or the OSCI score on admission. The OSCI score was collected in the most acute phase of COVID-19 as it was reported that COVID-19 can get worse between 7-10 days due to the cytokine storm regardless of the initial disease severity.
The quasi-experimental design is less robust than an RCT. Participants in the comparator group did not receive vitamin D placebo. Moreover, there was no randomization. It is plausible that the participants who regularly received vitamin D supplementation (Group 1) were treated better by their family physicians than the others, thereby exhibiting more stable chronic diseases such as cardiovascular comorbidities. It is also plausible that patients or relatives refused taking vitamin D supplementation in Group 3, because the conditions of patients were too severe for them to take the supplements. The authors noted that the history did not differ between the 3 groups and that their demographical and health characteristics were similar at baseline. However, the proportion of women who are likely to suffer from osteoporosis and may have received corresponding treatment that includes vitamin D. |
|
|
Study limitations (reviewer) |
Estimations of vitamin D status based on supplementation may be incorrect as it relies on medicine compliance. |
Study arms
|
Regular vitamin D supplementation (N = 29) |
|
Vitamin D supplementation after COVID-19 (N = 16) |
|
Non-supplemented comparator (N = 32) |
Characteristics
Study-level characteristics
|
Study (N = 77) |
|
|
Ethnicity |
|
|
Custom value |
NA |
|
Study (N = 77) |
|
|
BMI |
|
|
Custom value |
NA |
|
Use of immune suppressing treatments |
|
|
Custom value |
NA |
|
Socioeconomic status |
|
|
Custom value |
NA |
|
Previous history of COVID-19 |
|
|
Custom value |
NA |
|
Other supplement use |
|
|
Custom value |
NA |
|
Timing of vitamin D measurements |
|
|
Custom value |
NA |
|
Shielding status |
|
|
Custom value |
NA |
|
Living in care homes |
|
|
Custom value |
NA |
|
Vitamin D status |
|
Study (N = 77) |
|
|
Custom value |
NA |
Arm-level characteristics
|
Regular vitamin D supplementation (N = 29) |
Vitamin D supplementation after COVID-19 (N = 16) |
Non-supplemented comparator (N = 32) |
|
|
Age |
|||
|
MedianIQR |
88 (87 to 93) |
85 (84 to 89) |
88 (84 to 92) |
|
% Female |
|||
|
Sample Size |
n = 20 ; % = 69 |
n = 5 ; % = 31.3 |
n = 13 ; % = 40.6 |
|
Comorbidities |
|||
| Severe undernutrition
Serum albumin concentration <30 g/L |
|||
|
Sample Size |
n = 9 ; % = 31 |
n = 3 ; % = 18.8 |
n = 9 ; % = 28.1 |
|
Haematological and solid cancers |
|||
|
Sample Size |
n = 10 ; % = 34.5 |
n = 4 ; % = 25 |
n = 13 ; % = 40.6 |
|
Hypertension |
|||
|
Sample Size |
n = 18 ; % = 62.1 |
n = 10 ; % = 62.5 |
n = 21 ; % = 65.6 |
|
Cardiomyopathy |
|||
|
Sample Size |
n = 13 ; % = 44.8 |
n = 11 ; % = 68.8 |
n = 18 ; % = 56.3 |
|
Regular vitamin D supplementation (N = 29) |
Vitamin D supplementation after COVID-19 (N = 16) |
Non-supplemented comparator (N = 32) |
|
|
Number of acute health issues at hospital admission |
|||
|
MedianIQR |
3 (2 to 4) |
3.5 (2 to 5) |
2.5 (1 to 4) |
|
CRP at admission (mg/L) |
|||
|
MedianIQR |
44 (19 to 110) |
69 (15.5 to 140) |
59 (29 to 166) |
|
Use of antibiotics |
|||
|
Sample Size |
n = 23 ; % = 79.3 |
n = 14 ; % = 87.5 |
n = 22 ; % = 68.8 |
|
Use of systemic corticosteroids |
|||
|
Sample Size |
n = 6 ; % = 20.7 |
n = 2 ; % = 12.5 |
n = 5 ; % = 15.6 |
|
Use of pharmacological treatments of respiratory disorders |
|||
|
Sample Size |
n = 1 ; % = 3.5 |
n = 2 ; % = 12.5 |
n = 7 ; % = 21.9 |
|
Glycated haemoglobin (%) |
|||
|
MedianIQR |
6 (5.5 to 6.6) |
6.4 (6 to 8.2) |
6.2 (5.9 to 6.7) |
Outcomes
COVID-19 outcomes
| Regular vitamin D supplementation | Vitamin D supplementation after COVID- 19 | Non-supplemented comparator | |
|
N = 29 |
N = 16 |
N = 32 |
|
|
Severe COVID-19 defined as an OSCI score for COVID-19 in acute phase ≥5
Polarity: Lower values are better |
|||
|
Sample Size |
n = 3 ; % = 10.3 |
n = 4 ; % = 25 |
n = 10 ; % = 31.3 |
|
14-day mortality Polarity: Lower values are better |
|||
|
Sample Size |
n = 2 ; % = 6.9 |
n = 3 ; % = 18.8 |
n = 10 ; % = 31.3 |
Association between vitamin D supplementation and COVID-19 outcomes
Results from the Cox regression. Adjusted for age, gender, GIR score, severe undernutrition, history of cancer, history of hypertension, history of cardiomyopathy, glycated haemoglobin, number of acute health problems, use of antibiotics, use of systemic corticosteroids, use of treatments of respiratory disorders.
| Regular vitamin D supplementation vs Non-supplemented comparator | Vitamin D supplementation after COVID-19 vs Non- supplemented comparator | |
|
N1 = 29, N2 = 32 |
N1 = 16, N2 = 32 |
|
|
Mortality Evaluated by Cox regression Polarity: Lower values are better |
||
|
Hazard ratio/95% CI |
0.07 (0.01 to 0.61) |
0.37 (0.06 to 2.21) |
| Regular vitamin D supplementation vs Non-supplemented comparator | Vitamin D supplementation after COVID-19 vs Non- supplemented comparator | |
|
N1 = 29, N2 = 32 |
N1 = 16, N2 = 32 |
|
|
Severe COVID-19 Evaluated by multiple logistic regression Polarity: Lower values are better |
||
|
Odds ratio/95% CI |
0.08 (0.01 to 0.81) |
0.46 (0.07 to 2.85) |
| Section | Question | Answer |
| Study participation | Summary Study participation | Moderate risk of bias (Important baseline characteristics, such as BMI, ethnicity, use of other supplements and socioeconomic status not included) |
| Study Attrition | Study Attrition Summary | Low risk of bias (no attrition reported) |
| Prognostic factor measurement | Prognostic factor Measurement Summary | Moderate risk of bias (Method for ensuring prognostic factor was received appropriately for each group not reliable - vitamin D supplements assumed to be taken at home and adherence cannot be guaranteed) |
| Outcome Measurement | Outcome Measurement Summary | Moderate risk of bias (outcomes were objective and/or a valid, recognised tool for measuring COVID-19 severity, completed by geriatrician but 14 days was short for follow-up) |
| Study Confounding | Study Confounding Summary | High risk of bias (Important confounders, such as BMI, ethnicity, use of other supplements and socioeconomic status not included) |
| Statistical Analysis and Reporting | Statistical Analysis and Presentation Summary | High risk of bias (Many factors adjusted for in small cohort likely to lead to overfitting; Important confounders, such as BMI, ethnicity, use of other supplements and socioeconomic status not accounted for in analyses) |
| Overall risk of bias and directness | Risk of Bias | High |
| Directness | Directly applicable |
| Hastie, 2020 |
| Bibliographic Reference | Hastie, Claire E; Mackay, Daniel F; Ho, Frederick; Celis-Morales, Carlos A; Katikireddi, Srinivasa Vittal; Niedzwiedz, Claire L; Jani, Bhautesh D; Welsh, Paul; Mair, Frances S; Gray, Stuart R; O’Donnell, Catherine A; Gill, Jason Mr; Sattar, Naveed; Pell, Jill P; Vitamin D
concentrations and COVID-19 infection in UK Biobank.; Diabetes & metabolic syndrome; 2020; vol. 14 (no. 4); 561-565 |
Study details
| Study design | Case-control study |
| Trial registration (if reported) | Not applicable. |
| Study start date | 16-Mar-2020 |
| Study end date | 14-Apr-2020 |
| COVID-19 prevalence at the time of the study |
Higher prevalence (e.g. during peak of first wave) |
|
Aim of the study |
The study hypothesised that blood 25 hydroxyvitamin D (25(OH)D) would negatively associate with COVID-19 infection, which may explain the higher incidence of COVID-19 in ethnic minority participants. The study used the data collected by the UK Biobank study which recruited 502,624 participants between 2006 and 2010, which followed up people over time to identify causes of disease in middle and old age. It matched this data with COVID-19 tests taken by participants. |
| County/ Geographical location |
England, Scotland and Wales. |
| Study setting | UK Biobank and Public Health England data. |
|
Population description |
UK Biobank recruited 502,624 participants aged between 37-73 years. Complete data was available for 348,598 participants.
The population studied in this correlation study was a subset of this, 1474 individuals who had undergone COVID-19 tests (2724 tests total). |
|
Inclusion criteria |
- Complete data in UK Biobank study
- Taken a COVID-19 test |
| Exclusion criteria | None reported. |
|
Vitamin D status measurements |
Blood samples were taken when participants were first recruited into the study between 2006-2010. Biochemical assays were conducted at a central laboratory alongside other tests and assays. No detail on which tests was used to quantify 25 hydroxyvitamin D or other vitamin D concentration in the blood was reported. Minimum detectable value was imputed at 10 nmol/L if vitamin D concentration was too low to detect, and with the maximum detectable value of 375 nmol/L if the concentration was too high for detection. |
| Methods used to confirm COVID-19 infection | Test results data provided by Public Health England, including specimen date, origin (if the person was an inpatient or not), and if the test was positive or negative for COVID-19. A positive result was defined as at least 1 positive COVID-19 test, data available for 16th March 2020 and 14th April 2020. |
| Intervention | Not applicable. |
| Comparator (where applicable) | Not applicable. |
| Methods for population selection/allocation |
Not applicable |
| Methods for case- matching with control |
Not reported. Adjustment conducted in analyses. |
|
Methods of data analysis |
Three models were conducted: 1) a correlation between 25(OH)D and confirmed COVID-19 infection adjusting for covariates; 2) a correlation between ethnicity and COVID-19 infection adjusting for 25(OH)D; 3) a correlation looking for an interaction between ethnicity*25(OH)D and COVID-19 infection.
1) The association between 25(OH)D, as a continuous variable, and COVID-19 infection was explored with a univariable logistic regression analysis. The model was adjusted for sex, month of assessment, Townsend deprivation quintile, household income, self- reported health rating, smoking status, BMI quintile, ethnicity, age at assessment, diabetes, systolic blood pressure, diastolic blood pressure, and long-standing illness, disability or infirmity. Sensitivity analyses were conducted with participants were categorised as vitamin D deficient (<25 nmol/L) or not deficient; and another conducted with participants categorised as vitamin D insufficient (<50 nmol/L) or sufficient.
2) The association between ethnicity and COVID-19 infection was explored with univariable logistic regression analysis. The model was first adjusted for 25(OH)D and then sex, month of assessment, Townsend deprivation quintile, household income, self-reported health |
| rating, smoking status, BMI quintile, age at assessment, diabetes, systolic blood pressure, diastolic blood pressure, and long-standing illness, disability or infirmity.
3) The possible interaction between 25(OH)D and ethnicity and its association with COVID-19 infection was explored with multivariable analysis.
All analyses were conducted on Stata v14. |
|
| Attrition/loss to follow-up | Not applicable. |
|
Source of funding |
CEH is funded by Health Data Research-UK (Ref. Edin-1).
SVK acknowledges funding from a NRS Senior Clinical Fellowship (SCAF/15/02), the Medical Research Council (MC_UU_12017/13) and the Scottish Government Chief Scientist Office (SPHSU13). CLN is supported by the Medical Research Council (MR/R024774/1). NS receives funding from the British Heart Foundation Research Excellence Award (RE/18/6/34217). |
|
Study limitations (Author) |
Not representative of the general population: participants live in less socioeconomically deprived areas, are predominantly white, and have fewer self-reported health conditions.
Baseline measures for 25(OH)D were collected over a decade before the study was conducted and it would have been preferable to have more up-to-date measurements or measurements taken immediately before development of COVID-19. |
|
Study limitations (Reviewer) |
Regarding the old 25(OH)D measurements used in the study, they defend their data by saying that concentrations vary more by season than year and generally track over time. However, this neglects other causes of vitamin D change as it also tracks health state, or covariates to health state.
All other variables measured by the UK Biobank study would have been collected at the same time. Variables that can change, such as blood pressure and smoking status, may have changed over the decade since the study was conducted. There is no reason to assume that vitamin D levels from a decade ago would correlate with levels currently. The changes may be net neutral but there may be certain groups whose smoking status, for example, has changed more towards smoke-free or smoking more, but it would not be possible to deduce who this would be. This may mean that some variables that are used for adjustment do not provide accurate data for people’s current state leading to misleading results.
The results are reported as “COVID” and “no COVID”. The COVID group rightly contains only people who have had a confirmed COVID infection. However, the “no COVID” group contains everyone else in the UK Biobank database who has not had a confirmed COVID diagnosis, but also people who have not been tested. Considering the projections that showed many more people were infected at this |
| time than had been tested, including people who were symptomatic and isolating and the people who took tests may be more ill than people who did not. Therefore, this may not accurately represent the 25(OH)D levels of people with and without COVID infection.
Vitamin D assay limits of 10nmol/L and 375nmol/L under- or overestimate some values that are lower or higher than thee concentrations. Many people have serum 25(OH)D <10μg/ml and therefore a linear correlation would also overestimate how low these values are, not accurately representing the data. |
Study arms
| No COVID (N = 348149)
Number of people in the UK Biobank register who either received a negative COVID test or who were not tested. |
| COVID (N = 449)
Number of people in the UK Biobank register who received a positive COVID test. |
Characteristics
Study-level characteristics
| Study (N = 348598) | |
| Vitamin D by ethnicity | |
| White | |
| MedianIQR | 48.1 (33.8 to 63.4) |
| Black | |
| MedianIQR | 29.9 (21 to 41.3) |
| South Asian | |
| MedianIQR | 22.1 (14.5 to 33.7) |
| Other | |
| MedianIQR | 33.7 (23.3 to 47.4) |
Arm-level characteristics
| No COVID (N = 348149) | COVID (N = 449) | |
| Age | ||
| MedianIQR | 57 (49 to 63) | 58 (49 to 64) |
| Gender
%female |
||
| Sample Size | n = 179758 ; % = 51.63 | n = 184 ; % = 40.98 |
| Ethnicity | ||
| White | ||
| Sample Size | n = 331464 ; % = 95.21 | n = 385 ; % = 85.75 |
| Black | ||
| Sample Size | n = 5022 ; % = 1.44 | n = 32 ; % = 7.13 |
| South Asian | ||
| Sample Size | n = 5917 ; % = 1.7 | n = 19 ; % = 4.23 |
| Other | ||
| Sample Size | n = 5746 ; % = 1.65 | n = 13 ; % = 2.9 |
| Comorbidities | ||
| Diabetes | ||
| Sample Size | n = 18825 ; % = 5.41 | n = 49 ; % = 10.91 |
| BMI | ||
| Underweight | ||
| Sample Size | n = 1759 ; % = 0.51 | n = 2 ; % = 0.45 |
| Normal weight | ||
| Sample Size | n = 115410 ; % = 33.15 | n = 95 ; % = 21.16 |
| Overweight | ||
| Sample Size | n = 148210 ; % = 42.57 | n = 194 ; % = 43.21 |
| Obese | ||
| Sample Size | n = 82770 ; % = 23.77 | n = 158 ; % = 35.19 |
| Use of immune suppressing treatments | ||
| No COVID (N = 348149) | COVID (N = 449) | |
| Socioeconomic status
Townsend deprivation quintile - higher is more deprived |
||
| One | ||
| Sample Size | n = 70669 ; % = 10.37 | n = 51 ; % = 11.36 |
| Two | ||
| Sample Size | n = 70726 ; % = 20.31 | n = 76 ; % = 16.93 |
| Three | ||
| Sample Size | n = 70644 ; % = 20.29 | n = 64 ; % = 14.25 |
| Four | ||
| Sample Size | n = 70270 ; % = 20.18 | n = 105 ; % = 23.39 |
| Five | ||
| Sample Size | n = 65840 ; % = 18.91 | n = 143 ; % = 31.85 |
| Previous history of COVID-19 | ||
| Custom value | NA | NA |
| Other supplement use | ||
| Custom value | NA | NA |
| Timing of vitamin D measurements | ||
| Custom value | NA | NA |
| Shielding status | ||
| Custom value | NA | NA |
| Living in care homes | ||
| Custom value | NA | NA |
| Current smoker | ||
| Sample Size | n = 36112 ; % = 10.37 | n = 51 ; % = 11.36 |
| Self-reported health rating | ||
| Excellent | ||
| Sample Size | n = 60508 ; % = 17.38 | n = 45 ; % = 10.02 |
| No COVID (N = 348149) | COVID (N = 449) | |
| Good | ||
| Sample Size | n = 203640 ; % = 58.49 | n = 227 ; % = 50.56 |
| Fair | ||
| Sample Size | n = 69676 ; % = 20.01 | n = 133 ; % = 29.62 |
| Poor | ||
| Sample Size | n = 14325 ; % = 4.11 | n = 44 ; % = 9.8 |
| Long standing illness, disability or infirmity | ||
| Sample Size | n = 110679 ; % = 31.79 | n = 204 ; % = 45.43 |
| Systolic blood pressure
mmHg |
||
| MedianIQR | 138 (125 to 151) | 138 (127 to 153) |
| Diastolic blood pressure
mmHg |
||
| MedianIQR | 82 (75 to 89) | 83 (76 to 90) |
| Vitamin D
nmol/L |
||
| MedianIQR | 47.2 (32.7 to 62.7) | 43.8 (28.7 to 61.6) |
Outcomes
Association between vitamin D and confirmed COVID-19 infection
Multivariable results adjusted for ethnicity, sex, month of assessment, Townsend deprivation quintile, household income, self-reported health rating, smoking status, BMI category, age at assessment, diabetes, SBP, DBP, and long-standing illness, disability or infirmity.
| COVID vs No COVID | |
| N1 = 449, N2 = 348149 | |
| Vitamin D
Polarity: Lower values are better |
|
| Univariate |
| COVID vs No COVID | |
| N1 = 449, N2 = 348149 | |
| Odds ratio/95% CI | 0.99 (0.99 to 1) |
| Multivariable | |
| Odds ratio/95% CI | 1 (1 to 1.01) |
| Vitamin D deficient (<25 nmol/L)
Polarity: Lower values are better |
|
| Univariable | |
| Odds ratio/95% CI | 1.37 (1.07 to 1.76) |
| Multivariable | |
| Odds ratio/95% CI | 0.92 (0.71 to 1.21) |
| Vitamin D insufficient (<50 nmol/L)
Polarity: Lower values are better |
|
| Univariable | |
| Odds ratio/95% CI | 1.19 (0.99 to 1.44) |
| Multivariable | |
| Odds ratio/95% CI | 0.88 (0.72 to 1.08) |
Association between ethnicity and confirmed COVID-19 infection.
Multivariable model in this table adjusted for sex, month of assessment, Townsend deprivation quintile, household income, self-reported health rating, smoking status, BMI category, age at assessment, diabetes, SBP, DBP, and long-standing illness, disability or infirmity.
| COVID vs No COVID | |
| N1 = 449, N2 = 348149 | |
| Univariable
Polarity: Lower values are better |
|
| White (referent) | |
| Odds ratio/95% CI | 1 (1 to 1) |
| Black | |
| Odds ratio/95% CI | 5.49 (3.82 to 7.88) |
| COVID vs No COVID | |
| N1 = 449, N2 = 348149 | |
| South Asian | |
| Odds ratio/95% CI | 2.76 (1.74 to 4.39) |
| Other | |
| Odds ratio/95% CI | 1.95 (1.12 to 3.39) |
| Adjusted for vitamin D concentration
Polarity: Lower values are better |
|
| White (referent) | |
| Odds ratio/95% CI | 1 (1 to 1) |
| Black | |
| Odds ratio/95% CI | 5.32 (3.68 to 7.7) |
| South Asian | |
| Odds ratio/95% CI | 2.65 (1.65 to 4.25) |
| Other | |
| Odds ratio/95% CI | 1.9 (1.09 to 3.32) |
| Multivariable
Polarity: Lower values are better |
|
| White (referent) | |
| Odds ratio/95% CI | 1 (1 to 1) |
| Black | |
| Odds ratio/95% CI | 4.3 (2.92 to 6.31) |
| South Asian | |
| Odds ratio/95% CI | 2.42 (1.5 to 3.93) |
| Other | |
| Odds ratio/95% CI | 1.87 (1.07 to 3.28) |
| Section | Question | Answer |
|
Study participation |
Summary Study participation |
High risk of bias
(People in the cohort were over-represented in factors of better health, white ethnicity and less deprived compared with the general population. Outreach by UK Biobank may not be sufficient to gather data for all eligible participants.) |
|
Study Attrition |
Study Attrition Summary |
Moderate risk of bias
(Difficult to assess the affect and potential bias of not including people that did not have complete data and therefore were not included in this study.) |
| Prognostic factor measurement | Prognostic factor Measurement Summary | High risk of bias
(Timing of vitamin D measurement allows significant risk of bias.) |
| Outcome Measurement | Outcome Measurement Summary | High risk of bias
(Due to lack of COVID-19 testing for all participants.) |
|
Study Confounding |
Study Confounding Summary | Moderate risk of bias
(Analyses adjust for confounding but the accuracy of confounders is doubtful since they were measured over a decade before the study took place.) |
|
Statistical Analysis and Reporting |
Statistical Analysis and Presentation Summary |
Moderate risk of bias
(All people, including people who had not taken a COVID-19 test were included. There may be selection bias between people who were tested and people who were not. To account for this potential bias, subgroup analyses could have been conducted to address this). |
| Overall risk of bias and directness |
Risk of Bias |
High
(Based on age of vitamin D measurements and excluding participants who do not have complete outcome data.) |
| Directness | Partially applicable
(Vitamin D status taken from historical measurements, 2006-2010) |
| Hernandez, 2020 |
| Bibliographic
Reference |
Hernandez, JL; Nan, D; Fernandez-Ayala, M; Garcia-Unzueta, M; Hernandez-Hernandez, M; Vitamin D Status in Hospitalized Patients
With SARS-CoV-2 Infection; The Journal of Clinical Endocrinology & Metabolism; 2020; (no. earlyonline) |
Study details
| Study design | Case-control study |
| Study start date | 10-Mar-2020 |
| Study end date | 20-May-2020 |
| COVID-19 prevalence at the time of the study |
Higher prevalence (e.g. during peak of first wave) |
| Aim of the study | The study aimed to assess the serum 25OHD levels in hospitalized patients with COVID-19 compared to population-based controls. In addition, it looked at the possible association between serum 25OHD concentrations and COVID-19 severity and mortality. |
| County/ Geographical location |
Santander, Spain. |
| Study setting | Hospital for positive cohort; community for control cohort. |
|
Population description |
Patients with confirmed COVID-19 who were admitted to the University Hospital Marqués de Valdecilla made up the COVID-19 arm. They were recruited 10th - 30th March 2020. 19 patients were on vitamin D supplementation (11 taking cholecalciferol, 10 25,000 U monthly and 1 5,600 U weekly; 8 patients were taking calcifediol 0.266 mg monthly).
Concerning immunomodulatory therapy, patients were selected for tocilizumab according to the institutional protocol. Thus, tocilizumab was indicated if there was clinical worsening with PaO2/FIO2 ratio (PaFI) <300 and high serum acute-phase reactant levels when no contraindication for its use was present.
Controls were taken from the Camargo cohort during their last follow-up visit in January-March 2020. |
|
Inclusion criteria |
COVID-19 arm: COVID-19 positive inpatients.
Controls: participants who completed their last follow-up appointment in January-March 2020. |
|
Exclusion criteria |
Patients or controls with malabsorption disorders, liver cirrhosis, serum creatinine levels >1.9 mg/dl, or previous treatment with anticonvulsants.
Controls who took vitamin D supplements for more than 3 months were also excluded. They were allowed in the COVID-19 positive arm but were analysed separately. |
|
Vitamin D status measurements |
Serum samples from Covid-19 patients were provided by the IDIVAL Biobank samples collection (internal code 2020-126).
Serum 25OHD concentrations were determined in controls by a fully automated electrochemiluminescence system (Elecsys 2010, Roche Diagnostics, GmbH, Mannheim, Germany). The detection limit of serum 25OHD was 4 ng/ml. The intra-assay coefficient of variation (CV) was 5% and inter-assay was 7.5%.
In COVID-19 patients, serum 25OHD levels were obtained at admission and assessed by automated competitive chemiluminescence assay (Liaison XL, DiaSorin Inc, Stillwater MN, USA). Our laboratory is DEQAS (Vitamin D External Quality Assessment Scheme) certified for this parameter. The detection limit of serum 25OHD was 4 ng/ml. The intra-assay and interassay CV were 2.58% and 7.83%, respectively. |
| Methods used to confirm COVID-19 infection | Qualitative detection of RNA from the SARS-CoV-2 was performed by using Real-Time PCR. Blood samples from the controls were obtained from an antecubital vein in the morning after a requested 12- hour overnight fast. The serum was divided into 0.5-ml aliquots and stored at -40ºC. Routine biochemical parameters were measured by standard automated methods in a Technicon Dax autoanalyzer (Technicon Instruments, CO. USA). |
| Methods for case- matching with control |
Cases were sex-matched with population-based controls. |
|
Methods of data analysis |
4 models were used to compare 25(OH)D levels and associate 25(OH)D with clinical and laboratory parameters: 1) simple correlation between 25(OH)D and clinical indications; 2) comparing 25(OH)D levels between COVID-19 patients and controls that took into account confounding factors; 3) in COVID-19 patients only, simple association between 25(OH)D and disease severity; 4) in COVID-19 patients only, association between 25(OH)D and disease severity taking into account confounding factors.
1) Spearman rho was used to assess the relationships between serum 25(OH)D levels and several clinical and laboratory parameters. Serum 25(OH)D levels were stratified into four categories: below 10 ng/ml, between 10 and 20 ng/ml, between 20 and 30 ng/ml, and above 30 ng/ml. Vitamin D deficiency was defined as serum 25(OH)D levels <20 ng/ml (50 nmol/l).
2) A multivariable general linear model was set up to compare serum 25(OH)D levels between COVID-19 patients and controls (Bonferroni test), adjusting for confounding variables.
3+4) In the group of COVID-19 patients, univariable and multivariable binary logistic regression analyses were used to assess the association between vitamin D (as a continuous variable, or expressed as vitamin D deficiency or as quintiles) and the dependent variable of severity of the disease.
Demographic, clinical, and outcome data of COVID-19 patients were gathered from the hospital records, stored in a computerized database, and independently reviewed by two researchers. Missing data were not imputed. Smoking status was coded as current or |
|
Study limitations (authors) |
non-smoker. Immunosuppression included prolonged use (≥3 months) of corticoids (>10 mg/day of prednisone or equivalent) or immunomodulatory agents, and bone marrow or organ transplantation. Overall, the criteria for ICU admission were those of the guidelines by the American Thoracic Society and Infectious Diseases Society of America and the critical care ethic recommendations for the SARS-CoV-2 pandemic by the Intensive Medicine Spanish Society.
The endpoint variable for COVID-19 severity has been defined as the composite of admission to the intensive care unit (ICU), requirement for mechanical ventilation, or in-hospital mortality. Clinical outcomes were monitored up to May 20, 2020.
Continuous variables were expressed as mean +/- SD or median and interquartile range (IQR) and compared with the Student’s t-test or Mann-Whitney U-test depending on how the data was distributed. Categorical variables were presented as sample numbers and percentages and compared using chi square test or the Fisher’s exact test as appropriate for the distribution of the data.
Two-sided p-value of <0.05 considered statistically significant.
A post-hoc power analysis with the present sample size and the obtained difference in serum 25OHD levels between cases and controls yields a power of 100% to detect this difference. In fact, a difference of 2.1 ng/ml between groups already yields a potency of 89.8%. Nevertheless, due to the sample size and the lower number of events (especially mortality) in COVID-19 patients with and without vitamin D deficiency, the post-hoc power analysis for the severity endpoints was lower than 40%. |
| Problems inherent to an observational study that does not permit to establish whether vitamin D is simply a biomarker of exposure or a biomarker of effect on the disease.
Other vitamin D-related parameters such as the free fraction of 25OHD, 1,25 dihydroxyvitamin D, and vitamin D-binding protein were not measured.
The number of COVID-19 patients who were on oral vitamin D supplements is too small and on different dosages to draw solid conclusions on its role in the clinical outcomes of the disease.
The study was conducted in a single Spanish tertiary-care hospital, and data may not be generalized to other settings, ethnicities, or countries, especially those with specific policies for vitamin D supplementation or food fortification.
The methods to assess serum 25OHD levels in cases and controls were different, although there was a very good correlation between both techniques.
No dietary assessment has been carried out, and therefore information on dietary habits is lacking. |
| There could be differences in the clinical decisions made before hospitalisation and ICU admission due to this study not being in the UK and changes over the course of the pandemic. |
Study arms
| COVID-19 positive (N = 197)
Patients aged 18 or over with confirmed COVID-19 admitted to hospital who are not taking vitamin D supplements. |
| COVID-19 positive on vitamin D (N = 19)
COVID-19 positive patients who were taking vitamin D supplements for at least 3 months before they were admitted to hospital. |
| Control (N = 197)
Controls who were not COVID-19 positive |
Characteristics
Arm-level characteristics
| COVID-19 positive (N = 197) | COVID-19 positive on vitamin D (N = 19) | Control (N = 197) | |
| Age | |||
| MedianIQR | 61 (47.5 to 70) | 60 (59 to 75) | 61 (56 to 66) |
| Gender
Male |
|||
| Sample Size | n = 123 ; % = 63.4 | n = 7 ; % = 36.8 | n = 123 ; % = 62.4 |
| Ethnicity | |||
| Custom value | NA | NA | NA |
| Comorbidities | |||
| Cardiovascular disease | |||
| Sample Size | n = 21 ; % = 10.7 | n = 3 ; % = 15.8 | n = 22 ; % = 11.2 |
| Hypertension | |||
| COVID-19 positive (N = 197) | COVID-19 positive on vitamin D (N = 19) | Control (N = 197) | |
| Sample Size | n = 76 ; % = 38.6 | n = 12 ; % = 63.2 | n = 87 ; % = 44.2 |
| Diabetes | |||
| Sample Size | n = 34 ; % = 17.3 | n = 0 ; % = 0 | n = 31 ; % = 15.7 |
| COPD | |||
| Sample Size | n = 15 ; % = 7.6 | n = 2 ; % = 10.5 | n = 9 ; % = 4.6 |
| Active cancer | |||
| Sample Size | n = 7 ; % = 3.6 | n = 0 ; % = 0 | n = 8 ; % = 4.1 |
| BMI | |||
| Mean/SD | 29.2 (4.7) | 30.9 (6.3) | 28.9 (4) |
| Use of immune suppressing treatments | |||
| Sample Size | n = 16 ; % = 8.1 | n = 6 ; % = 31.6 | n = 2 ; % = 1 |
| Socioeconomic status | |||
| Custom value | NA | NA | NA |
| Previous history of COVID-19 | |||
| Custom value | NA | NA | NA |
| Other supplement use
ACE1/ARA2 agents |
|||
| Sample Size | n = 58 ; % = 29.4 | n = 7 ; % = 36.8 | n = 47 ; % = 23.9 |
| Timing of vitamin D measurements | |||
| Shielding status | |||
| Custom value | NA | NA | NA |
| Living in care homes | |||
| Custom value | NA | NA | NA |
| Vitamin D (ng/mL) | |||
| Mean/SD | 13.8 (7.2) | 21.1 (5.9) | 20.9 (7.4) |
| Current smoker | |||
| Sample Size | n = 14 ; % = 7.1 | n = 2 ; % = 10.5 | n = 34 ; % = 17.3 |
Outcomes
Main characteristics of COVID-19 patients according to the presence of vitamin D deficiency.
Includes comparisons between vitamin D deficient and sufficient participants in the COVID-19 not taking vitamin D supplementation only.
| Deficient 25(OH)D <20 ng/ml; | Sufficient 25(OH)D ≥20 ng/ml | p value | |
| N = 162 | N = 35 | ||
| Age
Polarity: Not set |
|||
| MedianIQR | 62 (48 to 70.3) | 58.4 (45 to 69) | 0.29 |
| Sex
Male Polarity: Not set |
|||
| Sample Size | n = 106 ; % = 65.4 | n = 17 ; % = 48.6 | 0.062 |
| BMI (kg/m²) Polarity: Not set | |||
| Mean/SD | 29 (4.9) | 29.8 (4.1) | 0.43 |
| Comorbidities
Polarity: Not set |
|||
| Cardiovascular disease | |||
| Sample Size | n = 21 ; % = 13 | n = 0 ; % = 0 | 0.029 |
| Hypertension | |||
| Sample Size | n = 68 ; % = 42 | n = 8 ; % = 22.9 | 0.035 |
| COPD | |||
| Sample Size | n = 13 ; % = 8 | n = 2 ; % = 5.7 | 0.99 |
| Active cancer | |||
| Sample Size | n = 7 ; % = 4.3 | n = 0 ; % = 0 | 0.36 |
| Immunosuppression | |||
| Sample Size | n = 11 ; % = 6.8 | n = 5 ; % = 14.3 | 0.17 |
| Diabetes | |||
| Sample Size | n = 28 ; % = 17.3 | n = 6 ; % = 17.1 | 0.98 |
| Deficient 25(OH)D <20 ng/ml; | Sufficient 25(OH)D ≥20 ng/ml | p value | |
| N = 162 | N = 35 | ||
| Pneumonia
Polarity: Not set |
|||
| Sample Size | n = 155 ; % = 95.7 | n = 33 ; % = 94.3 | 0.66 |
| PaO2/FlO2 ratio (PaFl) Polarity: Not set | |||
| MedianIQR | 444 (424 to 452) | 444 (436 to 452) | 0.17 |
| ACE1/ARA2 agents
Polarity: Not set |
|||
| Sample Size | n = 52 ; % = 32.1 | n = 6 ; % = 17.1 | 0.078 |
| Tocilizumab
Polarity: Not set |
|||
| Sample Size | n = 55 ; % = 34 | n = 8 ; % = 22.9 | 0.2 |
| ICU admission
Polarity: Not set |
|||
| Sample Size | n = 44 ; % = 27.2 | n = 6 ; % = 17.1 | 0.22 |
| Mechanical ventilation
Polarity: Not set |
|||
| Sample Size | n = 37 ; % = 84.1 | n = 6 ; % = 100 | 0.58 |
| Secondary infection
Polarity: Not set |
|||
| Sample Size | n = 38 ; % = 23.5 | n = 6 ; % = 17.1 | 0.42 |
| Death
Polarity: Not set |
|||
| Sample Size | n = 16 ; % = 10.2 | n = 4 ; % = 11.4 | 0.77 |
| Composite severity endpoint
Polarity: Not set |
|||
| Sample Size | n = 111 ; % = 68.5 | n = 27 ; % = 77.1 | 0.31 |
| Length of stay
Polarity: Not set |
|||
| MedianIQR | 12 (8 to 17) | 8 (6 to 14) | 0.013 |
| 25(OH)D |
| Deficient 25(OH)D <20 ng/ml; | Sufficient 25(OH)D ≥20 ng/ml | p value | |
| N = 162 | N = 35 | ||
| Polarity: Not set | |||
| Mean/SD | 13.8 (7.2) | 21.1 (5.9) |
Main features in COVID-19 patients with or without oral vitamin D supplements at admission.
| COVID-19 positive | COVID-19 positive on vitamin D | Control | |
| N = 197 | N = 19 | ||
| Pneumonia
Polarity: Not set |
|||
| Sample Size | n = 188 ; % = 95.4 | n = 18 ; % = 94.7 | 0.99 |
| PaO2/FlO2 ratio
Polarity: Not set |
|||
| MedianIQR | 44 (428 to 452) | 444 (432 to 452) | 0.52 |
| PaFi <300
Polarity: Not set |
|||
| Sample Size | n = 52 ; % = 26.4 | n = 1 ; % = 5.3 | 0.049 |
| Tocilizumab
Polarity: Not set |
|||
| Sample Size | n = 63 ; % = 32 | n = 1 ; % = 5.3 | 0.015 |
| ICU admission
Polarity: Not set |
|||
| Sample Size | n = 50 ; % = 25.4 | n = 1 ; % = 5.3 | 0.05 |
| Mechanical ventilation
Polarity: Not set |
|||
| Sample Size | n = 43 ; % = 86 | n = 1 ; % = 5.3 | 1 |
| Secondary infection
Polarity: Not set |
|||
| Sample Size | n = 44 ; % = 22.3 | n = 2 ; % = 10.5 | 0.38 |
| Death
Polarity: Not set |
|||
| Sample Size | n = 20 ; % = 10.4 | n = 2 ; % = 10.5 | 1 |
| COVID-19 positive | COVID-19 positive on vitamin D | Control | |
| N = 197 | N = 19 | ||
| Composite severity endpoint
Polarity: Not set |
|||
| Sample Size | n = 59 ; % = 29.9 | n = 3 ; % = 15.8 | 0.19 |
| Length of stay (days)
Polarity: Not set |
|||
| MedianIQR | 12 (8 to 16) | 8 (6 to 14) | 0.11 |
Relationship between serum vitamin D levels and composite severity endpoint
Within the COVID-19 positive arm who were not supplemented with vitamin D, vitamin D levels are correlated with disease severity. The OR corresponds to a unit of vitamin D in 1 ng/ml. Adjusted model used age, smoking, hypertension, diabetes mellitus, history of cardiovascular events, immunosuppression, body mass index, serum corrected calcium, glomerular filtration rate and the month of vitamin D determination.
|
COVID-19 positive vs COVID-19 positive |
|
| N1 = 197, N2 = 197 | |
| Composite severity endpoint
Polarity: Lower values are better |
|
| Unadjusted | |
| Odds ratio/95% CI | 1.55 (0.66 to 3.65) |
| Adjusted | |
| Odds ratio/95% CI | 1.13 (0.27 to 4.77) |
Vitamin D levels in COVID-19 patients and controls.
Multivariable general linear model adjusted for age, smoking, hypertension, diabetes mellitus, history of cardiovascular events, immunosuppression, body mass index, serum corrected calcium, glomerular filtration rate and the month of vitamin D determination. Only includes COVID-19 patients not on vitamin D supplementation, and controls.
| COVID-19 positive | COVID-19 positive on vitamin D | Control | |
| N = 197 | N = 19 | ||
| Vitamin D level (ng/mL) Polarity: Not set |
| COVID-19 positive | COVID-19 positive on vitamin D | Control | |
| N = 197 | N = 19 | ||
| Mean/95% CI | 11.9 (9.6 to 14.3) | 21.2 (19.7 to 22.7) | 0.01 |
| Section | Question | Answer |
|
Study participation |
Summary Study participation | Moderate risk of bias
(Controls only sex-matched with cases and did not control or report ethnicity, which has been shown to be associated with vitamin D status and COVID-19 morbidity and mortality.) |
| Study Attrition | Study Attrition Summary | Low risk of bias
(No attrition reported.) |
|
Prognostic factor measurement |
Prognostic factor Measurement Summary |
High risk of bias
(Two different assays with using different methods were used to measure serum 25(OH)D for cases and controls. Different assays have different limits of detection and chemical concentrations cannot be compared directly to one another unless the data has been transformed in a certain way, e.g. diagnostic test accuracy metrics.) Low risk of bias (Between vitamin D supplemented and non-vitamin D supplemented cases, the method is the same, so analyses comparing these should be considered at low risk of bias for this domain.) |
| Outcome Measurement | Outcome Measurement Summary | Low risk of bias
(HCP knowing the vitamin D status of participants in hospital unlikely to affect how they assessed patients as they were unaware the data would be used for this study at the time of assessment.) |
|
Study Confounding |
Study Confounding Summary | Moderate risk of bias
(Unsure if the study encountered missing data or how they accounted for it. Non-reporting of confounders/baseline characteristics has been covered in another domain so has not contributed to the overall decision of this domain.) |
| Statistical Analysis and Reporting | Statistical Analysis and Presentation Summary | Low risk of bias |
|
Overall risk of bias and directness |
Risk of Bias |
High
(High risk of bias for data comparing cases and controls due to different methods of measuring vitamin D. Note: moderate risk of bias for data that compares the vitamin D supplemented and non-supplemented arms because vitamin D was measured in the same way.) |
| Section | Question | Answer |
| Directness | Directly applicable (Note: there could be differences in the clinical decisions made before hospitalisation and ICU admission due to this study not being in the UK and changes over the course of the pandemic) |
| Karahan, 2020 |
| Bibliographic
Reference |
Karahan, S.; Katkat, F.; Impact of Serum 25(OH) Vitamin D Level on Mortality in Patients with COVID-19 in Turkey; Journal of Nutrition,
Health and Aging; 2020 |
Study details
| Study design | Case-control study |
| Trial registration (if reported) | Not reported. |
| Study start date | 01-Apr-2020 |
| Study end date | 20-May-2020 |
| COVID-19 prevalence at the time of the study |
Higher prevalence (e.g. during peak of first wave) |
| Aim of the study | To evaluate the association of vitamin D status with disease severity and mortality in patients with COVID-19. |
| County/ Geographical location |
Istanbul, Turkey. |
| Study setting | Research and training hospital. |
| Population description | The population was made up of 149 COVID-19 patients who were admitted into the hospital with confirmed COVID-19. |
| Inclusion criteria | Adult admitted to Health Sciences University, Bagcilar Training and Research Hospital with COVID-19. |
|
Exclusion criteria |
No vitamin D (25(OH)D) values.
Clinical presentation compatible with COVID-19 but who did not have a PCR-based test for SARS-CoV2. Paediatric patients. |
| Vitamin D status measurements | Serum 25(OH) vitamin D levels were studied by electrochemiluminescence method. Patients were stratified into different groups according to their serum 25(OH) vitamin D levels. Serum 25(OH) vitamin D level >30 ng/mL was accepted as normal. Vitamin D insufficiency and deficiency were defined as serum 25(OH) vitamin D levels of 21-29 ng/mL and <20 ng/mL, respectively. |
|
Methods used to confirm COVID-19 infection |
PCR but specific method not mentioned.
Classification of the severity of COVID-19 was done using the Chinese Clinical Guideline for classification of COVID-19 severity. Patient symptoms, laboratory values, and results of imaging studies performed at admission are used to determine severity of COVID-19.
Mild symptoms: Mild clinical symptoms and normal lung on radiologic imaging.
Moderate disease: Fever and pulmonary symptoms along with pneumonia on radiologic imaging.
Severe disease: The presence of any of the following criteria: i) respiratory distress (≥ 30 breaths/min); ii) oxygen saturation ≤ 93% at rest; iii) PaO2/FiO2 ≤ 300 mmHg or chest imaging shows obvious lesion progression > 50% within 24-48 hours).
Critical disease: The presence of any of the following criteria: i) respiratory failure and need for mechanical ventilation; ii) shock; iii) other organ failures that requires ICU care.
Since patients with mild COVID-19 were not hospitalized according to the Turkish national guidelines, the study did not involve any patient with mild disease. The study combined severe and critical COVID-19 in a single group named “severe-critical disease”. Patients were stratified into either to moderate or critical-severe COVID-19 groups. |
|
Methods of data analysis |
This evidence table has included patients as they were classified in the study. Patients were classified as either moderate or severe/critical COVID-19, and then also classified as survived or deceased. The study compared patients in the moderate arm to the severe/critical arm; and then compared patients who survived and patients who died.
Descriptive statistics were presented for continuous variables as either mean +/-standard deviation or median-interquartile range and compared with Independent Samples t-test or Mann-Whitney U tests depending on the distribution type of the data. Categorical |
| variables were reported as numbers and percentages and compared between groups by chi-square/Fisher exact test depending on the distribution of the data. Normal distribution was checked with the Kolmogorov-Smirnov test.
To evaluate the bivariate correlation between the serum 25(OH) vitamin D level and inflammatory marker, Pearson’s correlation was used. Univariate and multivariable logistic regression analyses were used to determine the independent associates of mortality. For all analyses, a p-value of <0.05 was considered significant.
SPSS 26.0 (IBM Corporation, NY, US) was used to perform all statistical analyses. |
|
| Attrition/loss to follow-up | No loss to follow-up. |
| Source of funding | No sources of funding declared. |
| Study limitations (Author) | As a retrospective analysis, confounders that can impact the mortality rate were not controlled.
The sample size is small. |
| Study limitations (Reviewer) | Due to the inclusion criteria, there was not a wide range of COVID-19 disease classifications as it omitted the mild cases. |
Study arms
| Moderate COVID-19 (N = 47)
Participants who had moderate COVID-19 |
| Severe-critical COVID-19 (N = 102)
Participants who made severe or critical COVID-19 |
| Survived (N = 80)
Participants who survived |
| Deceased (N = 69)
Participants who died |
Characteristics
Arm-level characteristics
| Moderate COVID-19 (N = 47) | Severe-critical COVID-19 (N = 102) | Survived (N = 80) | Deceased (N = 69) | |
| Age | ||||
| Mean/SD | 56.1 (15.2) | 67 (14.1) | 60 (15.1) | 67.7 (14.1) |
| Gender
Female |
||||
| Sample Size | n = 24 ; % = 51.1 | n = 44 ; % = 43.1 | n = 40 ; % = 50 | n = 28 ; % = 40.6 |
| Ethnicity | ||||
| Custom value | NA | NA | NA | NA |
| Comorbidities | ||||
| Coronary heart disease | ||||
| Sample Size | n = 3 ; % = 6.4 | n = 41 ; % = 40.2 | n = 13 ; % = 16.3 | n = 19 ; % = 27.5 |
| Hypertension | ||||
| Sample Size | n = 15 ; % = 31.9 | n = 70 ; % = 68.6 | n = 41 ; % = 51.3 | n = 44 ; % = 63.8 |
| Diabetes | ||||
| Sample Size | n = 12 ; % = 25.5 | n = 49 ; % = 48 | n = 25 ; % = 31.3 | n = 36 ; % = 52.2 |
| COPD | ||||
| Sample Size | n = 4 ; % = 8.5 | n = 11 ; % = 10.8 | n = 8 ; % = 10 | n = 7 ; % = 10.1 |
| Malignancy | ||||
| Sample Size | n = 6 ; % = 12.8 | n = 17 ; % = 16.7 | n = 9 ; % = 11.3 | n = 14 ; % = 20.3 |
| Chronic kidney disease | ||||
| Sample Size | n = 2 ; % = 4.3 | n = 27 ; % = 26.5 | n = 9 ; % = 11.3 | n = 20 ; % = 29 |
| Chronic atrial fibrillation | ||||
| Sample Size | n = 0 ; % = 0 | n = 15 ; % = 14.7 | n = 2 ; % = 2.5 | n = 13 ; % = 18.8 |
| Congestive heart failure | ||||
| Sample Size | n = 0 ; % = 0 | n = 18 ; % = 17.6 | n = 4 ; % = 5 | n = 14 ; % = 20.3 |
| BMI | ||||
| Moderate COVID-19 (N = 47) | Severe-critical COVID-19 (N = 102) | Survived (N = 80) | Deceased (N = 69) | |
| Custom value | NA | NA | NA | NA |
| Use of immune suppressing treatments | ||||
| Custom value | NA | NA | NA | NA |
| Socioeconomic status | ||||
| Custom value | NA | NA | NA | NA |
| Previous history of COVID-19 | ||||
| Custom value | NA | NA | NA | NA |
| Other supplement use | ||||
| Custom value | NA | NA | NA | NA |
| Timing of vitamin D measurements | ||||
| Custom value | NA | NA | NA | NA |
| Shielding status | ||||
| Custom value | NA | NA | NA | NA |
| Living in care homes | ||||
| Custom value | NA | NA | NA | NA |
| Smokers | ||||
| Sample Size | n = 10 ; % = 21.3 | n = 41 ; % = 40.2 | n = 21 ; % = 26.3 | n = 30 ; % = 43.5 |
Outcomes
Vitamin D level and status
| Moderate COVID-19 | Severe-critical COVID-19 | Survived | Deceased | |
| N = 47 | N = 102 | N = 80 | N = 69 | |
| 25(OH)D level (ng/mL) Polarity: Not set | ||||
| Mean/SD | 26.3 (8.4) | 10.1 (6.2) | 19.3 (11.2) | 10.4 (6.4) |
| Vitamin D status
Polarity: Not set |
| Moderate COVID-19 | Severe-critical COVID-19 | Survived | Deceased | |
| N = 47 | N = 102 | N = 80 | N = 69 | |
| ≤20 | ||||
| Sample Size | n = 8 ; % = 17 | n = 95 ; % = 93.1 | n = 39 ; % = 48.8 | n = 64 ; % = 92.8 |
| 21-29 | ||||
| Sample Size | n = 27 ; % = 57.4 | n = 7 ; % = 6.9 | n = 29 ; % = 6.3 | n = 5 ; % = 7.2 |
| ≤30 | ||||
| Sample Size | n = 12 ; % = 25.5 | n = 0 ; % = 0 | n = 12 ; % = 15 | n = 0 ; % = 0 |
Univariate and multivariable logistic regression analysis showing independent predictors of in-hospital mortality
Univariate analyses were conducted on age and comorbidities to assess if they were independently correlated with mortality from COVID-19. Serum 25(OH)D concentration was included in a multivariable model controlling for age and comorbidities.
|
Deceased vs Survived |
|
| N1 = 69, N2 = 80 | |
| Age
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 1.04 (1.01 to 1.06) |
| Smoking
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 2.16 (1.09 to 4.3) |
| Hyperlipidaemia
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 3.12 (1.45 to 6.72) |
| Diabetes
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 2.4 (1.23 to 4.68) |
| Chronic kidney disease
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 3.22 (1.35 to 7.66) |
| Chronic atrial fibrillation
Polarity: Lower values are better |
|
Deceased vs Survived |
|
| N1 = 69, N2 = 80 | |
| Odds ratio/95% CI | 9.05 (1.97 to 41.72) |
| Congestive heart failure
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 4.84 (1.51 to 15.49) |
| Acute kidney injury
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 4 (1.23 to 13.05) |
| eGFR
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 0.98 (0.97 to 0.99) |
| 25(OH)D level
Univariate Polarity: Lower values are better |
|
| Univariate | |
| Odds ratio/95% CI | 0.9 (0.86 to 0.94) |
| Multivariable | |
| Odds ratio/95% CI | 0.93 (0.88 to 0.98) |
| Section | Question | Answer |
| Study participation | Summary Study participation | Moderate risk of bias
(No descriptive statistics on ethnicity or BMI.) |
| Study Attrition | Study Attrition Summary | Low risk of bias |
| Prognostic factor measurement | Prognostic factor Measurement Summary | Low risk of bias |
| Outcome Measurement | Outcome Measurement Summary | Low risk of bias |
| Study Confounding | Study Confounding Summary | Moderate risk of bias
(Ethnicity and BMI not reported.) |
| Section | Question | Answer |
| Statistical Analysis and Reporting | Statistical Analysis and Presentation Summary | Low risk of bias |
| Overall risk of bias and directness |
Risk of Bias |
Moderate
(Important confounders of ethnicity and BMI was not reported as baseline characteristics nor used in the multivariable model assessing the association between 25(OH)D level and mortality.) |
| Directness | Partially applicable
(Historic vitamin D measurements used) |
| Kaufman, 2020 |
| Bibliographic
Reference |
Kaufman, Harvey W; Niles, Justin K; Kroll, Martin H; Bi, Caixia; Holick, Michael F; SARS-CoV-2 positivity rates associated with circulating
25-hydroxyvitamin D levels.; PloS one; 2020; vol. 15 (no. 9); e0239252 |
Study details
| Study design | Case-control study |
| Trial registration (if reported) | Not reported. |
| Study start date | 09-Mar-2020 |
| Study end date | 19-Jun-2020 |
| COVID-19 prevalence at the time of the study |
Higher prevalence (e.g. during peak of first wave) |
| Aim of the study | To assess the association of circulating 25-hydroxyvitamin D [25(OH)D] levels, a measure of vitamin D status, with positivity for SARS- CoV-2. |
| County/ Geographical location |
US, data included from all 50 states and the District of Columbia. |
| Study setting | Results analysed from clinical laboratory results. |
|
Population description |
The population was recruited from test results from a clinical laboratory. A Quest Diagnostics-wide unique patient identifier was used to match all results of SARS-CoV-2 testing with 25(OH)D results from the preceding 12 months. |
| Inclusion criteria | Had a SARS-CoV-2 test result and vitamin D measurements. |
| Exclusion criteria | Specimens with inconclusive results (one out of two SARS-CoV-2 targets detected) or missing residential zip code data, which are needed to assign race/ethnicity proportions and latitude. |
|
Vitamin D status measurements |
Total 25(OH)D was measured using a chemiluminescent immunoassay (DiaSorin LIAISON1XL 25-hydroxyvitamin D, total) or a laboratory-developed test based on liquid chromatograph/tandem mass spectrometry. The laboratory categorizes 25(OH)D results
<20 ng/mL as deficient, 20-29 ng/mL as suboptimal, and >30 ng/mL as optimal. The laboratory assays are standardized and performed identically throughout Quest Diagnostics. When multiple 25(OH)D results were available, the most recent was used. |
|
Methods used to confirm COVID-19 infection |
All SARS-CoV-2 RNA NAATs were performed by Quest Diagnostics using one of four United States Food and Drug Administration (FDA) Emergency Use Authorized tests (Quest Diagnostics SARS-CoV-2 RNA [COVID-19], Qualitative NAAT; Hologic Panther Fusion SARS-CoV-2 assay; Roche Diagnostics cobas1SARS-CoV-2 test; or Hologic Aptima SARS-CoV-2 assay). We combined results from all four tests due to their very similar sensitivity and specificity.
Analysis was limited to one SARS-CoV-2 result per patient. Patients were considered positive if at least one test result indicated positivity. |
| Intervention | Not applicable. |
| Comparator (where applicable) | Not applicable. |
| Methods for population selection/allocation |
Described above. |
| Methods for case- matching with control |
Not applicable. |
|
Methods of data analysis |
Comparisons of categorical and continuous variables were done by chi-square and t-test as appropriate.
25(OH)D values were binned. For most ethnic groups, 25(OH)D values were grouped into bins of two values, e.g. 20-21 ng/ml. For black non-Hispanic and Hispanic zip codes only, 25(OH)D values were put in bins of 2 values from 20-29 ng/ml and into bins of 5 values after 30 ng/ml, e.g. 30-35 ng/ml, because of the low numbers of people with values over 30ng/ml. Vitamin D was adjusted for seasonality with a model based on a previous 25(OH)D study, using Quest Diagnostics results that fit the present study, according to the authors.
Age was stratified into 2 groups, under 60 years old and 60 years old and above.
Participants did not have specific ethnicities linked to them - their ethnicity in the study was based on their zip code and their likelihood of being a certain ethnicity. Therefore, people were categorised into the following groups “predominately black non-Hispanic”, pre- dominantly Hispanic” and “predominantly white non-Hispanic”. Race/ethnicity proportions were taken as reported by the zip code in the 2018 5-year American Community Services. Zip codes with estimated proportions of black non-Hispanic population over 50% are referred to as “pre-dominantly black non-Hispanic”. The same pattern was followed for “pre-dominantly Hispanic” and “predominantly white non-Hispanic” zip codes.
The correlation between 25(OH)D values and infection were fitted the best by the weighted second-order polynomial regression. Multivariable logistic regression was performed using a stepwise entry criterion of p<0.05, after excluding participants with missing values.
Analyses were performed using SAS Studio 3.6 on SAS 9.4, and R v3.6.1. |
|
Attrition/loss to follow-up |
Participants were excluded for lack of zip code data or inconclusive SARS-CoV-2 results, as specified in the exclusion criteria. No difference between included and excluded participants on infection rates, age and gender.
In the multivariable model, only participants with no missing data were included (n=188,028; 98%). |
| Source of funding | Quest Diagnostics provide salaries to authors JKN, BC, MHK and HWK, and consulting fees for MFH who did not have any addition role in the study design, data collection or analysis, decision to publish, or preparation of the manuscript. |
|
Study limitations (author) |
Testing for SARS-CoV-2 was based on selection factors, including presence and gravity of symptoms and exposure to infected individuals.
High-risk groups, such as healthcare workers and first responders, are also more likely to be tested.
Another limitation is that race/ethnicity estimates were based on aggregate U.S. Census proportions by zip code. |
| There may be many other potentially confounding factors that were neither identified nor controlled for in this study. The multivariable model displayed poor overall fit and correlation statistics, given SARS-CoV-2 can infect anyone. | |
|
Study limitations (reviewer) |
No baseline characteristics table and baseline characteristics poorly reported. Ethnicity was only partly reported - the 3 ethnicities mentioned do not make up the whole cohort.
Concerning the estimated ethnicities, there could be higher positivity in hispanic/black populations in white areas and vice versa not related to vit D status. This estimation also masks whether some ethnicities are more susceptible to COVID-19 than others. |
Study arms
| Entire cohort (N = 191779)
Results were presented as entire cohort. |
Characteristics
Study-level characteristics
| Study (N = 191779) | |
| Age | |
| MedianIQR | 54 (40.4 to 64.7) |
| Gender
Female |
|
| Sample Size | n = 130473 ; % = 68 |
| Ethnicity | |
| Predominantly black non-Hispanic | |
| Sample Size | n = 9529 ; % = 5 |
| Predominantly Hispanic | |
| Sample Size | n = 26242 ; % = 13.7 |
| Predominantly white non-Hispanic | |
| Sample Size | n = 112281 ; % = 58.5 |
| Comorbidities |
| Study (N = 191779) | |
| Custom value | NA |
| BMI | |
| Custom value | NA |
| Use of immune suppressing treatments | |
| Custom value | NA |
| Socioeconomic status | |
| Custom value | NA |
| Previous history of COVID-19 | |
| Custom value | NA |
| Other supplement use | |
| Custom value | NA |
| Timing of vitamin D measurements | |
| Custom value | NA |
| Shielding status | |
| Custom value | NA |
| Living in care homes | |
| Custom value | NA |
Outcomes
Association between lower SARS-CoV-2 positivity rates and higher circulating 25(OH)D levels
Odds ratios (ORs) are presented as risk of infection per ng/ml. OR = 1 is no difference in risk per ng/ml, lower ORs indicate a lower risk of infection at a higher 25(OH)D.
| Entire cohort vs Entire cohort | |
| N1 = 191779 | |
| 25(OH)D level
Polarity: Lower values are better |
| Entire cohort vs Entire cohort | |
| N1 = 191779 | |
| Unadjusted | |
| Odds ratio/95% CI | 0.98 (0.98 to 0.98) |
| Adjusted
Adjusted for vitamin D seasonality. Only included 188,028 participants without missing values. |
|
| Odds ratio/95% CI | 0.98 (0.98 to 0.99) |
| Predominantly black non-Hispanic
All other zip codes apart from predominantly black/Hispanic were used as reference. Polarity: Lower values are better |
|
| Unadjusted | |
| Odds ratio/95% CI | 2.04 (1.93 to 2.17) |
| Adjusted
Adjusted for vitamin D seasonality. |
|
| Odds ratio/95% CI | 2.03 (1.91 to 2.15) |
| Predominantly Hispanic
All other zip codes apart from predominantly black/Hispanic were used as reference. Polarity: Lower values are better |
|
| Unadjusted | |
| Odds ratio/95% CI | 1.61 (1.54 to 1.67) |
| Adjusted
Adjusted for vitamin D seasonality. |
|
| Odds ratio/95% CI | 1.95 (1.87 to 2.04) |
| Section | Question | Answer |
| Study participation | Summary Study participation | High risk of bias
(Baseline characteristics not adequately described and not clear where data has initially come from.) |
| Section | Question | Answer |
|
Study Attrition |
Study Attrition Summary | Low risk of bias
(Some loss to follow-up but only 2%. Loss to follow-up in this case is participants not included in modelling due to missing data. Loss is small and random.) |
| Prognostic factor measurement | Prognostic factor Measurement Summary | Low risk of bias
(Measurement of 25(OH)D conducted by one company.) |
| Outcome Measurement | Outcome Measurement Summary |
Low risk of bias |
| Study Confounding | Study Confounding Summary | High risk of bias
(Very few confounders measured, only season vitamin D was measured, age, gender and ethnicity.) |
|
Statistical Analysis and Reporting |
Statistical Analysis and Presentation Summary |
High risk of bias
(Presentation of data is not adequate, fundamental baseline characteristics table is missing. The number of participants in the reported ethnic groups do not add up to the total number analysed in the unadjusted nor the adjusted models. Many people would have their ethnicity incorrectly classified due to the way the study used zip code as a proxy. The only confounder that the adjusted model adjusts for is the season vitamin D was measured. 25(OH)D concentrations are binned, losing information about the linear relationship between 25(OH)D and risk of infection.) |
|
Overall risk of bias and directness |
Risk of Bias |
High
(Data planning presentation poor and missing important details, very few confounders measured, only season vitamin D was measured, age, gender and ethnicity. Baseline characteristics not adequately described and not clear where data has initially come from.) |
|
Directness |
Indirectly applicable
(Vitamin status data was historical (preceding 12 months) where vitamin level may have changed before SAR-CoV- 2 testing. Also, the outcome is SAR-CoV-2 positive, not COVID-19) |
| Macaya, 2020 |
| Bibliographic Reference | Macaya, Fernando; Espejo Paeres, Carolina; Valls, Adrian; Fernandez-Ortiz, Antonio; Gonzalez Del Castillo, Juan; Martin-Sanchez, F Javier; Runkle, Isabelle; Rubio Herrera, Miguel Angel; Interaction between age and vitamin D deficiency in severe COVID-19 infection.;
Nutricion hospitalaria; 2020; vol. 37 (no. 5); 1039-1042 |
Study details
| Study design | Case series |
| Trial registration (if reported) | Not reported. |
| Aim of the study | The aim of this study was to explore the association between vitamin D deficiency and the development of severe COVID-19. |
| County/ Geographical location |
Madrid, Spain. |
| Study setting | Emergency department of a tertiary hospital. |
| Population description | A cohort of consecutive patients admitted with COVID-19 between 3rd March and 31st March 2020. |
| Inclusion criteria | 1) a positive reverse-transcriptase polymerase chain reaction for SARS-CoV-2, and 2) an available measurement of serum 25- hydroxyvitamin D (25(OH)D) (chemiluminescent immunoassay, Abbott Diagnostics) at admission or within the 3 previous months. |
| Exclusion criteria | None reported. |
| Vitamin D status measurements | Described in the inclusion criteria. |
| Methods used to confirm COVID-19 infection |
Described in the inclusion criteria. |
| Intervention | Not applicable. |
| Comparator (where applicable) | Not applicable. |
| Methods for population selection/allocation |
Described above. |
| Methods for case- matching with control |
Not applicable. |
|
Methods of data analysis |
For continuous variables, a univariate analysis with a t-test or rank-sum test was conducted. For categorical variables, a chi-square test or Fisher’s exact test was used. The association between the composite COVID-19 outcome (death, admission to the intensive care unit, and/or need for higher oxygen flow than provided by a nasal cannula) and 25(OH)D level was conducted by multivariable logistic regression adjusting for obesity, sex, age and advanced kidney disease. Age group-specific analyses defined by percentile 50, were conducted. All analyses were conducted using Stata 13. |
| Attrition/loss to follow-up | 11/91 had not completed follow-up by the time statistical analyses were conducted. The study did not explain how it dealt with attrition, but based on the percentages, it appears they only includes people who completed follow-up (n=80). |
| Source of funding | The Fundación interhospitalaria para la Investigación Cardiovascular covered the publication charge for this study. |
| Study limitations (Author) | Small sample size.
More likely to have vitamin D measurements for the elderly, in patients with comorbidities, renal disease and obesity. |
|
Study limitations (reviewer) |
Retrospective study.
Did not account for attrition statistically, only excluded participants from all analyses.
Did not describe how many vitamin D measurements were up to 3 months old or taken on admission. As patients were admitted in March, measurements from 3 months previously would have been taken in mid-winter where vitamin D levels are lower.
There could be differences in the clinical decisions made before hospitalisation and ICU admission due to this study not being in the UK and changes over the course of the pandemic |
Study arms
| Non-severe COVID-19 (N = 49) |
| Severe COVID-19 (N = 31)
Defined by the composite endpoint: death, admission to the intensive care unit, and/or need for higher oxygen flow than that provided by a nasal cannula. |
Characteristics
Arm-level characteristics
| Non-severe COVID-19 (N = 49) | Severe COVID-19 (N = 31) | |
| Age | ||
| MedianIQR | 63 (50 to 72) | 75 (66 to 84) |
| Under 67 years (n=40)
Non-severe, n=30; Severe, n=10 |
||
| MedianIQR | 51.5 (44 to 63) | 54.5 (45 to 66) |
| Gender
Male |
||
| Sample Size | n = 14 ; % = 29 | n = 21 ; % = 68 |
| Under 67 years (n=40)
Non-severe, n=30; Severe, n=10 |
||
| Sample Size | n = 6 ; % = 20 | n = 8 ; % = 80 |
| Ethnicity | ||
| Custom value | NA | NA |
| Comorbidities | ||
| BMI
Number of people ≥ 30 kg/m2 |
||
| Sample Size | n = 12 ; % = 24 | n = 14 ; % = 45 |
| Under 67 years (n=40)
Non-severe, n=30; Severe, n=10 |
||
| Sample Size | n = 5 ; % = 17 | n = 5 ; % = 50 |
| Use of immune suppressing treatments | ||
| Custom value | NA | NA |
| Socioeconomic status | ||
| Custom value | NA | NA |
| Previous history of COVID-19 | ||
| Custom value | NA | NA |
| Other supplement use | ||
| Custom value | NA | NA |
| Timing of vitamin D measurements |
| Non-severe COVID-19 (N = 49) | Severe COVID-19 (N = 31) | |
| Custom value | NA | NA |
| Shielding status | ||
| Custom value | NA | NA |
| Living in care homes | ||
| Custom value | NA | NA |
| Smoking history | ||
| Sample Size | n = 6 ; % = 12 | n = 7 ; % = 23 |
| Under 67 years (n=40)
Non-severe, n=30; Severe, n=10 |
||
| Sample Size | n = 4 ; % = 13 | n = 1 ; % = 10 |
| Hypertension | ||
| Sample Size | n = 20 ; % = 57 | n = 30 ; % = 67 |
| Under 67 years (n=40)
Non-severe, n=30; Severe, n=10 |
||
| Sample Size | n = 9 ; % = 30 | n = 5 ; % = 50 |
| Diabetes | ||
| Sample Size | n = 20 ; % = 41 | n = 12 ; % = 39 |
| Under 67 years (n=40)
Non-severe, n=30; Severe, n=10 |
||
| Sample Size | n = 8 ; % = 27 | n = 2 ; % = 20 |
| Cardiac disease | ||
| Sample Size | n = 11 ; % = 22 | n = 8 ; % = 26 |
| Under 67 years (n=40)
Non-severe, n=30; Severe, n=10 |
||
| Sample Size | n = 2 ; % = 7 | n = 1 ; % = 10 |
| Advanced chronic kidney disease
CKD-EPI < 30 mL/min/m2 |
||
| Sample Size | n = 12 ; % = 24 | n = 14 ; % = 45 |
| Under 67 years (n=40)
Non-severe, n=30; Severe, n=10 |
| Non-severe COVID-19 (N = 49) | Severe COVID-19 (N = 31) | |
| Sample Size | n = 5 ; % = 17 | n = 5 ; % = 50 |
| Chronic respiratory disease | ||
| Sample Size | n = 8 ; % = 16 | n = 5 ; % = 16 |
| Under 67 years (n=40)
Non-severe, n=30; Severe, n=10 |
||
| Sample Size | n = 1 ; % = 3 | n = 0 ; % = 0 |
| Vitamin D supplements
Number of people who took vitamin D supplements |
||
| Sample Size | n = 24 ; % = 49 | n = 20 ; % = 65 |
| Under 67 years (n=40)
Non-severe, n=30; Severe, n=10 |
||
| Sample Size | n = 15 ; % = 50 | n = 6 ; % = 60 |
| serum 25(OH)D (ng/mL) | ||
| MedianIQR | 19 (9 to 30) | 13 (8 to 25) |
| Under 67 years (n=40)
Non-severe, n=30; Severe, n=10 |
||
| MedianIQR | 22 (11 to 31) | 11 (9 to 12) |
| Vitamin D deficiency
25(OH)D <20ng/mL |
||
| Sample Size | n = 25 ; % = 51 | n = 20 ; % = 65 |
| Under 67 years (n=40)
Non-severe, n=30; Severe, n=10 |
||
| Sample Size | n = 13 ; % = 43 | n = 10 ; % = 100 |
Outcomes
Risk of reaching composite outcome
This model associated variables with the composite outcome in a multivariable regression model. The variables included in the model were obesity, cardiac disease, age, sex, advanced chronic kidney disease. The variables shown below are those reported in the study.
|
Severe COVID-19 vs Non-severe COVID-19 |
|
| N1 = 31, N2 = 49 | |
| Vitamin D deficiency
Polarity: Not set |
|
| Odds ratio/95% CI | 3.2 (0.9 to 11.4) |
| Over 75 years
Polarity: Not set |
|
| Odds ratio/95% CI | 10.4 (2 to 54.8) |
| Gender
Male Polarity: Not set |
|
| Odds ratio/95% CI | 6.2 (2 to 19.5) |
| Section | Question | Answer |
| Study participation | Summary Study participation | Moderate risk of bias
(Ethnicity, socioeconomic status, and use of immune suppressing treatments not included.) |
| Study Attrition | Study Attrition Summary | High risk of bias
(No information for people who had dropped out or why they had dropped out.) |
|
Prognostic factor measurement |
Prognostic factor Measurement Summary |
Low risk of bias
(Appropriate methods for measuring vitamin D were used. The study reported that there was missing outcome data but not for vitamin D measurements. Missing outcome data is considered in other domains and has not been considered as a factor for this domain.) |
| Outcome Measurement | Outcome Measurement Summary | High risk of bias
(Difficult to judge when follow-up was halted. The study only says “until data analysis was conducted”. No detail on how outcomes were measured are reported.) |
| Study Confounding | Study Confounding Summary | High risk of bias
(Ethnicity, socioeconomic status, and use of immune suppressing treatments not included.) |
| Statistical Analysis and Reporting | Statistical Analysis and Presentation Summary | High risk of bias
(The results are not reported in full, only those that are critical to the paper or that are significant.) |
| Section | Question | Answer |
|
Overall risk of bias and directness |
Risk of Bias |
High
(No information for people who had dropped out or why they had dropped out. Ethnicity, socioeconomic status, and use of immune suppressing treatments not included. The results are not reported in full, only those that are critical to the paper or that are significant.) |
| Directness | Directly applicable (There could be differences in the clinical decisions made before hospitalisation and ICU admission due to this study not being in the UK and changes over the course of the pandemic) |
| Meltzer, 2020 |
| Bibliographic
Reference |
Meltzer, David O; Best, Thomas J; Zhang, Hui; Vokes, Tamara; Arora, Vineet; Solway, Julian; Association of Vitamin D Status and Other
Clinical Characteristics With COVID-19 Test Results.; JAMA network open; 2020; vol. 3 (no. 9); e2019722 |
Study details
| Study design | Retrospective cohort study |
| Trial registration (if reported) | Not reported. |
| Aim of the study | To assess if people would be more likely to test positive for COVID-19 if they had deficient vitamin D level measurement before COVID- 19 testing. |
| County/ Geographical location |
Chicago, US. |
| Study setting | University of Chicago Medicine. |
|
Population description |
Data was obtained for all 4313 patients tested for COVID-19 at the university between 3rd March 2020 and 10th April 2020.
Age, sex, and race/ethnicity were also obtained from the electronic health record. The most recent data was obtained during the study period up to 14 days before COVID-19 testing to calculate body mass index and the following International Statistical Classification of |
| Diseases, Tenth Revision, Clinical Modification (ICD-10-CM)-based Elixhauser comorbidity clusters potentially related to COVID-19 and/or vitamin D metabolism: hypertension, diabetes, chronic pulmonary disease, pulmonary circulation disorders, depression, immunosuppression, liver disease, and chronic kidney disease. | |
| Inclusion criteria | Tested for COVID-19 in the study period and had a vitamin D measurement within the past 12 months. |
| Exclusion criteria | People who had vitamin D testing within 14 days of COVID-19 testing in case the infection confounded the vitamin D results. |
|
Vitamin D status measurements |
Vitamin D was measured up to a year before the COVID-19 test. The authors were aware that levels may have changed, therefore they estimated whether the participants would likely be still sufficient/deficient based on their vitamin D concentration at measurement and if their vitamin D supplementation had changed since that measurement was taken. Participants were then categorised into 1 of 4 groups: : likely deficient (last level deficient and treatment not increased), likely sufficient (last level not deficient and treatment not decreased), and 2 groups with uncertain deficiency (last level deficient and treatment increased, and last level not deficient and treatment decreased). A more detailed explanation on how the study categorised participants is below:
Patients were deemed to be vitamin D deficient if their most recent serum vitamin D levels within 1 year before their first COVID-19 tests were less than 20 ng/mL for 25-hydroxycholecalciferol (to convert to nanomoles per litre, multiply by 2.496) or less than 18 pg/ml for 1,25-dihydroxycholecalciferol (to convert to picomoles per litre, multiply by 2.4) and deemed not deficient if their most recent levels were equal to or greater than 20 ng/mL or equal to or greater than 18 pg/ml, respectively. Vitamin D treatment was defined by report in the electronic health record of vitamin D either in the patient medication list or prescription orders. Vitamin D3 dosing was defined based on most recent daily dose recorded over the past year excluding the 14 days before testing: none, 1 to 1000 IU or a multivitamin, 2000 IU, or greater than or equal to 3000 IU. Indicators for treatment with vitamin D2 and calcitriol were also included. Possible changes in patients’ vitamin D treatment after the time of their last vitamin D level were accounted for by categorizing changes in treatment between the date of the last vitamin D level and 14 days before COVID-19 testing as increased, unchanged, or decreased according to the following ordering: calcitriol was considered the highest treatment category followed in decreasing order by greater than or equal to 3000 IU D3, 2000 IU D3, D2, 1-1000 IU D3 or multivitamin, and no vitamin D. The data was then combined on last vitamin D level measurements with changes in treatment after that last vitamin D level to assign each patient to 1 of 4 categories reflecting their likelihood of being vitamin D deficient at the time of COVID-19 testing: likely deficient (last level deficient and treatment not increased), likely sufficient (last level not deficient and treatment not decreased), and 2 groups with uncertain deficiency (last level deficient and treatment increased, and last level not deficient and treatment decreased). |
| Methods used to confirm COVID-19 infection | COVID-19 test status was determined by any positive COVID-19 polymerase chain reaction test result, with the Centers for Disease Control and Prevention or Viacor test used until in-house testing with the test from Roche (cobas) began on March 15, 2020. Because of test supply, testing at UCM was limited to persons presenting with potential symptoms of COVID-19 admitted to the hospital or health care workers with COVID-19 symptoms and exposure. |
| Intervention | Not applicable. |
| Comparator (where applicable) | Not applicable. |
| Methods for population selection/allocation |
Described above. |
| Methods for case- matching with control |
Not applicable. |
|
Methods of data analysis |
Basic descriptive statistics were reviewed for all variables. In comparing patients with last vitamin D levels that were deficient and patients with last levels that were not deficient, Fisher’s exact test was used for binary variables and the t-test for continuous variables. A multivariable generalized linear model with binomial residuals and log-link function was estimated with the covariates noted above. A piecewise linear spline with a single knot at 50 improved model fit over models with unadjusted age or more complex parameterizations. Statistical significance was defined as P < 0.05. All tests were 2-tailed. |
|
Source of funding |
Supported by the Learning Health Care System Core of the University of Chicago/Rush University Institute for Translational Medicine (ITM) Clinical and Translational Science Award (ITM 2.0: Advancing Translational Science in Metropolitan Chicago, UL1TR002389, Solway, Contact PI) and the African American Cardiovascular Pharmacogenetic Consortium (U54-MD010723, Meltzer). |
|
Study limitations (authors) |
Vitamin D deficiency may be a consequence associated with a range of chronic health conditions or behavioural factors that plausibly increase COVID-19 risk. The authors defend the results by saying they are robust and include a broad set of demographic and comorbidity indicators that have either physiological reasons for consideration or have been suggested to influence COVID-19 outcomes.
Neither patients who were deficient in vitamin D and had increased treatment nor patients who were not deficient in vitamin D who had decreased treatment were more likely than patients who were not vitamin D deficient and at least maintained their current treatment (ie, had nondeficient status) to test positive for COVID-19. If the observed association were due to confounding by behavioural or other health factors, such associations might have been expected, although our limited sample size might be inadequate to identify such effects.
The data are limited to those available in the UCM electronic health record. Patterns of vitamin D screening, treatment, or COVID-19 testing at UCM or in other institutions might have somehow selected for patients who induced an association between observed vitamin D status and testing positive for COVID-19.
They considered whether specific versions of this broad range of alternative hypotheses might explain our findings, including the idea that vitamin D treatment not recorded at UCM prior to COVID-19 testing might have biased our results. Analysis of medication information reported at the time of COVID-19 testing did not identify changes in vitamin D dosing.
Only a few individuals received higher doses of vitamin D3 or had relative high vitamin D levels, limiting power to assess whether vitamin D dose or levels are associated with the likelihood of COVID-19. Calcitriol was also included in defining vitamin D deficiency and |
| included patients treated with vitamin D2 or calcitriol, which are often used in patients with chronic kidney disease or hypoparathyroidism. Sensitivity analysis were robust at omitting these patients.
The sample is overrepresented in persons with vitamin D deficiency because of the large number of African American individuals, adults with chronic illness, and health care workers, all living in a northern city and exposed to COVID-19 during winter. Vitamin D deficiency is highly prevalent in the US but could be a smaller risk factor in other populations. |
|
| Study limitations (reviewer) | Estimations of vitamin D status based on supplementation may be incorrect as it relies on medicine compliance. |
Study arms
| Full cohort (N = 489) |
Characteristics
Study-level characteristics
| Study (N = 489) | |
| Age | |
| Mean/SD | 49.2 (18.4) |
| Vitamin D deficient
<20 ng/mL |
|
| Mean/SD | 45.9 (17.6) |
| Vitamin D sufficient
≥20 ng/mL |
|
| Mean/SD | 51 (18.6) |
| Gender
Female |
|
| Sample Size | n = 366 ; % = 75 |
| Vitamin D deficient
<20 ng/mL |
|
| Sample Size | n = 133 ; % = 77 |
| Study (N = 489) | |
| Vitamin D sufficient
≥20 ng/mL |
|
| Sample Size | n = 233 ; % = 74 |
| Ethnicity
BAME |
|
| Sample Size | n = 331 ; % = 68 |
| Vitamin D deficient
<20 ng/mL |
|
| Sample Size | n = 142 ; % = 83 |
| Vitamin D sufficient
≥20 ng/mL |
|
| Sample Size | n = 189 ; % = 60 |
| Comorbidities | |
| BMI | |
| Mean | 29.8 |
| Vitamin D deficient
<20 ng/mL |
|
| Mean | 30.4 |
| Vitamin D sufficient
≥20 ng/mL |
|
| Mean | 29.4 |
| Use of immune suppressing treatments | |
| Custom value | NA |
| Socioeconomic status | |
| Custom value | NA |
| Previous history of COVID-19 | |
| Custom value | NA |
| Other supplement use | |
| Custom value | NA |
| Study (N = 489) | |
| Timing of vitamin D measurements
Number evaluated in the past year |
|
| Sample Size | n = 489 ; % = 100 |
| Shielding status | |
| Custom value | NA |
| Living in care homes | |
| Custom value | NA |
| Vitamin D sufficiency
Number of people who fall into each sufficiency category |
|
| Likely deficient
Answer was yes to most recent vitamin D level within 1 year being deficient (<20 ng/mL); dose was stable or decreased after last visit. Vitamin D dose was rank ordered as follows: calcitriol > 3000+ IU D3 > 2000 IU D3 > D2 > 1-1000 IU D3/multivitamin > no vitamin D. |
|
| Sample Size | n = 124 ; % = 25 |
| Uncertain deficiency
Answer was yes to most recent vitamin D level within 1 year being deficient (<20 ng/mL); dose was increased after last visit. Vitamin D dose was rank ordered as follows: calcitriol > 3000+ IU D3 > 2000 IU D3 > D2 > 1-1000 IU D3/multivitamin > no vitamin D. |
|
| Sample Size | n = 48 ; % = 10 |
| Uncertain deficiency
Answer was no to most recent vitamin D level within 1 year being deficient (<20 ng/mL); dose was decreased after last visit. Vitamin D dose was rank ordered as follows: calcitriol > 3000+ IU D3 > 2000 IU D3 > D2 > 1-1000 IU D3/multivitamin > no vitamin D. |
|
| Sample Size | n = 30 ; % = 5 |
| Likely sufficient
Answer was no to most recent vitamin D level within 1 year being deficient (<20 ng/mL); dose was stable or increased after last visit. Vitamin D dose was rank ordered as follows: calcitriol > 3000+ IU D3 > 2000 IU D3 > D2 > 1-1000 IU D3/multivitamin > no vitamin D. |
|
| Sample Size | n = 287 ; % = 59 |
Arm-level characteristics
| Full cohort (N = 489) | |
| Hypertension | |
| Sample Size | n = 261 ; % = 53 |
| Full cohort (N = 489) | |
| Vitamin D deficient
<20 ng/mL |
|
| Sample Size | n = 89 ; % = 52 |
| Vitamin D sufficient
≥20 ng/mL |
|
| Sample Size | n = 172 ; % = 54 |
| Diabetes | |
| Sample Size | n = 137 ; % = 28 |
| Vitamin D deficient
<20 ng/mL |
|
| Sample Size | n = 51 ; % = 30 |
| Vitamin D sufficient
≥20 ng/mL |
|
| Sample Size | n = 86 ; % = 27 |
| Chronic pulmonary disease | |
| Sample Size | n = 117 ; % = 24 |
| Vitamin D deficient
<20 ng/mL |
|
| Sample Size | n = 43 ; % = 25 |
| Vitamin D sufficient
≥20 ng/mL |
|
| Sample Size | n = 74 ; % = 23 |
| Pulmonary circulation disorders | |
| Sample Size | n = 20 ; % = 4 |
| Vitamin D deficient
<20 ng/mL |
|
| Sample Size | n = 9 ; % = 5 |
| Vitamin D sufficient
≥20 ng/mL |
|
| Sample Size | n = 11 ; % = 3 |
| Depression |
| Full cohort (N = 489) | |
| Sample Size | n = 119 ; % = 24 |
| Vitamin D deficient
<20 ng/mL |
|
| Sample Size | n = 45 ; % = 26 |
| Vitamin D sufficient
≥20 ng/mL |
|
| Sample Size | n = 74 ; % = 23 |
| Chronic kidney disease | |
| Sample Size | n = 116 ; % = 24 |
| Vitamin D deficient
<20 ng/mL |
|
| Sample Size | n = 36 ; % = 21 |
| Vitamin D sufficient
≥20 ng/mL |
|
| Sample Size | n = 80 ; % = 25 |
| Liver disease | |
| Sample Size | n = 56 ; % = 11 |
| Vitamin D deficient
<20 ng/mL |
|
| Sample Size | n = 17 ; % = 10 |
| Vitamin D sufficient
≥20 ng/mL |
|
| Sample Size | n = 39 ; % = 12 |
| Comorbidities with immunosuppression | |
| Sample Size | n = 105 ; % = 21 |
| Vitamin D deficient
<20 ng/mL |
|
| Sample Size | n = 36 ; % = 21 |
| Vitamin D sufficient
≥20 ng/mL |
|
| Sample Size | n = 69 ; % = 22 |
| Full cohort (N = 489) | |
| Most recent active vitamin D treatment before COVID-19 test
Participants are listed by vitamin D treatment, and vitamin D sufficiency below |
|
| None, vitamin D deficient | |
| Sample Size | n = 80 ; % = 47 |
| None, vitamin D sufficient | |
| Sample Size | n = 132 ; % = 42 |
| 1-1000 IU D3/multivitamin, vitamin D deficient | |
| Sample Size | n = 28 ; % = 16 |
| 1-1000 IU D3/multivitamin, vitamin D sufficient | |
| Sample Size | n = 85 ; % = 27 |
| 2000 IU D3, vitamin D deficient | |
| Sample Size | n = 7 ; % = 4 |
| 2000 IU D3, vitamin D sufficient | |
| Sample Size | n = 53 ; % = 17 |
| ≥3000 IU D3, vitamin D deficient | |
| Sample Size | n = 10 ; % = 6 |
| ≥3000 IU D3, vitamin D sufficient | |
| Sample Size | n = 10 ; % = 3 |
| D2, vitamin D deficient | |
| Sample Size | n = 44 ; % = 26 |
| D2, vitamin D sufficient | |
| Sample Size | n = 32 ; % = 10 |
| Calcitriol, vitamin D deficient
The study reports <5 people in this group and to preserve confidentiality, the actual frequency counts were masked |
|
| Sample Size | n = 5 ; % = 2 |
| Calcitriol, vitamin D sufficient | |
| Sample Size | n = 5 ; % = 2 |
| COVID-19 positive |
| Full cohort (N = 489) | |
| Vitamin D deficient
<20 ng/mL |
|
| Sample Size | n = 32 ; % = 19 |
| Vitamin D sufficient
≥20 ng/mL |
|
| Sample Size | n = 39 ; % = 12 |
Outcomes
Multivariable Association of Vitamin D Deficiency and Treatment with Testing Positive for COVID-19
Age, sex, ethnicity, employee status, vitamin D status, comorbidity indicators and BMI were in included in this model.
| Full cohort vs Full cohort | |
| N1 = 489 | |
| Most recent vitamin D <20 ng/mL
Polarity: Lower values are better |
|
| Likely deficient
Answer was yes to most recent vitamin D level within 1 year being deficient (<20 ng/mL); dose was stable or decreased after last visit. Vitamin D dose was rank ordered as follows: calcitriol > 3000+ IU D3 > 2000 IU D3 > D2 > 1-1000 IU D3/multivitamin > no vitamin D. |
|
| Sample Size | n = 124 ; % = 25, n = 287 ;
% = 59 |
| Odds ratio/95% CI | 1.77 (1.12 to 2.81) |
| Uncertain deficiency
Answer was yes to most recent vitamin D level within 1 year being deficient (<20 ng/mL); dose was increased after last visit. Vitamin D dose was rank ordered as follows: calcitriol > 3000+ IU D3 > 2000 IU D3 > D2 > 1-1000 IU D3/multivitamin > no vitamin D. |
|
| Sample Size | n = 48 ; % = 10, n = 287 ;
% = 59 |
| Odds ratio/95% CI | 1.1 (0.49 to 2.43) |
| Uncertain deficiency
Answer was no to most recent vitamin D level within 1 year being deficient (<20 ng/mL); dose was decreased after last visit. Vitamin D dose was rank ordered as follows: calcitriol > 3000+ IU D3 > 2000 IU D3 > D2 > 1-1000 IU D3/multivitamin > no vitamin D. |
| Full cohort vs Full cohort | |
| N1 = 489 | |
| Sample Size | n = 30 ; % = 5, n = 287 ; %
= 59 |
| Odds ratio/95% CI | 1.09 (0.43 to 2.82) |
| Likely sufficient [reference]
Answer was no to most recent vitamin D level within 1 year being deficient (<20 ng/mL); dose was stable or increased after last visit. Vitamin D dose was rank ordered as follows: calcitriol > 3000+ IU D3 > 2000 IU D3 > D2 > 1-1000 IU D3/multivitamin > no vitamin D. |
|
| Sample Size | n = 287 ; % = 59, n = 287 ;
% = 59 |
| Odds ratio | 1 |
| Age
linear spline Polarity: Lower values are better |
|
| 50+ | |
| Sample Size | n = 260 ; % = 53 |
| Odds ratio/95% CI | 1.05 (1.01 to 1.09) |
| 50+ | |
| Sample Size | n = 229 ; % = 47 |
| Odds ratio/95% CI | 1.02 (1 to 1.05) |
| Sex
Polarity: Lower values are better |
|
| Male [reference] | |
| Sample Size | n = 123 ; % = 25 |
| Odds ratio | 1 |
| Female | |
| Sample Size | n = 366 ; % = 75 |
| Odds ratio/95% CI | 0.87 (0.52 to 1.44) |
| Race
Polarity: Lower values are better |
|
| White [reference] | |
| Sample Size | n = 158 ; % = 32 |
| Full cohort vs Full cohort | |
| N1 = 489 | |
| Odds ratio | 1 |
| Other than white | |
| Sample Size | n = 331 ; % = 68 |
| Odds ratio/95% CI | 2.54 (1.26 to 5.12) |
| Comorbidities
Polarity: Lower values are better |
|
| Hypertension | |
| Sample Size | n = 261 ; % = 53 |
| Odds ratio/95% CI | 1.08 (0.6 to 1.97) |
| Diabetes | |
| Sample Size | n = 137 ; % = 28 |
| Odds ratio/95% CI | 0.78 (0.49 to 1.26) |
| Chronic pulmonary disease | |
| Sample Size | n = 117 ; % = 24 |
| Odds ratio/95% CI | 0.91 (0.55 to 1.52) |
| Pulmonary circulation disorders | |
| Sample Size | n = 20 ; % = 4 |
| Odds ratio/95% CI | 0.64 (0.23 to 1.79) |
| Depression | |
| Sample Size | n = 119 ; % = 24 |
| Odds ratio/95% CI | 1.22 (0.74 to 2.02) |
| Chronic kidney disease | |
| Sample Size | n = 116 ; % = 24 |
| Odds ratio/95% CI | 0.8 (0.49 to 1.32) |
| Liver disease | |
| Sample Size | n = 56 ; % = 11 |
| Odds ratio/95% CI | 0.99 (0.47 to 2.08) |
| Full cohort vs Full cohort | |
| N1 = 489 | |
| Comorbidities with immunosuppression | |
| Sample Size | n = 105 ; % = 21 |
| Odds ratio/95% CI | 0.39 (0.2 to 0.76) |
| BMI
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 1.02 (1 to 1.05) |
| Mean | 29.8 |
| Section | Question | Answer |
|
Study participation |
Summary Study participation |
Moderate risk of bias
(Adequate participation from eligible participants. People only included if they had a recent vitamin D test. People were excluded if they had had a vitamin D test 14 days before they had their COVID-19 test as positivity or reasons for presenting with symptoms may affect vitamin D levels. This may bias results.) |
| Study Attrition | Study Attrition Summary | Low risk of bias
(No attrition reported.) |
|
Prognostic factor measurement |
Prognostic factor Measurement Summary |
High risk of bias
Moderate risk of bias (Vitamin D status at the time of the study was estimated depending on participant’s status at time of vitamin D testing and if their supplements had changed since then. However, as sensitivity analyses showed that removing people who had less certain vitamin D status did not change the results of the multivariable analyses, preventing this domain from being classed as high risk of bias.) |
| Outcome Measurement | Outcome Measurement Summary | Low risk of bias
(COVID-19 tested by RT-PCR at the same site) |
| Section | Question | Answer |
|
Study Confounding |
Study Confounding Summary | Moderate risk of bias
(Some confounders missing from study, such as use of immune suppressing treatments. Socioeconomic status was included in some modelling but only reported as not different to the model that was reported.) |
|
Statistical Analysis and Reporting |
Statistical Analysis and Presentation Summary |
Low risk of bias
(Multivariable analysis allows adjustment for confounders. Model covariates was chosen based on comorbidity clusters potentially related to COVID-19 and/or vitamin D metabolism as listed in the International Statistical Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM)-based Elixhauser.) |
|
Overall risk of bias and directness |
Risk of Bias |
Moderate
(Some confounders not reported, vitamin D status estimated but sensitivity analyses show no difference between models, only people with recent vitamin D test included which restricts the pool to people who have presented with vitamin D deficiency or symptoms of that or related conditions.) |
| Directness | Partially applicable
(Historical vitamin D measurements used) |
| Merzon, 2020 |
| Bibliographic Reference | Merzon, Eugene; Tworowski, Dmitry; Gorohovski, Alessandro; Vinker, Shlomo; Golan Cohen, Avivit; Green, Ilan; Frenkel-Morgenstern, Milana; Low plasma 25(OH) vitamin D level is associated with increased risk of COVID-19 infection: an Israeli population-based study.; The
FEBS journal; 2020 |
Study details
| Study design | Case-control study |
| Trial registration (if reported) | Not reported. |
| Study start date | 01-Feb-2020 |
| Study end date | 30-Apr-2020 |
| Aim of the study | To determine associations between low plasma 25(OH)D and the risk of COVID-19 infection and hospitalization using population-based data. |
| County/ Geographical location |
Leumit Health Services, Israel. |
| Study setting | Community |
| Population description | 14,000 people of the Leumit Health Services who were tested for COVID-19 in the study period were eligible. |
| Inclusion criteria | At least one previous blood test for plasma vitamin D (25(OH)D). |
| Exclusion criteria | Not reported. |
|
Vitamin D status measurements |
Blood was transported on ice and processed within 4 hours of collection using DiaSorin Chemiluminescence assay.
‘Suboptimal’ or ‘low’ plasma 25(OH)D level was defined as plasma 25-hydroxyvitamin D, or 25(OH)D, concentration below the level of 30 ng/ml. |
| Methods used to confirm COVID-19 infection | Referrals for viral tests were according to Israeli Ministry of Health guidelines by physicians based on symptoms. Tests were run using the AllplexTM 2019-nCoV Assay (Seegene Inc., Seoul, Korea). |
| Intervention | Not applicable |
| Comparator (where applicable) | Not applicable |
| Methods for case- matching with control |
Not applicable |
|
Methods of data analysis |
Descriptive statistics compared demographic characteristics between COVID-19 positive and COVID-19 negative participants. Continuous variables were reported as means (95% CI) and compared using student’s t-test for normally distributed data. Fisher’s exact or chi-square test was used for categorical variables, displayed as counts and percentages.
Univariate analyses were conducted to assess the association between baseline characteristics and COVID-19 infection and hospitalisation. Multivariable analyses assessed the association between 25(OH)D levels and COVID-19 infection and hospitalisation, adjusting for demographic variables, and psychiatric and somatic disorders. These were reported as odds ratios (ORs) and 95% CIs. |
| Differences were considered significant at p=0.05.
All analyses were conducted using Stata 12. |
|
| Attrition/loss to follow-up | No attrition reported. |
| Source of funding | The study was funded by COVID-19 Data Sciences Institute (DSI) grant (for MFM, #247017). All authors have indicated they have no financial relationships relevant to this manuscript to disclose. |
| Study limitations (Author) | Retrospective database design.
Vitamin D levels were tested according to the presentation of symptoms, and not according to population-wide testing |
| Study limitations (reviewer) | Historic vitamin D measurements are unlikely to reflect actual status at time of COVID-19 test. |
Study arms
| COVID-19 positive (N = 782) |
| COVID-19 negative (N = 7025) |
Characteristics
Arm-level characteristics
| COVID-19 positive (N = 782) | COVID-19 negative (N = 7025) | |
| Age | ||
| Mean/95% CI | 35.58 (34.29 to 36.67) | 47.35 (46.87 to 47.85) |
| Gender
Male |
||
| Sample Size | n = 385 ; % = 49.23 | n = 2849 ; % = 40.56 |
| Ethnicity | ||
| Custom value | NA | NA |
| Comorbidities |
| COVID-19 positive (N = 782) | COVID-19 negative (N = 7025) | |
| Low vitamin D level
Plasma 25(OH)D <30 ng/ml |
||
| Sample Size | n = 703 ; % = 89.9 | n = 5965 ; % = 84.91 |
| Smoking | ||
| Sample Size | n = 127 ; % = 16.24 | n = 1136 ; % = 16.17 |
| Depression/Anxiety | ||
| Sample Size | n = 73 ; % = 9.34 | n = 817 ; % = 11.63 |
| Schizophrenia | ||
| Sample Size | n = 15 ; % = 1.92 | n = 141 ; % = 2.01 |
| Dementia | ||
| Sample Size | n = 27 ; % = 3.45 | n = 427 ; % = 6.08 |
| Diabetes | ||
| Sample Size | n = 154 ; % = 19.69 | n = 1578 ; % = 22.46 |
| Hypertension | ||
| Sample Size | n = 174 ; % = 22.25 | n = 1962 ; % = 27.93 |
| Cardiovascular disease | ||
| Sample Size | n = 78 ; % = 9.97 | n = 1172 ; % = 16.68 |
| Chronic lung disorders | ||
| Sample Size | n = 66 ; % = 8.44 | n = 935 ; % = 13.31 |
| Obesity | ||
| Sample Size | n = 235 ; % = 30.05 | n = 1900 ; % = 27.05 |
| BMI | ||
| Mean/95% CI | 27.32 (26.88 to 27.77) | 27.36 (27.22 to 27.52) |
| Use of immune suppressing treatments | ||
| Custom value | NA | NA |
| Socioeconomic status | ||
| Low-medium | ||
| Sample Size | n = 601 ; % = 83.7 | n = 4418 ; % = 67.73 |
| COVID-19 positive (N = 782) | COVID-19 negative (N = 7025) | |
| High-medium | ||
| Sample Size | n = 117 ; % = 16.3 | n = 2105 ; % = 32.27 |
| Previous history of COVID-19 | ||
| Custom value | NA | NA |
| Other supplement use | ||
| Custom value | NA | NA |
| Timing of vitamin D measurements | ||
| Custom value | NA | NA |
| Shielding status | ||
| Custom value | NA | NA |
| Living in care homes | ||
| Custom value | NA | NA |
| Mean vitamin D (ng/mL) | ||
| Mean/95% CI | 19 (18.4 to 19.6) | 20.6 (20.3 to 20.8) |
| Plasma 25(OH)D level categories | ||
| Sufficiency >30ng/mL
Odds ratio is a crude comparison between the positive and negative cohorts |
||
| Sample Size Odds ratio | n = 79 ; % = 10.1 1 | n = 106 ; % = 15.1 |
| Insufficiency 29-20 ng/mL
Odds ratio is a crude comparison between the positive and negative cohorts |
||
| Sample Size Odds ratio/95% CI | n = 598 ; % = 76.5 1.59 (1.24 to 2.02) | n = 5050 ; % = 71.8 |
| Deficiency <20 ng/mL
Odds ratio is a crude comparison between the positive and negative cohorts |
||
| Sample Size Odds ratio/95% CI | n = 105 ; % = 13.4 1.58 (1.13 to 2.09) | n = 915 ; % = 13.1 |
Outcomes
Multivariable logistic regression analysis associating low vitamin D level with COVID-19 outcomes
Controlling for multiple conditions, OR with 95% confidence interval (CI). Results are presented unadjusted and adjusted.
|
COVID-19 positive vs COVID-19 negative |
|
| N1 = 782, N2 = 7025 | |
| Infection with COVID-19
Polarity: Lower values are better |
|
| Unadjusted | |
| Odds ratio/95% CI | 1.58 (1.24 to 2.01) |
| Adjusted | |
| Odds ratio/95% CI | 1.5 (1.13 to 1.98) |
| Hospitalisation
Polarity: Lower values are better |
|
| Unadjusted | |
| Odds ratio/95% CI | 2.09 (1.01 to 4.31) |
| Adjusted | |
| Odds ratio/95% CI | 1.95 (0.99 to 4.78) |
| Section | Question | Answer |
| Study participation | Summary Study participation | Moderate risk of bias
(Cohort of people with potentially biased vitamin D levels and not reporting ethnicity.) |
| Study Attrition | Study Attrition Summary | Low risk of bias |
| Prognostic factor measurement | Prognostic factor Measurement Summary | Low risk of bias |
| Outcome Measurement | Outcome Measurement Summary | Low risk of bias
(Diagnosis done by verified test.) |
| Study Confounding | Study Confounding Summary | Moderate risk of bias
(Ethnicity, immunosuppressants and vitamin D supplements not listed.) |
| Section | Question | Answer |
| Statistical Analysis and Reporting | Statistical Analysis and Presentation Summary | Low risk of bias
(Analyses presented adequately, no suspicion of bias) |
| Overall risk of bias and directness |
Risk of Bias |
Moderate
(Ethnicity, immunosuppressants and vitamin D supplements not listed. Cohort of people with potentially biased vitamin D levelsz.) |
| Directness | Partially applicable
(Historic vitamin D measurements used) |
| Radujkovic, 2020 |
| Bibliographic
Reference |
Radujkovic, Aleksandar; Hippchen, Theresa; Tiwari-Heckler, Shilpa; Dreher, Saida; Boxberger, Monica; Merle, Uta; Vitamin D Deficiency
and Outcome of COVID-19 Patients.; Nutrients; 2020; vol. 12 (no. 9) |
Study details
| Study design | Retrospective cohort study |
| Trial registration (if reported) | Not reported. |
| Study start date | 18-Mar-2020 |
| Study end date | 18-Jun-2020 |
| Aim of the study | To explore possible associations between vitamin D status and disease severity and survival in COVID-19 patients. |
| County/ Geographical location |
Heidelberg, Germany |
| Study setting | Hospital |
| Population description | Consecutive symptomatic SARS-CoV2-positive patients admitted to the Medical University Hospital Heidelberg enrolled onto a prospective non-interventional register. |
| Inclusion criteria | Participant consent and serum samples available for analysis. |
| Exclusion criteria | None reported. |
|
Vitamin D status measurements |
To assess whole-body vitamin D status of patients, levels of total 25(OH)D were measured retrospectively in cryopreserved (-80°C) serum samples collected in gel tubes at the time of admission and SARS-CoV-2 testing. Serum levels of total 25(OH)D were quantified using a commercially available immunoassay (ADVIA Centaur Vitamin D Total Assay®, Siemens Healthcare GmbH, Erlangen, Germany).
All measurements were carried out at the Department of Clinical Chemistry of the Heidelberg University Hospital using accredited laboratory methods (certified according to ISO 15189 by Germany’s national accreditation body). |
|
Methods used to confirm COVID-19 infection |
RNA from nasopharyngeal and oropharyngeal swabs was analysed using QIAGEN kits on QIASymphony or QIAcube devices. Varies RT-PCR reagent mixes were used: LightMix Modular SARS and Wuhan CoV E-gene, LightMix Modular SARS and Wuhan CoV N-gene, LightMix ModularWuhan CoV RdRP-gene, and LightMix Modular AV RNA Extraction Control (as internal control) from TIB MOLBIOL Syntheselabor GmbH (Berlin, Germany), and LightCycler Multiplex RNA Virus Master (Roche, Mannheim, Germany).
The decision for inpatient versus outpatient admission was based on the level of spontaneous oxygen saturation (SpO2 ≤ 93%), comorbidities, and the overall performance status. With regard to established COVID-19 severity classifications, all inpatients had severe disease (defined as tachypnoea [≥30 breaths per min], oxygen saturation ≤ 93% at rest, or PaO2/FiO2 ratio < 300 mm Hg) or critical disease (respiratory failure requiring mechanical ventilation, septic shock, or other organ dysfunction or failure that requires intensive care).
Outpatients included in the analysis had symptomatic disease presenting with fever, cough, sore throat, myalgia, and/or fatigue. Outpatients were visited in their home quarantine on a regular basis and their clinical conditions were regularly evaluated employing “Coronataxis” (i.e., home visits by medical students, nursing staff, and a supervising physician) which were implemented by the University Hospital Heidelberg and the regional health authorities. |
| Intervention | Not applicable |
| Comparator (where applicable) | Not applicable |
| Methods for population selection/allocation |
Described elsewhere |
| Methods for case- matching with control |
Not applicable |
|
Methods of data analysis |
Categorical data of patient characteristics were compared using Fisher’s exact test. Continuous Median follow-up time was calculated by the reverse Kaplan-Meier method.
Two endpoints were reported: severe course of disease (need for invasive mechanical ventilation and/or death, IMV/D, as a composite endpoint) and death of any cause.
Survival was calculated from the date of first presentation/admission and SARS-CoV-2 testing to last follow-up or death of any cause. Patients alive were censored at the date of last contact. Severe course of the disease was determined as time from the date of first presentation/admission and SARS-CoV-2 testing to IMV/D. Patients who were alive without necessary IMV were censored at the time of the last contact. Vitamin D deficiency was defined as serum total 25(OH)D level < 12 ng/mL (equivalent to <30 nM). In addition, the cut-point of 25(OH)D < 20 ng/mL (<50 nM) reflecting “Vitamin D insufficiency” was analysed. For uni- and multivariable analysis of the associations between Vitamin D status and severe course of the disease and survival, Cox regression models were applied. For the multivariable analyses, additional prognostic factors including age, gender, and presence of comorbidity were chosen to reflect confounders demonstrated to be associated with risk of death. All statistical tests were two-sided at a significance level of 5%. Hazard ratios (HR) were estimated with 95% confidence interval (95% CI). Calculations were done using IBM® SPSS® Statistics, Version 24.0.0. |
| Source of funding | No external funding declared. |
|
Other details |
Oxygen therapy included oxygen delivery via nasal cannula, high-low nasal oxygen therapy (HFNO), and invasive mechanical ventilation (IMV). Criteria for initiation of IMV were failure to maintain adequate ventilation or oxygenation in spite of high FiO2 delivery.
Hospitalized patients were treated with standard supportive care including antibiotic and antifungal therapy, whereas additional immunomodulatory therapy was inconsistently applied (azithromycin, hydroxychloroquine, tocilizumab, anakinra, prednisolone, maraviroc, Cytosorb, and plasmapheresis). Routine CT scans were performed at hospital admission for most patients. |
|
Study limitations (authors) |
A single-centre, retrospective, and observational study. In particular, since the number of events is rather low, the results require confirmation in larger patient cohorts analysing a higher number of events and considering additional potential confounders like obesity (as reflected by the body mass index) or other specific comorbidities.
It should be noted that without randomized controlled trial evidence, no causal association between Vitamin D deficiency and severity/outcome of COVID-19 can be inferred. However, since no causal treatment for COVID-19 is available, identification of modifiable prognostic factors may help to improve outcomes. |
|
Study limitations (reviewer) |
Such a large number of adjustments for low event rates lends itself to overfitting, limiting generalisability outside of this cohort. It also can cause imprecise effect estimates and large effects can arise from small changes in the raw data. Therefore the uncertainty in these effect estimates is very high and likely to change with further publication of evidence.
There could be differences in the clinical decisions made before ventilation due to this study not being in the UK and changes over the course of the pandemic. |
Study arms
| Vitamin D <12 ng/mL (N = 41)
Participants with serum vitamin D less than 12 ng/mL |
| Vitamin D ≥12 ng/mL (N = 144)
Participants with serum vitamin D greater than or equal to 12 ng/mL |
Characteristics
Arm-level characteristics
| Vitamin D <12 ng/mL (N = 41) | Vitamin D ≥12 ng/mL (N = 144) | |
| Age | ||
| MedianIQR | 66 (53 to 78) | 58 (47 to 67) |
| Inpatients
n=93 |
||
| MedianIQR | 71 (54 to 79) | 62 (50 to 70) |
| Outpatients
n=92 |
||
| MedianIQR | 60 (48 to 77) | 55 (42 to 63) |
| Gender
Male |
||
| Sample Size | n = 23 ; % = 56 | n = 72 ; % = 50 |
| Inpatients |
| Vitamin D <12 ng/mL (N = 41) | Vitamin D ≥12 ng/mL (N = 144) | |
| n=93 | ||
| Sample Size | n = 19 ; % = 66 | n = 40 ; % = 62 |
| Outpatients
n=92 |
||
| Sample Size | n = 4 ; % = 33 | n = 32 ; % = 40 |
| Ethnicity | ||
| Custom value | NA | NA |
| Comorbidities
Any |
||
| Sample Size | n = 22 ; % = 54 | n = 55 ; % = 38 |
| Inpatients
n=93 |
||
| Sample Size | n = 19 ; % = 66 | n = 33 ; % = 52 |
| Outpatients
n=92 |
||
| Sample Size | n = 3 ; % = 25 | n = 22 ; % = 27 |
| BMI | ||
| Custom value | NA | NA |
| Use of immune suppressing treatments | ||
| Custom value | NA | NA |
| Socioeconomic status | ||
| Custom value | NA | NA |
| Previous history of COVID-19 | ||
| Custom value | NA | NA |
| Other supplement use
Vitamin D |
||
| Sample Size | n = 0 ; % = 0 | n = 6 ; % = 7 |
| Inpatients
n=93 |
||
| Sample Size | n = 0 ; % = 0 | n = 6 ; % = 9 |
| Vitamin D <12 ng/mL (N = 41) | Vitamin D ≥12 ng/mL (N = 144) | |
| Outpatients
n=92 |
||
| Sample Size | n = 0 ; % = 0 | n = 0 ; % = 0 |
| Timing of vitamin D measurements | ||
| Custom value | NA | NA |
| Shielding status | ||
| Custom value | NA | NA |
| Living in care homes | ||
| Custom value | NA | NA |
| Cardiovascular disease | ||
| Sample Size | n = 18 ; % = 44 | n = 40 ; % = 28 |
| Inpatients
n=93 |
||
| Sample Size | n = 17 ; % = 59 | n = 28 ; % = 44 |
| Outpatients
n=92 |
||
| Sample Size | n = 1 ; % = 8 | n = 12 ; % = 15 |
| Diabetes | ||
| Sample Size | n = 8 ; % = 20 | n = 11 ; % = 8 |
| Inpatients
n=93 |
||
| Sample Size | n = 6 ; % = 21 | n = 7 ; % = 11 |
| Outpatients
n=92 |
||
| Sample Size | n = 2 ; % = 17 | n = 4 ; % = 5 |
| Chronic kidney disease | ||
| Sample Size | n = 2 ; % = 5 | n = 6 ; % = 4 |
| Inpatients
n=93 |
||
| Sample Size | n = 2 ; % = 7 | n = 6 ; % = 9 |
| Vitamin D <12 ng/mL (N = 41) | Vitamin D ≥12 ng/mL (N = 144) | |
| Outpatients
n=92 |
||
| Sample Size | n = 0 ; % = 0 | n = 0 ; % = 0 |
| Chronic lung disease | ||
| Sample Size | n = 6 ; % = 15 | n = 9 ; % = 6 |
| Inpatients
n=93 |
||
| Sample Size | n = 6 ; % = 21 | n = 4 ; % = 6 |
| Outpatients
n=92 |
||
| Sample Size | n = 0 ; % = 0 | n = 5 ; % = 6 |
| History of malignancy | ||
| Sample Size | n = 5 ; % = 12 | n = 12 ; % = 8 |
| Inpatients
n=93 |
||
| Sample Size | n = 4 ; % = 14 | n = 5 ; % = 8 |
| Outpatients
n =92 |
||
| Sample Size | n = 1 ; % = 8 | n = 7 ; % = 9 |
| Maximum oxygen therapy | ||
| Sample Size | n = 26 ; % = 54 | n = 54 ; % = 38 |
| Inpatients
n=93 |
||
| Sample Size | n = 26 ; % = 90 | n = 54 ; % = 86 |
| Outpatients
n=92 |
||
| Sample Size | n = 0 ; % = 0 | n = 0 ; % = 0 |
Outcomes
Associations of Vitamin D Status with the Endpoints Invasive Mechanical Ventilation and/or Death and Death
Results from Cox regression analysis and reverse Kaplan Meier survival analysis. Composite endpoint includes mechanical ventilation and death to produce hazard ratios of cumulative incidence. Survival analysis assesses death only. Adjusted values take into account age, gender, and comorbidities.
|
Vitamin D <12 ng/mL vs Vitamin D ≥12 ng/mL |
|
| N1 = 41, N2 = 144 | |
| Cumulative incidence, whole cohort
Polarity: Lower values are better |
|
| Unadjusted | |
| Hazard ratio/95% CI | 7.66 (3.53 to 16.63) |
| Adjusted | |
| Hazard ratio/95% CI | 6.12 (2.79 to 13.42) |
| Cumulative incidence, inpatients only
Polarity: Lower values are better |
|
| Unadjusted | |
| Hazard ratio/95% CI | 5.24 (2.41 to 11.42) |
| Adjusted | |
| Hazard ratio/95% CI | 4.65 (2.11 to 10.25) |
| Death, whole cohort
Polarity: Lower values are better |
|
| Unadjusted | |
| Hazard ratio/95% CI | 18.05 (5.14 to 63.43) |
| Adjusted | |
| Hazard ratio/95% CI | 14.73 (4.16 to 52.19) |
| Death, inpatients only
Polarity: Lower values are better |
|
| Unadjusted | |
| Hazard ratio/95% CI | 12.31 (3.5 to 43.34) |
| Adjusted | |
| Hazard ratio/95% CI | 11.51 (3.24 to 40.92) |
| Section | Question | Answer |
| Study participation | Summary Study participation | Moderate risk of bias
(Ethnicity not included.) |
| Study Attrition | Study Attrition Summary | Low risk of bias
(No attrition reported.) |
| Prognostic factor measurement | Prognostic factor Measurement Summary | Low risk of bias
(Immunoassay conducted for all participants.) |
| Outcome Measurement | Outcome Measurement Summary | Low risk of bias
(Standardised outcome measurements, events hard to misclassify.) |
| Study Confounding | Study Confounding Summary | Moderate risk of bias
(Ethnicity and immunosuppressant use missing.) |
| Statistical Analysis and Reporting | Statistical Analysis and Presentation Summary | Low risk of bias |
| Overall risk of bias and directness | Risk of Bias | Moderate
(Possible confounding caused by not controlling for ethnicity and immunosuppressants.) |
| Directness | Directly applicable (There could be differences in the clinical decisions made before ventilation due to this study not being in the UK and changes over the course of the pandemic) |
| Raisi-Estabragh, 2020 |
| Bibliographic Reference | Raisi-Estabragh, Zahra; McCracken, Celeste; Bethell, Mae S; Cooper, Jackie; Cooper, Cyrus; Caulfield, Mark J; Munroe, Patricia B; Harvey, Nicholas C; Petersen, Steffen E; Greater risk of severe COVID-19 in Black, Asian and Minority Ethnic populations is not explained by
cardiometabolic, socioeconomic or behavioural factors, or by 25(OH)-vitamin D status: study of 1326 cases from the UK Biobank.; Journal of public health (Oxford, England); 2020; vol. 42 (no. 3); 451-460 |
Study details
| Study design | Case-control study |
| Trial registration (if reported) | Not reported. |
| Study start date | 16-Mar-2020 |
| Study end date | 18-May-2020 |
| COVID-19 prevalence at the time of the study |
Higher prevalence (e.g. during peak of first wave) |
| Aim of the study | By using the UK Biobank cohort, the study aimed to test if the different rates of COVID-19 across sex and ethnicities could be explained by cardiometabolic, socioeconomic, lifestyle and behavioural factors. Vitamin D was also tested as part of these factors. |
| County/ Geographical location |
UK |
| Study setting | Community |
| Population description | People were recruited into the UK Biobank study between 2006-2010. It aims to capture the health of a broad range of the population to track outcomes of people and assess if there are common factors causing disease of middle/old age. People were recruited by post, everyone living within 10 miles of the 22 UK Biobank assessment centres were invited to participate. |
|
Inclusion criteria |
Aged 40-69 years old, as the UKBiobank protocol specifies.
Taken a COVID-19 test. |
| Exclusion criteria | Unable to provide consent. |
|
Vitamin D status measurements |
Measurements were taken during the 2006-2010 recruitment period. Vitamin D was measured at a central laboratory with a biochemical test, [Clinical Laboratory Improvement Amendments (CLIA) analysis on a DiaSorin Ltd. LIASON XL]. It limited results to between 10 nmol/L and 375 nmol/L. Any results above or below those thresholds were undetectable and were labelled accordingly as either 10 or 375 nmol/L. |
| Methods used to confirm COVID-19 infection |
Data matched with Public Health England COVID-19 test results released to UK Biobank researchers. |
| Intervention | Not applicable. |
| Comparator (where applicable) | Not applicable. |
|
Methods for population selection/allocation |
People were included in this current study if they had taken a COVID-19 test. Both people with positive and negative test results were included to allow associations to be drawn.
The study states that “As UK testing during this period was almost entirely restricted to hospitalized patients, researchers have been advised that COVID-19 positive status can be taken as surrogate for severe disease.” |
| Methods for case- matching with control |
Not applicable. |
|
Methods of data analysis |
Participants were grouped into COVID-19 positive and negative cohorts.
2 models relevant to the protocol were conducted in the analyses: 1) individual correlations between each of the variables and COVID infection; 2) correlations between multiple variables and COVID infection.
1) Univariate logistic regression was performed for every variable individually to assess the association between them and SARS-CoV-2 infection. Models were run for the whole cohort and then separately for men and women, and for white and non-white participants.
2) Multivariable logistic regression models were run to associated groups of variables with COVID-19 infection, one of which included vitamin D levels. The variables included in this model were: sex, age, ethnicity and vitamin D. Variables were added to the model simultaneously.
Adjustments were made based on the season the vitamin D measurement was made and ethnicity. Therefore, seasonal adjustment was conducted separately for white and BAME populations and an intercept added to the adjusted variables to maintain the difference between the two groups.
All analyses were conducted on R v3.6.2 and R Studio v1.2.5019 |
| Attrition/loss to follow-up | Not applicable |
|
Source of funding |
Z.R.E. is supported by a British Heart Foundation Clinical Research Training Fellowship (FS/17/81/33318). S.E.P., P.B.M. and M.J.C. acknowledge support from the Barts Biomedical Research Centre funded by the National Institute for Health Research (NIHR). N.C.H. and C.C. acknowledge support from the UK Medical Research Council, NIHR Southampton Biomedical Research Centre, University of Southampton and University Hospital Southampton and NIHR Oxford Biomedical Research Centre, University of Oxford. |
|
Study limitations (Author) |
Not possible to evaluate causal relationships.
Possible changes to vitamin D concentrations between baseline measurement taken when participants were recruited over a decade before this current analysis was performed.
Limited age range.
Wider social, economic and behavioural factors that likely to impact on the infection rate of COVID-19 than could be studied in UK Biobank.
People’s occupations could be a factor in transmission.
Aggregating all BAME populations overlooks differences between ethnicities. |
|
Study limitations (Reviewer) |
Other variables, apart from vitamin D status, could have changed since the participants were first recruited. Particularly if they have not updated their information, or not had to access health services, when it is most likely details are updated. This could bias results in unpredictable ways resulting in misleading conclusions.
For limitations concerning using UK Biobank data, see limitations in the evidence table for Hastie 2020. |
Study arms
| COVID positive (N = 1326)
People who tested positive with COVID-19 |
| COVID negative (N = 3184)
People who tested negative with COVID-19 |
Characteristics
Arm-level characteristics
| COVID positive (N = 1326) | COVID negative (N = 3184) | |
| Age | ||
| Standardised Mean/SD | 68.11 (9.23) | 68.91 (8.72) |
| Gender
Male |
||
| Sample Size | n = 696 ; % = 52.5 | n = 1505 ; % = 47.3 |
| Ethnicity | ||
| White | ||
| Sample Size | n = 1141 ; % = 86 | n = 2927 ; % = 91.9 |
| Black | ||
| Sample Size | n = 76 ; % = 5.7 | n = 91 ; % = 2.9 |
| Chinese | ||
| Sample Size | n = 6 ; % = 0.5 | n = 3 ; % = 0.1 |
| Mixed | ||
| Sample Size | n = 9 ; % = 0.7 | n = 24 ; % = 0.8 |
| Other
Ethnicity was missing for n = 11 test positive and n = 16 test negative participants, so were included in ‘other’ |
||
| Sample Size | n = 34 ; % = 2.6 | n = 61 ; % = 1.9 |
| Comorbidities | ||
| Diabetes | ||
| Sample Size | n = 217 ; % = 16.4 | n = 449 ; % = 14.1 |
| Hypertension | ||
| Sample Size | n = 624 ; % = 47.1 | n = 1457 ; % = 45.8 |
| High cholesterol | ||
| Sample Size | n = 437 ; % = 33 | n = 1034 ; % = 32.5 |
| Prior MI | ||
| Sample Size | n = 96 ; % = 7.2 | n = 242 ; % = 7.6 |
| BMI (kg/m²)
Please note IQR is reported as +/- and not as a range. |
||
| COVID positive (N = 1326) | COVID negative (N = 3184) | |
| MedianIQR | 28.04 (21.57 to 34.51) | 27.41 (21.04 to 33.78) |
| Use of immune suppressing treatments | ||
| Custom value | NA | NA |
| Socioeconomic status | ||
| MedianIQR | -0.91 (4.43 to -6.25) | -1.55 (-6.55 to 3.45) |
| Previous history of COVID-19 | ||
| Custom value | NA | NA |
| Other supplement use | ||
| Custom value | NA | NA |
| Timing of vitamin D measurements
No individual data available |
||
| Custom value | NA | NA |
| Shielding status | ||
| Custom value | NA | NA |
| Living in care homes | ||
| Custom value | NA | NA |
| Smoking
Smokers |
||
| Sample Size | n = 683 ; % = 51.1 | n = 1653 ; % = 51.9 |
Outcomes
Univariate logistic regression models exposures associations with COVID-19 status in whole cohort, men, and women within the tested sample
Results presented below show how likely people in the cohort were of testing positive for COVID-19 based on one characteristic at a time. An odds ratio and 95% confidence interval higher than 1 means the variable is associated with higher COVID-19 infection and vice versa.
|
COVID positive vs COVID negative |
|
| N1 = 1326, N2 = 3184 | |
| Sex (male) |
|
COVID positive vs COVID negative |
|
| N1 = 1326, N2 = 3184 | |
| Polarity: Lower values are better | |
| Odds ratio/95% CI | 1.23 (1.08 to 1.4) |
| Age
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 0.99 (0.98 to 1) |
| Men | |
| Odds ratio/95% CI | 0.99 (0.98 to 1) |
| Women | |
| Odds ratio/95% CI | 0.98 (0.97 to 0.99) |
| Non-white ethnicity
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 1.85 (1.51 to 2.28) |
| Men | |
| Odds ratio/95% CI | 2.09 (1.55 to 2.83) |
| Women | |
| Odds ratio/95% CI | 1.69 (1.27 to 2.25) |
| Townsend deprivation score
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 1.04 (1.02 to 1.06) |
| Men | |
| Odds ratio/95% CI | 1.04 (1.02 to 1.07) |
| Women | |
| Odds ratio/95% CI | 1.05 (1.02 to 1.07) |
| Household size
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 1.12 (1.06 to 1.17) |
| Men | |
| Odds ratio/95% CI | 1.11 (1.03 to 1.2) |
|
COVID positive vs COVID negative |
|
| N1 = 1326, N2 = 3184 | |
| Women | |
| Odds ratio/95% CI | 1.12 (1.05 to 1.21) |
| Generations in household
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 1.26 (1.11 to 1.43) |
| Men | |
| Odds ratio/95% CI | 1.21 (1.01 to 1.45) |
| Women | |
| Odds ratio/95% CI | 1.35 (1.14 to 1.61) |
| Family/friends visits
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 0.84 (0.72 to 0.98) |
| Men | |
| Odds ratio/95% CI | 0.85 (0.7 to 1.04) |
| Women | |
| Odds ratio/95% CI | 0.87 (0.69 to 1.11) |
| Socialisation habits
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 1.04 (0.91 to 1.19) |
| Men | |
| Odds ratio/95% CI | 1.14 (0.94 to 1.39) |
| Women | |
| Odds ratio/95% CI | 0.94 (0.77 to 1.14) |
| Diabetes
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 1.19 (1 to 1.42) |
| Men | |
| Odds ratio/95% CI | 1.18 (0.94 to 1.49) |
|
COVID positive vs COVID negative |
|
| N1 = 1326, N2 = 3184 | |
| Women | |
| Odds ratio/95% CI | 1.12 (0.84 to 1.49) |
| Hypertension
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 1.05 (0.93 to 1.2) |
| Men | |
| Odds ratio/95% CI | 0.99 (0.83 to 1.19) |
| Women | |
| Odds ratio/95% CI | 1.05 (0.87 to 1.26) |
| BMI (kg/m²)
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 1.02 (1.01 to 1.04) |
| Men | |
| Odds ratio/95% CI | 1.03 (1.01 to 1.05) |
| Women | |
| Odds ratio/95% CI | 1.02 (1 to 1.03) |
| Smoking
Smoker, current or previous Polarity: Lower values are better |
|
| Odds ratio/95% CI | 0.98 (0.87 to 1.12) |
| Men | |
| Odds ratio/95% CI | 1.07 (0.89 to 1.29) |
| Women | |
| Odds ratio/95% CI | 0.84 (0.7 to 1.01) |
| Vitamin D (nmol/L) Polarity: Lower values are better | |
| Odds ratio/95% CI | 1 (0.99 to 1) |
| Men | |
| Odds ratio/95% CI | 1 (0.99 to 1) |
|
COVID positive vs COVID negative |
|
| N1 = 1326, N2 = 3184 | |
| Women | |
| Odds ratio/95% CI | 1 (0.99 to 1) |
Multivariable logistic regression model testing the role of vitamin D in determining risk of COVID-19
Variables added to the model were sex, age, ethnicity and vitamin D. An odds ratio and 95% confidence interval higher than 1 indicates that the variable is associated with a positive COVID-19 test when the other variables are adjusted.
|
COVID positive vs COVID negative |
|
| N1 = 1326, N2 = 3184 | |
| Male sex
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 1.31 (1.14 to 1.5) |
| Age
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 0.99 (0.98 to 1) |
| Men | |
| Odds ratio/95% CI | 1 (0.99 to 1.01) |
| Women | |
| Odds ratio/95% CI | 0.99 (0.97 to 1) |
| BAME ethnicity
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 1.77 (1.41 to 2.22) |
| Men | |
| Odds ratio/95% CI | 2.02 (1.45 to 2.82) |
| Women | |
| Odds ratio/95% CI | 1.6 (1.16 to 2.18) |
| Vitamin D
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 1 (1 to 1) |
|
COVID positive vs COVID negative |
|
| N1 = 1326, N2 = 3184 | |
| Men | |
| Odds ratio/95% CI | 1 (1 to 1.01) |
| Women | |
| Odds ratio/95% CI | 1 (1 to 1.01) |
| Section | Question | Answer |
| Study participation | Summary Study participation | Moderate risk of bias
(From the initial UK Biobank sampling.) |
|
Study Attrition |
Study Attrition Summary |
Moderate risk of bias
(Key variable, ethnicity was missing for n = 11 test positive and n = 16 test negative participants, these participants are included as part of ‘other ethnicity’ in the baseline demographics table but were excluded from subsequent modelling)) |
| Prognostic factor measurement | Prognostic factor Measurement Summary | High risk of bias
(Length of time between measuring vitamin D and when it was used to associate with COVID-19 infection.) |
| Outcome Measurement | Outcome Measurement Summary | Low risk of bias |
| Study Confounding | Study Confounding Summary | Low risk of bias
(Low risk of bias due to small numbers affected.) |
| Statistical Analysis and Reporting | Statistical Analysis and Presentation Summary | Low risk of bias |
| Overall risk of bias and directness | Risk of Bias | High
(Bias in measurement of prognostic factor, vitamin D, could significantly bias results.) |
| Directness | Partially applicable
(vitamin D status and demographic data are historical) |
| Ye, 2020 |
| Bibliographic Reference | Ye, Kun; Tang, Fen; Liao, Xin; Shaw, Benjamin A; Deng, Meiqiu; Huang, Guangyi; Qin, Zhiqiang; Peng, Xiaomei; Xiao, Hewei; Chen, Chunxia; Liu, Xiaochun; Ning, Leping; Wang, Bangqin; Tang, Ningning; Li, Min; Xu, Fan; Lin, Shao; Yang, Jianrong; Does Serum Vitamin D
Level Affect COVID-19 Infection and Its Severity?-A Case-Control Study.; Journal of the American College of Nutrition; 2020; 1-8 |
Study details
| Study design | Case-control study |
| Trial registration (if reported) | Not reported. |
| Aim of the study | To assess the association between 25(OH)D levels and COVID-19 disease, its severity, and its clinical characteristics in a Chinese population. |
| County/ Geographical location |
Nanning, China. |
| Study setting | Hospital |
|
Population description |
The cases included all patients with COVID-19 disease, from February 16th, 2020 to March 16th, 2020, treated at the Yongwu Hospital of The People’s Hospital of Guangxi Zhuang Autonomous Region (n=62).
Controls were recruited from The Physical Examination Center at the Institution with no medical disorder (hypertension, diabetes, renal disease, and pneumonia), and were frequency matched by sex and age as the cases (n=80).
The study also compared participants with different severity of COVID-19. The participants in the mild/moderate (n=50) and severe/critical (n=10) arms are from the COVID-19 positive cohort. 2 participants from the positive cohort were asymptomatic and do not appear in either mild/moderate or severe/critical arms. |
| Inclusion criteria | Cases had to be positive according to the COVID-19 test conducted. |
| Exclusion criteria | None reported. |
|
Vitamin D status measurements |
Serum samples were collected for each patient at admission and stored at -80 degrees C before measuring the concentration of 25- hydroxyvitamin D (25(OH)D) using an electrochemiluminescent immunoassay (ECLIA) with a Roche Elecsys 10100/201 system.
The study followed the Endocrine Society clinical practice guideline, vitamin D deficiency (VDD) was defined as a 25(OH)D<50 nmol/L, vitamin D insufficiency as 50 nmol/L≤25(OH)D<75 nmol/L and vitamin D sufficiency as 25(OH)D≥75 nmol/L. |
|
Methods used to confirm COVID-19 infection |
The cases of COVID-19 were diagnosed according to the guidelines of the National Health Commission of China and were confirmed by positive SARS-CoV-2 RNA with throat swab samples (Sansure Biotechnology, Changsha, Hunan, China).
Severe COVID-19 case was defined according to the guidelines of the National Health Commission of China. Severe cases met at least one of the following criteria: 1) breathing rate >30/min
2) pulse oximeter oxygen saturation (SpO2) ≤93% when breathing ambient air
3) ratio of partial oxygen pressure (PaO2) to the fraction of inspired oxygen (FiO2) ≤300mmHg (1mmHg = 0.133 kpa)
4) lung imaging showing significant progression of >50% with 24 to 48 hours Critical cases met at least 1 of the following criteria: 1) respiratory failure (PaO2 <60mmHg when breathing ambient air)
2) hemodynamic shock (persisting hypotension requiring vasopressors to maintain MAP ≥65mmHg and serum lactate level >2 mmol/L despite volume resuscitation
3) organ failure or admittance to intensive care unit (ICU).
All other cases were classed as mild/moderate. |
| Intervention | Not applicable |
| Comparator (where applicable) | Not applicable |
| Methods for population selection/allocation |
Not applicable |
| Methods for case- matching with control |
Frequency matched controls to cases by sex and age. |
|
Methods of data analysis |
Potential confounding variables include age, sex, and comorbidities (those listed in the characteristics table) which are suspected risk factors of COVID-19 or thought to be associated with both COVID-19 infection and vitamin D deficiency.
Continuous variables were presented as mean (SD) and compared with students t test, if normally distributed, or median [IQR] and Mann-Whitney U test if non-normally distributed. Categorical variables were presented as number and percentage and analysed by chi- square test or Fisher’s exact test when counts were expected to be <1.
2 model were reported: an association between all measured risk factors and severe/critical disease; 2) an association between all measured risk factors and cases vs controls. These were conducted by unconditional logistic regression.
SPSS 22.0 was used to conduct the analyses. |
| Attrition/loss to follow-up | Not loss to follow-up reported. |
| Source of funding | The study was supported by the Guangxi Critical Infectious Disease Center (2020281) and Nanning Science and Technology Foundation (2018030). |
|
Study limitations (Author) |
Controls were not tested for COVID-19, creating bias towards the null as infected people can be asymptomatic.
Small sample size means study power was calculated to be 76.4%, lower than the standard 80% considered acceptable in epidemiological studies. This means the study may not have adequate power to accurately detect differences in clinical indicators between cases.
The protective effect of serum 25(OH)D>75nmol/l is quite possibly greater than observed here which could have been better demonstrated if a wider range of serum 25(OH)D was present in the controls.
The hospital is situated in a city, which takes local cases. The majority of participants were from urban areas and therefore may not be generalisable to other populations in China. |
| Cases’ vitamin D levels were taken during hospital admission after the onset of COVID-19. They defended this by saying that prior research has shown vitamin D level is stable and not affected by acute respiratory infection making it a good representation of vitamin D level prior to infection and admission. | |
| Study limitations (reviewer) | There could be differences in the clinical decisions made before ICU admission due to this study not being in the UK and changes over the course of the pandemic. |
Study arms
| Controls (N = 80) |
| SARS-CoV2 positive (N = 62) |
| Mild/moderate COVID-19 (N = 50)
Participants from the COVID-19 positive arm who had either mild or moderate symptoms |
| Severe/critical (N = 10)
Participants from the COVID-19 positive arm who had either severe or critical symptoms |
Characteristics
Arm-level characteristics
| Controls (N = 80) | COVID-19 positive (N = 62) | Mild/moderate COVID-19 (N = 50) | Severe/critical (N = 10) | |
| Age | ||||
| MedianIQR | 42 (31 to 52) | 43 (32 to 59) | 39 (30 to 49) | 65 (54 to 69) |
| Gender
Female |
||||
| Sample Size | n = 48 ; % = 60 | n = 39 ; % = 63 | n = 31 ; % = 62 | n = 6 ; % = 60 |
| Ethnicity | ||||
| Custom value | NA | NA | NA | NA |
| BMI | ||||
| Custom value | NA | NA | NA | NA |
| Controls (N = 80) | COVID-19 positive (N = 62) | Mild/moderate COVID-19 (N = 50) | Severe/critical (N = 10) | |
| Use of immune suppressing treatments | ||||
| Custom value | NA | NA | NA | NA |
| Socioeconomic status | ||||
| Custom value | NA | NA | NA | NA |
| Previous history of COVID-19 | ||||
| Custom value | NA | NA | NA | NA |
| Other supplement use | ||||
| Custom value | NA | NA | NA | NA |
| Timing of vitamin D measurements | ||||
| Custom value | NA | NA | NA | NA |
| Shielding status | ||||
| Custom value | NA | NA | NA | NA |
| Living in care homes | ||||
| Custom value | NA | NA | NA | NA |
| Vitamin D status | ||||
| Deficiency | ||||
| Sample Size | n = 15 ; % = 19 | n = 26 ; % = 42 | n = 18 ; % = 36 | n = 8 ; % = 80 |
| Non-deficiency | ||||
| Sample Size | n = 65 ; % = 81 | n = 36 ; % = 58 | n = 32 ; % = 64 | n = 2 ; % = 20 |
| Comorbidities | ||||
| Diabetes | ||||
| Sample Size | n = NA | n = NA | n = 3 ; % = 6 | n = 2 ; % = 20 |
| Hypertension | ||||
| Sample Size | n = NA | n = NA | n = 4 ; % = 8 | n = 2 ; % = 20 |
| Liver injury | ||||
| Sample Size | n = NA | n = NA | n = 0 ; % = 0 | n = 1 ; % = 10 |
| COPD | ||||
| Controls (N = 80) | COVID-19 positive (N = 62) | Mild/moderate COVID-19 (N = 50) | Severe/critical (N = 10) | |
| Sample Size | n = NA | n = NA | n = 1 ; % = 2 | n = 0 ; % = 0 |
| Asthma | ||||
| Sample Size | n = NA | n = NA | n = 0 ; % = 0 | n = 0 ; % = 0 |
| Renal failure
Defined as GRF <90mL/min*1.73m2 |
||||
| Sample Size | n = NA | n = NA | n = 8 ; % = 16 | n = 8 ; % = 80 |
Outcomes
Distribution of vitamin D level among different severity status
Participants are distributed into vitamin D sufficient, insufficient and deficient. Data are presented as sample size and percentage. Percentages represent of total number of events across whole case cohort (n=62) and not within each vitamin D status group.
| Sufficient | Insufficient | Deficient | P value | |
| N = 10 | N = 26 | N = 26 | ||
| Clinical classification of severity
Polarity: Not set |
||||
| Asymptomatic | ||||
| Sample Size | n = 2 ; % = 100 | n = 0 | n = 0 ; % = 0 | 0.004 |
| Mild/moderate | ||||
| Sample Size | n = 7 ; % = 14 | n = 25 ; % = 50 | n = 18 ; % = 36 | |
| Severe/critical | ||||
| Sample Size | n = 1 ; % = 10 | n = 1 ; % = 10 | n = 8 ; % = 80 | |
| Adverse events
Polarity: Not set |
||||
| Shock | ||||
| Sample Size | n = 0 ; % = 0 | n = 0 ; % = 0 | n = 4 ; % = 100 | 0.025 |
| Mechanical ventilation | ||||
| Sample Size | n = 0 ; % = 0 | n = 0 ; % = 0 | n = 5 ; % = 100 | 0.012 |
| Sufficient | Insufficient | Deficient | P value | |
| N = 10 | N = 26 | N = 26 | ||
| FiO2<300mmHg | ||||
| Sample Size | n = 1 ; % = 12.5 | n = 1 ; % = 12.5 | n = 3 ; % = 75 | 0.58 |
| Lung infiltration | ||||
| Sample Size | n = 0 ; % = 0 | n = 1 ; % = 20 | n = 4 ; % = 80 | 0.099 |
Multivariable logistic regression analysis of all potential risk factors as predictors of severe/critical COVID-19.
|
Severe/critical vs Mild/moderate COVID-19 |
|
| N1 = 10, N2 = 50 | |
| Vitamin deficiency
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 15.18 (1.23 to 187.45) |
| Age
Every ten years per level (level 1-9) Polarity: Lower values are better |
|
| Odds ratio/95% CI | 2.45 (0.83 to 7.23) |
| Gender
Female Polarity: Lower values are better |
|
| Odds ratio/95% CI | 4.21 (0.28 to 64.35) |
| Renal failure
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 14.14 (0.79 to 253.9) |
| Hypertension
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 0.52 (0.02 to 12.29) |
| Diabetes
Polarity: Lower values are better |
|
| Odds ratio/95% CI | 0.58 (0.02 to 18.17) |
Odds of vitamin D deficiency
Likelihood of being vitamin D deficient according to disease state. Includes comparisons between all cases vs healthy controls and mild/moderate vs severe/critical cases. Values are unadjusted comparisons.
| COVID-19 positive vs Controls |
Severe/critical vs Mild/moderate COVID-19 |
|
| N1 = 62, N2 = 80 | N1 = 10, N2 = 50 | |
| Vitamin D deficiency
Polarity: Lower values are better |
||
| Odds ratio/95% CI | 3.13 (1.47 to 6.66) | 7.11 (1.36 to 37.16) |
| Section | Question | Answer |
| Study participation | Summary Study participation | Moderate risk of bias
(No description of BMI or ethnicity. Large number of urban participants.) |
| Study Attrition | Study Attrition Summary | Low risk of bias |
| Prognostic factor measurement | Prognostic factor Measurement Summary | Low risk of bias |
| Outcome Measurement | Outcome Measurement Summary | Low risk of bias |
| Study Confounding | Study Confounding Summary | High risk of bias
(Missing BMI and ethnicity.) |
| Statistical Analysis and Reporting | Statistical Analysis and Presentation Summary | Low risk of bias |
| Overall risk of bias and directness | Risk of Bias | High
(Missing BMI and ethnicity from adjusting and baseline characteristics.) |
| Directness | Directly applicable (There could be differences in the clinical decisions made before hospitalisation and ICU admission due to this study not being in the UK and changes over the course of the pandemic) |
Effectiveness of vitamin D as a COVID-19 treatment
Effectiveness of vitamin D supplement as treatment for COVID-19
|
Quality assessment |
Number of participants |
Effect |
Quality |
||||||||
| No of studies | Design | Risk of bias | Inconsistency | Indirectness | Imprecision | Other considerations | Interventi on |
Control |
Univariate 95%CI |
Multivariable 95%CI |
|
| Odds ratio: ICU admission (follow-up: 2 months or until death; Better indicated by lower values) | |||||||||||
| 11 | randomised trial | very serious2 | no serious inconsistency | no serious indirectness | no serious imprecision | low number of participants |
50 |
26 |
OR 0.02 (0.002 to 0.17) |
Adj OR 0.033
(0.003 to 0.25) |
VERY LOW |
| Odds ratio: mortality (follow-up: 2 months; Better indicated by lower values) | |||||||||||
| 11 | randomised trial | very serious2 | no serious inconsistency | no serious indirectness | no serious imprecision | low number of participants |
50 |
26 |
OR 0.0974 (0.004 to 2.99) |
- |
VERY LOW |
1 Entrenas Castillo 2020
2 Downgraded by 2 levels: Reported outcome, mortality, not analysed in multivariable analysis. Only ICU was reported in this way, even though they are both listed on the clinical trials register as outcomes. Adjustment for multivariable analysis not fully explored or reported, only hypertension and diabetes are reported as definitively included in the model but does include “others”.
3 Diabetes and hypertension were unbalanced after randomisation so were adjusted for in multivariable analysis.
4 No events in intervention arm. 2 events in control arm. NICE-calculated OR.
Effectiveness of vitamin as a COVID-19 preventative measure
Association between vitamin D status and COVID-19 outcomes
Association between vitamin D status and COVID-19 cases
|
Quality assessment |
Effect |
Quality |
|||||||
| No of studies |
Design N |
Risk of bias | Inconsistency | Indirectness | Imprecision | Other considerations |
Univariate 95%CI |
Multivariable 95%CI |
|
| Vitamin D status definition: Vitamin D level (nmol/L) | |||||||||
| 11 | Case control Cases n=449 Control n=348,598 | Very serious15 | no serious inconsistency | very serious indirectness13 | no serious imprecision | none |
OR 0.99 (0.99 to 0.999) |
Adj OR 1.002
(0.998 to 1.01) |
VERY LOW |
| 13 | Case control Cases n=1326
Control n=3184 |
Very serious15 | no serious inconsistency | serious indirectness13 | no serious imprecision | none |
OR 1 (0.99 to 1) |
Adj OR 14 (1 to 1) |
VERY LOW |
| Vitamin D status definition: Vitamin D level (ng/ml) | |||||||||
| 15 | Case control Cases n=197 Control n=197 | Serious16 | no serious inconsistency | No serious indirectness6 | Very serious imprecision7 | none |
_ |
Mean difference8 Case=11.9 ng/ml Control=21.2 ng/ml MD= -9.3 ng/ml (p<0.0001) |
VERY LOW |
| 19 | Cohort N=191,779
Unclear no. of positive cases |
Very serious15 | no serious inconsistency | Very serious indirectness10 | no serious imprecision | none |
_ |
Adj OR 0.98411
(0.983 to 0.986) |
VERY LOW |
| Vitamin D status definition: Vitamin D level (nmol/L)*Ethnicity interaction term19 | |||||||||
| 11 | Case control
Cases n=449 Control n=348,598 |
Very serious15 | no serious inconsistency | serious indirectness2 | no serious imprecision | none |
_ |
Adj OR 0.902
(0.66 to 1.23) |
VERY LOW |
| Vitamin D status definition: Vitamin D deficient (<25 nmol/L) | |||||||||
| 11 | Case control Cases n=449 Control n=348,598 | Very serious15 | no serious inconsistency | serious indirectness13 | no serious imprecision | none |
OR 1.37 (1.07 to 1.76) |
Adj OR 0.922
(0.71 to 1.21) |
VERY LOW |
| Vitamin D status definition: Vitamin D insufficient (<50 nmol/L) | |||||||||
| 11 | Case control Cases n=449
Control n=348,598 |
Very serious15 | no serious inconsistency | serious indirectness13 | no serious imprecision | none |
OR 1.19 (0.99 to 1.44) |
Adj OR 0.882
(0.72 to 1.08) |
VERY LOW |
| Vitamin D status definition: Vitamin D deficient (<20 ng/mL) | |||||||||
| 112 | Cohort Positive n=71
Negative n=418 |
Serious16 | no serious inconsistency | serious indirectness20 | no serious imprecision | none |
_ |
Adj OR 1.7714
(1.12 to 2.81) |
VERY LOW |
| Vitamin D status definition: Vitamin D suboptimal (<30 ng/mL) | |||||||||
| 117 | Case control Cases n=782 Control n=7025 | serious16 | no serious inconsistency | serious indirectness20 | no serious imprecision | none |
OR 1.58 (1.24 to 2.01) |
Adj OR 1.518
(1.13 to 1.98) |
VERY LOW |
1 Hastie 2020
2 Adjusted for ethnicity, sex, month of assessment, Townsend deprivation quintile, household income, self-reported health rating, smoking status, BMI category, age at assessment, diabetes, SBP, DBP, and long-standing illness, disability or infirmity.
3 Raisi-Esrabragh 2020
4 Adjusted for sex, age and ethnicity.
5 Hernandez 2020
6 No downgrade as the historical control is in 2019, assuming population profiles have not changed since.
7 Downgrade 2 levels: small sample size assessed against [n = 100+50i], where i = number of independent variables adjusted.
8 Adjusted for age, smoking, hypertension, diabetes mellitus, history of cardiovascular events, immunosuppression, body mass index, serum corrected calcium, glomerular filtration rate and the month of vitamin D determination.
9 Kaufman 2020
10 Downgrade 2 levels: Vitamin status data was historical (preceding 12 months) where vitamin level may have changed before SAR-CoV-2 testing. Also, the outcome is SAR- CoV-2 positive, not COVID-19 (unclear proportion of asymptomatic positive cases).
11 Adjusted for gender, age, latitudes, ethnicity.
12 Meltzer 2020
13 Downgrade 2 level: Vitamin D status and demographic data are over a decade old.
14 Adjusted for hypertension, diabetes, chronic pulmonary disease, pulmonary circulation disorders, depression, immunosuppression, liver disease, and chronic kidney disease.
15 Downgrade 2 levels: high risk of bias assessed by QUIPS checklist.
16 Downgrade 1 level: moderate risk of bias assessed by QUIPS checklist.
17 Merzon 2020
18 Adjusted for multiple conditions and demographic variables.
19 This means to explore whether Ethnicity is an effect moderator of vitamin D level which impacts on its association with COVID-19 cases. The non-significant result suggested ethnicity has no interaction with vitamin D level and its association with COVID-19 cases.
20 Downgrade 1 level: Vitamin D status are historical.
Association between vitamin D status and severity of COVID-19
|
Quality assessment |
Effect |
Quality |
|||||||
| No of studies |
Design N |
Risk of bias | Inconsistency | Indirectness | Imprecision | Other considerations |
Univariate 95%CI |
Multivariable 95%CI |
|
| Vitamin D status definition: Vitamin D level (nmol/L)
Severity definition: composite severity endpoint (admission to the intensive care unit (ICU), requirement for mechanical ventilation, or in-hospital mortality) |
|||||||||
| 11 | Case series2 n=197 | Very serious13 | no serious inconsistency | no serious indirectness | very serious imprecision3 | none |
OR 1.55 (0.66 to 3.65) |
Adj OR 1.134
(0.27 to 4.77) |
VERY LOW |
| Vitamin D status definition: Vitamin D suboptimal (<12 ng/ml)
Severity definition: composite endpoint including mechanical ventilation and death |
|||||||||
| 15 | Cohort n=185 | Serious20 | no serious inconsistency | no serious indirectness | very serious imprecision3 | none |
HR 7.66 (3.53 to 16.63) |
Adj HR 6.126
(2.79 to 13.42) |
VERY LOW |
| Vitamin D status definition: Vitamin D deficiency (<50 nmol/L)
Severity definition: severe/critical cases defined as having one of the following: breathing rate >30/min, O2 saturation ≤93% at rest, PaO2/FiO2 ≤mmHg or lung imaging shows significant progression, respiratory failure (PaO2 <60mmHg), shock, organ failures that requires ICU care |
|||||||||
| 17 | Case control COVID-19 positive n=60 | Very serious13 | no serious inconsistency | no serious indirectness | very serious imprecision3 | none |
OR 7.11 (1.36 to 37.16) |
Adj OR 15.188
(1.23 to 187.45) |
VERY LOW |
| Vitamin D status definition: Vitamin D level (ng/mL)
Severity definition (univariate): Moderate - fever and pulmonary symptoms with pneumonia. Severe-Critical - respiratory distress, O2 saturation ≤93% at rest, PaO2/FiO2 ≤mmHg or chest imaging shows lesion, respiratory failure (mechanical ventilation), shock, organ failures that requires ICU care Severity definition (multivariable): Mortality |
|||||||||
| 19 | Cohort n=149 | Serious20 | no serious inconsistency | serious indirectness24 | very serious imprecision7 | none |
Mean difference (SD) Moderate = 26.3 (8.4) Severe-Critical = 10.1 (6.2) MD = -16.2 (-18.6 to - 13.8) |
Adj OR 0.9210
(0.88 to 0.98) |
VERY LOW |
| Vitamin D status definition: Vitamin D deficiency as <20 ng/mL
Severity definition: composite outcome defining severity of COVID-19 included death, admission to ICU, and/or need for higher oxygen flow than that provided by a nasal cannula. |
|||||||||
| 111 | Case series n=80 | Very serious13 | no serious inconsistency | no serious indirectness | very serious imprecision3 | none | _ | Adj OR 3.212
(0.9 to 11.4) |
VERY LOW |
| Vitamin D status definition: vitamin D sufficient if received a vitamin D booster supplement within a month of COVID-19 diagnosis. Severity definition: mortality, follow-up up to 2 months. | |||||||||
| 114 | Cohort n=66 | Very serious13 | no serious inconsistency | serious indirectness25 | very serious imprecision3 | none |
HR 0.21 (0.07 to 0.63) |
Adj HR 0.1115
(0.03 to 0.48) |
VERY LOW |
| Vitamin D status definition: vitamin D sufficient if received vitamin D supplements for a year before COVID-19 diagnosis Severity definition: 14-day COVID-19 mortality | |||||||||
| 116 | Cohort n=77 | Very serious13 | no serious inconsistency | no serious indirectness | very serious imprecision3 | none | _ | Adj HR 0.0717
(0.01 to 0.61) |
VERY LOW |
| Vitamin D status definition: vitamin D sufficient if received vitamin D supplement when diagnosed with COVID-19 Severity definition: 14-day COVID-19 mortality | |||||||||
| 116 | Cohort n=77 | Very serious13 | no serious inconsistency | no serious indirectness | very serious imprecision3 | none | _ | Adj HR 0.3717
(0.06 to 2.21) |
VERY LOW |
| Vitamin D status definition: vitamin D sufficient if received vitamin D supplements for a year before COVID-19 diagnosis Severity definition: Severe COVID-19 - OSCI score ≥ 5 | |||||||||
| 116 | Cohort n=77 | Very serious13 | no serious inconsistency | no serious indirectness | very serious imprecision3 | none | _ | Adj OR 0.0817
(0.01 to 0.81) |
VERY LOW |
| Vitamin D status definition: vitamin D sufficient if received vitamin D supplement when diagnosed with COVID-19 Severity definition: Severe COVID-19 - OSCI score ≥ 5 | |||||||||
| 116 | Cohort n=77 | Very serious13 | no serious inconsistency | no serious indirectness | very serious imprecision3 | none | _ | Adj OR 0.4617
(0.07 to 2.85) |
VERY LOW |
| Vitamin D status definition: Vitamin D suboptimal (<30 ng/mL) Severity definition: hospitalisation for COVID-19 symptoms | |||||||||
| 118 | Case control n=7807 | Serious20 | no serious inconsistency | serious indirectness19 | no serious imprecision | none |
OR 2.09 (1.01 to 4.31) |
Adj OR 1.9521
(0.99 to 4.78) |
LOW |
| Vitamin D status definition: Vitamin D suboptimal (<12 nmol/L) Severity definition: mortality | |||||||||
| 122 | Cohort n=185 | Serious20 | no serious inconsistency | no serious indirectness | very serious imprecision3 | none |
HR 18.05 (5.14 to 63.43) |
Adj HR 14.7323
(4.16 to 52.19) |
VERY LOW |
1 Hernandez 2020
2 This is sub-analysis of cases only from Hernandez 2020.
3 Downgrade 2 levels: very small sample size assessed against [n = 100+50i], where i = number of independent variables adjusted.
4 Adjusted for age, smoking, hypertension, diabetes mellitus, history of cardiovascular events, immunosuppression, body mass index, serum corrected calcium, glomerular filtration rate and the month of vitamin D determination.
5 Radujkovic 2020
6 Adjusted for age, gender and comorbidities. Includes values for whole cohort not for inpatients only.
7 Ye 2020 - values presented in GRADE table are compared with mild/moderate cases.
8 Adjusted for age, sex and comorbidities.
9 Karahan 2020 - Cohort of COVID-19 positive patients with moderate or severe/critical condition.
10 Adjusted for age, smoking, hyperlipidaemia, diabetes mellitus, chronic kidney disease, chronic atrial fibrillation, congestive heart failure, acute kidney injury, eGFR, haemoglobin, neutrophil count.
11 Macaya 2020
12 Adjusted for age, gender, obesity, severe CKD, cardiac disease.
13 Downgrade 2 levels: high risk of bias assessed by QUIPS checklist.
14 Annweiler 2020
15 Adjusted for age, gender, number of drugs daily taken, functional abilities, nutritional status, COVID-19 treatment with corticosteroids and/or hydroxychloroquine and/or dedicated antibiotics, and hospitalization for COVID-19.
16 Annweiler 2020a
17 Compared with non-supplemented group diagnosed with COVID-19. Adjusted for age, gender, GIR score, severe undernutrition, history of cancer, history of hypertension, history of cardiomyopathy, glycated haemoglobin, number of acute health problems, use of antibiotics, use of systemic corticosteroids, use of treatments of respiratory disorders. 18 Merzon 2020
19 Historic vitamin D values used for association.
20 Downgrade 1 level: moderate risk of bias assessed by QUIPS checklist.
21 Adjusted for multiple conditions and demographic variables.
22 Radujkovic 2020
23 Adjusted for age, gender, comorbidity.
24 Downgrade 1 level: People with mild severity of COVID-19 were purposely excluded.
25 Downgraded 1 level: vitamin dosing occurs before and after infection diagnosis causing problems with temporal separation of events.
Appendix F - Ongoing studies (clinicaltrials.gov)
Search date: 29/10/2020
Double-blind randomized design comparing 1000 I.U. vitamin D versus matched placebo in healthy young adults. (Reducing Asymptomatic Infection With Vitamin D in Coronavirus Disease).
https://ClinicalTrials.gov/show/NCT04476680
Parallel-group, placebo-controlled trial of vitamin D3 supplementation for the Prevention of COVID-19 With Oral Vitamin D Supplemental Therapy in Essential healthcare Teams (PROTECT).
https://ClinicalTrials.gov/show/NCT04483635
A multicenter randomized double-blinded placebo-controlled clinical trial with parallel groups design to investigate the Preventive and Therapeutic Effects of Oral 25-hydroxyvitamin D3 on Coronavirus (COVID-19) in Adults.
https://ClinicalTrials.gov/show/NCT04386850
Blinded randomized clinical trial on the efficacy of Vitamin D Supplementation to Prevent the Risk of Acquiring or Evolving Into the Severe Form of COVID-19 in Healthcare Workers Caring for Patients With the Disease.
https://ClinicalTrials.gov/show/NCT04535791
Phase 3 Randomised Controlled Trial of Vitamin D Supplementation to Reduce Risk and Severity of COVID-19 and Other Acute Respiratory Infections in the UK Population
https://ClinicalTrials.gov/show/NCT04579640
Open-label RCT of vitamin D3 2000 IU (50 μg) plus 30 mg of zinc gluconate per day for 2 months versus usual care in adults >60 years who are ‘institutionalised’ but asymptomatic. Incidence of COVID-19 infection is a secondary outcome (Seguy D. Impact of Zinc and Vitamin D3 Supplementation on the Survival of Aged Patients Infected With COVID-19 (ZnD3-CoVici)
https://clinicaltrials.gov/ct2/show/NCT04351490
Single group open-label study of a combination of hydroxychloroquine, vitamins C and D (form not specified), and zinc as prophylaxis in healthy healthcare workers who are at risk of COVID-19 (Haza S. A Study of Hydroxychloroquine, Vitamin C, Vitamin D, and Zinc for the Prevention of COVID-19 Infection (HELPCOVID-19)
https://clinicaltrials.gov/ct2/show/record/NCT04335084
Open label RCT investigating interventions to prevent progression of COVID-19. Interventions include hydroxychloroquine, azithromycin, zinc, vitamin D, vitamin B12 with or without vitamin C. (International ALLIANCE Study of Therapies to Prevent Progression of COVID-19)
https://clinicaltrials.gov/ct2/show/NCT04395768
A Randomized, Double-Blind, Placebo-Controlled Phase IIa Study of Hydroxychloroquine, Vitamin C, Vitamin D, and Zinc for the Prevention of COVID-19 Infection on medical workers who at elevated risk of COVID-19 due to exposure to positive patients in the Emergency Department or Intensive Care Unit.
https://ClinicalTrials.gov/show/NCT04335084
A prospective, double-blind, randomized, placebo-controlled study in two distinct cohorts to evaluate the efficacy and safety of hydroxychloroquine (with vitamin D) in the prevention of COVID-19 infection.
https://ClinicalTrials.gov/show/NCT04372017
As monotherapy
Randomised controlled trial of a single oral dose of 25,000 IU (625 μg) vitamin D (form not specified) versus usual care in patients who are infected with SARS-CoV-2 but do not have severe symptoms (Castillo MJ. Vitamin D on Prevention and Treatment of COVID-19 (COVITD-19)
https://clinicaltrials.gov/ct2/show/NCT04334005
RCT comparing single doses of vitamin D3, 50,000 IU to 200,000 IU (1250 Vs 5000 μg) in people with COVID-19 pneumonia >75 years of age, or >70 with low oxygen saturations (Annweiler C. COvid-19 and Vitamin D Supplementation: a Multicenter Randomized Controlled Trial of High Dose Versus Standard Dose Vitamin D3 in High-risk COVID-19 Patients (CoVitTrial)
https://clinicaltrials.gov/ct2/show/record/NCT04344041
RCT comparing vitamin D3 (500,000 IU single dose) to placebo in adults admitted to hospital with COVID-19. Primary outcomes are respiratory SOFA at one week and need for high dose oxygen or mechanical ventilation at 30 days. (Cholecalciferol to Improve the Outcomes of COVID-19 Patients (CARED))
https://clinicaltrials.gov/ct2/show/NCT04411446
Single group open-label study of vitamin D supplementation (form not specified) in adults with COVID-19 and vitamin D deficiency (threshold for deficiency not defined). Participants to receive 2 weeks of vitamin D supplementation (10-15,000 IU based on age), with a further 3 weeks treatment if vitamin D levels remain low. Outcomes include 25(OH)D level and
COVID-19 symptoms. (Vitamin D Testing and Treatment for COVID 19) https://clinicaltrials.gov/ct2/show/NCT04407286
Randomised controlled trial comparing 2 doses of vitamin D3 (50,000 IU twice/once weekly and 1,000 IU daily) with low dose vitamin D3 in adults with COVID-19. Primary outcome: COVID-19 symptom recovery at 3 weeks. (Vitamin D and COVID-19 Management)
https://clinicaltrials.gov/ct2/show/NCT04385940
A Cluster-Randomized, Double-Blind, Placebo-Controlled Study to Evaluate the Efficacy of Vitamin D3 Supplementation to Reduce Disease Severity in Persons With Newly Diagnosed COVID-19 Infection and to Prevent Infection in Household Members.
https://ClinicalTrials.gov/show/NCT04536298
Randomized Controlled Trial of High Dose of Vitamin D as Compared With Placebo to Prevent Complications Among COVID-19 Patients.
https://ClinicalTrials.gov/show/NCT04411446
Open label randomised parallel study on the Prevention and Treatment With Calcifediol of COVID-19 Coronavirus-induced Acute Respiratory Syndrome (SARS).
https://ClinicalTrials.gov/show/NCT04366908
A Multicentre Randomized Controlled Trial of High Dose Versus Standard Dose Vitamin D3 in High-risk COVID-19 Patients (CoVitTrial).
https://ClinicalTrials.gov/show/NCT04344041
Vitamin D Supplementation in Patients With COVID-19: A Randomized, Double-blind, Placebo-controlled Trial.
https://ClinicalTrials.gov/show/NCT04449718
Randomised parallel trial on the Effect of Vitamin D Administration on Prevention and Treatment of Mild Forms of Suspected Covid-19.
https://ClinicalTrials.gov/show/NCT04334005
Randomised Parallel Trial on The Role of Vitamin D in Mitigating COVID-19 Infection Severity: Focusing on Reducing Health Disparities in South Carolina (VitD-COVID19).
https://ClinicalTrials.gov/show/NCT04482673
Randomised Parallel Trial To determine the efficacy of high dose Vitamin D (an over-the- counter nutritional supplement) in preventing immune-related complications in outpatients with confirmed SARS-CoV-2 infection.
https://ClinicalTrials.gov/show/NCT04489628
An open label treatment study for people with COVID 19 and low levels of vitamin D. https://ClinicalTrials.gov/show/NCT04407286
Randomised Parallel Trial on the Effect of Vitamin D on Morbidity and Mortality of the COVID-19 (COVID-VIT-D).
https://ClinicalTrials.gov/show/NCT04552951
High Dose Vitamin-D Substitution in Patients With COVID-19: a Randomized Controlled, Multi Centre Study (VitCov).
https://ClinicalTrials.gov/show/NCT04525820
A double blind, randomized, controlled three weeks clinical trial on the efficacy of vitamin D (daily low dose versus weekly high dose) in COVID-19 patients in order to determine the relationship between baseline vitamin D deficiency and clinical characteristics and to asses patients’ response to vitamin D supplementation in week three and determine its association with disease progression and recovery.
https://ClinicalTrials.gov/show/NCT04385940
Efficacy of Vitamin D Treatment in Paediatric Patients Hospitalized by COVID-19: Open Controlled Clinical Trial.
https://ClinicalTrials.gov/show/NCT04502667
Randomised Double Blind Controlled Study on Short Term, High Dose Vitamin D Supplementation for COVID-19 (SHADE).
https://ClinicalTrials.gov/show/NCT04459247
Single group open-label study of a combination of hydroxychloroquine, vitamins C and D (form not specified), and zinc plus azithromycin as treatment for COVID-19 (Haza S. A Study of Quintuple Therapy to Treat COVID-19 Infection (HAZDpaC)
https://clinicaltrials.gov/ct2/show/NCT04334512
Open-label RCT of vitamin D (form not specified; 50,000 IU once weekly for 2/52) added to aspirin 81 mg daily for 2/52. Investigating whether early treatment with aspirin and vitamin D in COVID-19 can mitigate COVID-19-associated coagulopathy and reduce hospitalization rates. (The LEAD COVID-19 Trial: Low-risk, Early Aspirin and Vitamin D to Reduce COVID- 19 Hospitalizations)
https://clinicaltrials.gov/ct2/show/NCT04363840
Open label randomised parallel study on the Impact of Zinc and Vitamin D3 Supplementation on the Survival of Institutionalized Aged Patients Infected With COVID-19.
https://ClinicalTrials.gov/show/NCT04351490
Multi-center, prospective, randomized controlled trial is to investigate low-risk, early treatment with aspirin and vitamin D in COVID-19 can mitigate the prothrombotic state and reduce hospitalization rates. (The LEAD COVID-19 Trial).
https://ClinicalTrials.gov/show/NCT04363840
A Randomized, Double-Blind, Placebo-Controlled Phase IIa Study of Quintuple Therapy (Hydroxychloroquine, Azithromycin, Vitamin C, Vitamin D, and Zinc) to Treat COVID-19 Infection.
https://ClinicalTrials.gov/show/NCT04334512
A Phase II Double-Blind Randomized Placebo-Controlled Trial of Combination Therapy (Ivermectin, Doxycycline Hcl, Zinc, Vitamin D3, Vitamin C) to Treat COVID-19 Infection.
https://ClinicalTrials.gov/show/NCT04482686
Proof-of-concept Open-label Randomized Dose-response Comparison Study of Famotidine Plus Vitamins C and D3 for Adults With Probable COVID-19.
https://ClinicalTrials.gov/show/NCT04565392
Randomized Double-Blind Placebo-Controlled Proof-of-Concept Trial of a Plant Polyphenol (with Vitamin D3) for the Outpatient Treatment of Mild Coronavirus Disease (COVID-19).
https://ClinicalTrials.gov/show/NCT04400890
Therapies to Prevent Progression of COVID-19, Including Hydroxychloroquine, Azithromycin, Zinc, Vitamin D, Vitamin B12 With or Without Vitamin C, a Multi-centre, International, Randomized Trial: The International ALLIANCE Study.
https://ClinicalTrials.gov/show/NCT04395768
A Non-Randomised Pilot Study for COVID-19 Outpatient Treatment With the Combination of Ivermectin-azithromycin-cholecalciferol.
https://ClinicalTrials.gov/show/NCT04399746
Association between vitamin D status and COVID-19
Case-control study investigating whether serum 25(OH)D level correlates to COVID-19 disease severity in people not treated in critical care. (Do Vitamin D Levels Really Correlated With Disease Severity in COVID-19 Patients? (COVIDVIT))
https://clinicaltrials.gov/ct2/show/NCT04394390
Case-series investigating differences in vitamin D blood levels between COVID-19 patients with different degrees of disease severity (mild-severe disease compared with patients requiring critical care). (VITACOV: Vitamin D Polymorphisms and Severity of COVID-19 Infection)
https://clinicaltrials.gov/ct2/show/NCT04370808
Case-control study investigating serum zinc, vitamin D and vitamin B12 levels in pregnant women with COVID-19. (Evaluation of the Relationship Between Zinc Vitamin D and b12 Levels in the Covid-19 Positive Pregnant Women)
https://clinicaltrials.gov/ct2/show/NCT04407572
Case-control study investigating whether vitamin D levels affect outcomes in COVID-19 infection and whether vitamin D deficiency is associated with increased risk. (Investigating the Role of Vitamin D in the Morbidity of COVID-19 Patients)
https://clinicaltrials.gov/ct2/show/NCT04386044
Prospective cohort study investigating the association between vitamin D deficiency and worse outcomes in people admitted to hospital for COVID-19. (Increased Risk of Severe Coronavirus Disease 2019 in Patients With Vitamin D Deficiency (COVIT-D))
https://clinicaltrials.gov/ct2/show/NCT04403932
Retrospective Pilot Study of Vitamin D Status and Immune-inflammatory Status in Different UK Populations With COVID-19 Infection
https://ClinicalTrials.gov/show/NCT04519034
Prognostic Factors and Outcomes of COVID-19 Cases in Ethiopia: Multi-Site Cohort Study. To determine the epidemiological and clinical features of COVID-19 cases, immunological
and virological courses, interaction with nutritional status, and response to treatment for COVID-19 patients admitted to treatment centers in Ethiopia.
https://ClinicalTrials.gov/show/NCT04584424
Observational study to investigating the Role of Vitamin D in the Morbidity of COVID-19 Patients.
https://ClinicalTrials.gov/show/NCT04386044
An observational study on Vitamin D-related Polymorphisms and Vitamin D Levels as Risk Biomarkers of COVID-19 Infection Severity. (VITACOV).
https://ClinicalTrials.gov/show/NCT04370808
Prospective Cohort Study to Determine the Association Between Vitamin D Deficiency and Severity of the Disease in Patients With Coronavirus Disease.
https://ClinicalTrials.gov/show/NCT04403932
A Case-Control Study on N-terminal Pro B-type Natriuretic Peptide and Vitamin D Levels as Prognostic Markers in COVID-19 Pneumonia.
https://ClinicalTrials.gov/show/NCT04487951
A Case-Control Study to investigate Whether Vitamin D Levels Really Correlated With Disease Severity in COVID-19 Patients. (COVIDVIT)
https://ClinicalTrials.gov/show/NCT04394390
Retrospective observational unicentric study in nursing-home residents with COVID-19. Health status monitoring data available until May 15, 2020. For all participants, gender, age, disability, history and comorbidities, treatments, date of last vitamin D3 supplementation, results of last blood test, date of suspicion / diagnosis of COVID-19, COVID-19 OSCI score, and eventual hospitalization or death (surveillance data available until May 15, 2020) are collected.
https://ClinicalTrials.gov/show/NCT04435119
A case=Control Study of the Relationship Between Zinc Vitamin D and b12 Levels in the Covid-19 Positive Pregnant Women.
https://ClinicalTrials.gov/show/NCT04407572
Treatment
| Study | Code [Reason] |
| Lee, Joseph, van Hecke, Oliver, Roberts, Nia et al. (2020) Vitamin D: A rapid review of the evidence for treatment or prevention in COVID- 19. | - Not a peer-reviewed publication
A review published by CEBM posted as a webpage and not peer-reviewed. |
| Tan, Chuen Wen, Ho, Liam Pock, Kalimuddin, Shirin et al. (2020) Cohort study to evaluate effect of vitamin D, magnesium, and vitamin B12 in combination on severe outcome progression in older patients with coronavirus (COVID-19).
Nutrition (Burbank, Los Angeles County, Calif.) 7980: 111017 |
- Comparator in study does not match that specified in protocol
Vitamin D was used as a combination therapy, which was not balanced in the comparator arm. |
| Study | Code [Reason] |
| Sharma, S.K., Mudgal, S.K., Pai, V.S. et al. (2020) Vitamin d: A cheap yet effective bullet against coronavirus disease-19 - are we convinced yet?. National Journal of Physiology, Pharmacy and Pharmacology 10(7): 511-518 | - Review article but not a systematic review |
| Yousfi, Narimen, Bragazzi, Nicola Luigi, Briki, Walid et al. (2020) The COVID-19 pandemic: how to maintain a healthy immune system during the lockdown - a multidisciplinary approach with special focus on athletes. Biology of sport 37(3): 211-216 | - Review article but not a systematic review |
| Study | Code [Reason] |
| Abrishami, Alireza, Dalili, Nooshin, Mohammadi Torbati, Peyman et al. (2020) Possible association of vitamin D status with lung involvement and outcome in patients with COVID-19: a retrospective study. European journal of nutrition | - Study does not contain any relevant predictive values
Unclear what had been adjusted in the meta- analysis. |
| Alexander, Jan, Tinkov, Alexey, Strand, Tor A et al. (2020) Early Nutritional Interventions with Zinc, Selenium and Vitamin D for Raising Anti- Viral Resistance Against Progressive COVID-
19. Nutrients 12(8) |
- Review article but not a systematic review |
| Ali, Nurshad (2020) Role of vitamin D in preventing of COVID-19 infection, progression | - Review article but not a systematic review |
| Study | Code [Reason] |
| and severity. Journal of infection and public health 13(10): 1373-1380 | |
| Anonymous (2020) Do Low Vitamin D Levels Increase COVID-19 Risk?. The American journal of nursing 120(11): 16 | - Not a relevant study design
Letter to the journal |
| Arvinte, Cristian; Singh, Maharaj; Marik, Paul E (2020) Serum Levels of Vitamin C and Vitamin D in a Cohort of Critically Ill COVID-19 Patients of a North American Community Hospital Intensive Care Unit in May 2020: A Pilot Study. Medicine in drug discovery 8: 100064 | - Study does not contain any relevant predictive values
Dependent variable of interest is vitamin C, not vitamin D. |
| Baktash, Vadir, Hosack, Tom, Patel, Nishil et al. (2020) Vitamin D status and outcomes for hospitalised older patients with COVID-19.
Postgraduate medical journal |
- Study does not contain any relevant predictive values
Unadjusted analyses |
| Benskin, Linda L (2020) A Basic Review of the Preliminary Evidence That COVID-19 Risk and Severity Is Increased in Vitamin D Deficiency. Frontiers in public health 8: 513 | - Review article but not a systematic review |
| Carpagnano, G E, Di Lecce, V, Quaranta, V N et al. (2020) Vitamin D deficiency as a predictor of poor prognosis in patients with acute respiratory failure due to COVID-19. Journal of endocrinological investigation | - Study does not contain any relevant predictive values
Unadjusted analysis. |
| D’Avolio, Antonio, Avataneo, Valeria, Manca, Alessandra et al. (2020) 25-Hydroxyvitamin D Concentrations Are Lower in Patients with Positive PCR for SARS-CoV-2. Nutrients 12(5) | - Study does not contain any relevant predictive values
Unadjusted analysis. |
| de Lucena, Thays Maria Costa, da Silva Santos, Ariane Fernandes, de Lima, Brenda Regina et al. (2020) Mechanism of inflammatory response in associated comorbidities in COVID-19.
Diabetes & metabolic syndrome 14(4): 597-600 |
- Review article but not a systematic review Review methodology states that it systematically searched databases but no review protocol was provided. Review details physiology of vitamin D in the immune response but does not provide studies that show effectiveness of vitamin D in prevention or treatment, or the association between vitamin D status and covid-19 infection. |
| Eugene, Merzon, Dmitry, Tworowski, Alessandro, Gorohovski et al. Low plasma 25(OH) vitamin D3 level is associated with increased risk of COVID-19 infection: an Israeli population-based study. | - Duplicate reference |
| Fasano, Alfonso, Cereda, Emanuele, Barichella, Michela et al. (2020) COVID-19 in Parkinson’s Disease Patients Living in Lombardy, Italy. Mov. Disord | - Study does not contain any relevant predictive values
The study included ‘probable’ unconfirmed cases in the analysis, and the analysis is unadjusted. |
| Fasano, Alfonso, Cereda, Emanuele, Barichella, Michela et al. (2020) COVID-19 in Parkinson’s | - Study does not contain any relevant predictive values |
| Study | Code [Reason] |
| Disease Patients Living in Lombardy, Italy. Movement disorders : official journal of the Movement Disorder Society 35(7): 1089-1093 | The study included ‘probable’ unconfirmed cases in the analysis, and the analysis is unadjusted. |
| Faul, J.L., Kerley, C.P., Love, B. et al. (2020) Vitamin d deficiency and ards after sars-cov-2 infection. Irish Medical Journal 113(5): p84 | - Study does not contain any relevant predictive values
Unadjusted analysis. |
| Galmes, Sebastia; Serra, Francisca; Palou, Andreu (2020) Current State of Evidence: Influence of Nutritional and Nutrigenetic Factors on Immunity in the COVID-19 Pandemic Framework. Nutrients 12(9) | - Study does not contain any relevant predictive values
Ecological study using countries’ population mineral intake as parameter, then simple correlation to COVID-19 incidence or death. |
| Goncalves, T.J.M., Goncalves, S.E.A.B., Guarnieri, A. et al. (2020) Prevalence of obesity and hypovitaminosis D in elderly with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Clinical Nutrition ESPEN | - Study does not contain any relevant predictive values
Unadjusted analysis. |
| Gromova, O. A.; Yu, T. I.; Kh, G. G. (2020)
COVID-19 pandemic: Protective role of vitamin D. Farmakoekonomika 13(2): 132-145 |
- Study not reported in English
Published in Russian |
| Haj Bloukh, Samir, Edis, Zehra, Shaikh, Annis A et al. (2020) A Look Behind the Scenes at COVID-19: National Strategies of Infection Control and Their Impact on Mortality.
International journal of environmental research and public health 17(15) |
- Study does not contain any relevant predictive values
Narrative review of ecological studies using worldwide geographical distribution of SARS CoV2 from ECDC
- Review article but not a systematic review |
| Hamza, A., Ahmed, M., Ahmed, K. et al. (2020) Role of Vitamin D in Pathogenesis and Severity of Coronavirus Disease 2019 (COVID-19) Infection. Pakistan Journal of Medical and Health Sciences 14(2): 462-465 | - Study does not contain any relevant predictive values
Unadjusted values. |
| Hribar, Casey A; Cobbold, Peter H; Church, Frank C (2020) Potential Role of Vitamin D in the Elderly to Resist COVID-19 and to Slow Progression of Parkinson’s Disease. Brain sciences 10(5) | - Review article but not a systematic review |
| Ilie, Petre Cristian; Stefanescu, Simina; Smith, Lee (2020) The role of vitamin D in the prevention of coronavirus disease 2019 infection and mortality. Aging clinical and experimental research | - Study does not contain any relevant predictive values
Unadjusted analysis. |
| Im, Jae Hyoung, Je, Young Soo, Baek, Jihyeon et al. (2020) Nutritional status of patients with COVID-19. International journal of infectious diseases : IJID : official publication of the International Society for Infectious Diseases 100: 390-393 | - Study does not contain any relevant predictive values
Unadjusted analysis. |
| Study | Code [Reason] |
| Jovic, Thomas H, Ali, Stephen R, Ibrahim, Nader et al. (2020) Could Vitamins Help in the Fight Against COVID-19?. Nutrients 12(9) | - Review article but not a systematic review
- Does not contain a cohort of people and therefore no extractable data Also included animal studies. |
| Kara, Murat, Ekiz, Timur, Ricci, Vincenzo et al. (2020) ‘Scientific Strabismus’ or Two Related Pandemics: COVID-19 & Vitamin D Deficiency. The British journal of nutrition: 1-20 | - Study does not contain any relevant predictive values
Unadjusted analysis. |
| Karonova, T. L.; Andreeva, A. T.; Vashukova,
M. A. (2020) serum 25(oH)D level in patients with CoVID-19. Jurnal Infektologii 12(3): 21-27 |
- Study not reported in English
Reported in Russian. Only abstract in English. |
| Karonova, T.L., Vashukova, M.A., Gusev, D.A. et al. (2020) Vitamin D deficiency as a factor for immunity stimulation and lower risk of acute respiratory infections and COVID-19. Arter.
Hypertens. 3(26): 295-303 |
- Study not reported in English
Publication only available in Russian. |
| Laird, E.; Rhodes, J.; Kenny, R.A. (2020) Vitamin D and inflammation: Potential implications for severity of Covid-19. Irish Medical Journal 113(5): p81 | - Study does not contain any relevant predictive values
Unadjusted analysis. |
| Macaya, Fernando, Espejo Paeres, Carolina, Valls, Adrian et al. (2020) Interaction between age and vitamin D deficiency in severe COVID- 19 infection Interaccion entre la edad y el deficit de vitamina D en la COVID-19 grave. Nutricion hospitalaria | - Duplicate reference |
| Maghbooli, Zhila, Sahraian, Mohammad Ali, Ebrahimi, Mehdi et al. (2020) Vitamin D sufficiency, a serum 25-hydroxyvitamin D at least 30 ng/mL reduced risk for adverse clinical outcomes in patients with COVID-19 infection. PloS one 15(9): e0239799 | - Study does not contain any relevant predictive values
Multivariable analysis was conducted, but only p values were reported. Univariate analyses were reported more fully, but were not adjusted for confounding variables as specified in the protocol. |
| Mardani, R, Alamdary, A, Mousavi Nasab, S D et al. (2020) Association of vitamin D with the modulation of the disease severity in COVID-19. Virus research 289: 198148 | - Study does not contain outcomes of interest Only reports indirect outcome of interest (simple correlation with ACE level). |
| McKenna, M.J. and Flynn, M.A.T. (2020) Covid- 19, cocooning and vitamin d intake requirements. Irish Medical Journal 113(5): p79 | - Not a peer-reviewed publication
- Study does not contain any relevant predictive values
- Data not reported in an extractable format |
| Munshi, Ruhul, Hussein, Mohammad H, Toraih, Eman A et al. (2020) Vitamin D insufficiency as a potential culprit in critical COVID-19 patients. Journal of medical virology | - Study does not contain any relevant predictive values
Systematic review does not meet protocol inclusion criteria, also using flawed statistical |
| Study | Code [Reason] |
| analysis (pooling unadjusted data that with severe heterogeneity with I2 = 99.1%). | |
| Namayandeh, S. M. (2020) Vitamin D and coronavirus disease (COVID-19);is deficiency and maintenance supplementation therapy necessary?. Journal of Nutrition and Food Security 5(3): 187-191 | - Review article but not a systematic review |
| Orru, B, Szekeres-Bartho, J, Bizzarri, M et al. (2020) Inhibitory effects of Vitamin D on inflammation and IL-6 release. A further support for COVID-19 management?. European review for medical and pharmacological sciences 24(15): 8187-8193 | - Review article but not a systematic review |
| Padhi, S., Suvankar, S., Panda, V.K. et al. (2020) Lower levels of vitamin D are associated with SARS-CoV-2 infection and mortality in the Indian population: An observational study.
International Immunopharmacology 88: 107001 |
- Study does not contain any relevant predictive values
Unclear where the study obtained the vitamin D status data.
- Study does not contain outcomes of interest Study reports simple unadjusted correlation r. |
| Pereira, Marcos, Dantas Damascena, Alialdo, Galvao Azevedo, Laylla Mirella et al. (2020) Vitamin D deficiency aggravates COVID-19: systematic review and meta-analysis. Critical reviews in food science and nutrition: 1-9 | - Study does not contain any relevant predictive values
Inappropriate meta-analysis: heterogenous baseline characteristics of index studies, point estimates from index studies some are adjusted (on different variables) and some are unadjusted. |
| Pizzini, Alex, Aichner, Magdalena, Sahanic, Sabina et al. (2020) Impact of Vitamin D Deficiency on COVID-19-A Prospective Analysis from the CovILD Registry. Nutrients 12(9) | - Study does not contain any relevant predictive values
Unadjusted analysis. |
| Roselin, C. and Parameshwari, S. (2020) Role of vitamin d in boosting immunity against covid-
19. International Journal of Research in Pharmaceutical Sciences 11(specialissue1): 425-429 |
- Review article but not a systematic review |
| Rozga, Mary, Cheng, Feon W., Moloney, Lisa et al. (2020) Effects of Micronutrients or Conditional Amino Acids on COVID-19 Related Outcomes: An Evidence Analysis Center Scoping Review. Journal of the Academy of Nutrition and Dietetics | - Does not contain a population of people with COVID-19
One study was identified that assess the effect of cholecalciferol in patients with ventilator- related pneumonia. |
| Singh, S K; Jain, Rujul; Singh, Shipra (2020) Vitamin D deficiency in patients with diabetes and COVID- 19 infection. Diabetes & metabolic syndrome 14(5): 1033-1035 | - Review article but not a systematic review |
| Singh, Samer; Kaur, Rajinder; Singh, Rakesh Kumar (2020) Revisiting the role of vitamin D levels in the prevention of COVID-19 infection | - Study does not contain any relevant predictive values |
| Study | Code [Reason] |
| and mortality in European countries post infections peak. Aging clinical and experimental research 32(8): 1609-1612
Tan, Si Heng Sharon, Hong, Choon Chiet, Saha, Soura et al. (2020) Medications in COVID-19 patients: summarizing the current literature from an orthopaedic perspective. International orthopaedics
Yalcin Bahat, Pinar, Aldikactioglu Talmac, Merve, Bestel, Aysegul et al. (2020) Micronutrients in COVID-19 Positive Pregnancies. Cureus 12(9): e10609
Yilmaz, Kamil and Sen, Velat (2020) Is vitamin D deficiency a risk factor for COVID-19 in children?. Pediatric pulmonology
Yousfi, Narimen, Bragazzi, Nicola Luigi, Briki, Walid et al. (2020) The COVID-19 pandemic: how to maintain a healthy immune system during the lockdown - a multidisciplinary approach with special focus on athletes. Biology of sport 37(3): 211-216 |
Unclear where the study obtained the vitamin D status data, no baseline characteristics data.
- Study does not contain outcomes of interest Results presented as simple unadjusted correlation r.
- Data not reported in an extractable format
- Study does not contain outcomes of interest
- Study does not contain outcomes of interest Outcomes were blood levels of micronutrients including vitamin D. No correlation between vitamin D status and disease severity or any other outcome relevant to the review protocol was reported.
- Not a relevant study design Case series of pregnant women who tested positive for SARS-CoV2 infection and had their blood vitamin D levels measured.
- Study does not contain any relevant predictive values Unadjusted analysis.
- Review article but not a systematic review |