Take Home Points:
This study analyzed the prevalence of vitamin D insufficiency (VDI) among SARS-CoV-2 positive subjects hospitalized in a medical center situated in New Orleans. Through medical records, the authors performed a retrospective review of 20 Covid-19 patients with serum 25-hydroxycholecalciferol (25ODH) levels determined. Vitamin D insufficiency (VDI) was defined as a serum 25OHD < 30 ng/mL. Among these patients, 65% required intensive care unit (ICU). When they analyzed the 25ODH levels, 84.6% of ICU subjects presented VDI in compare to 57.1% of floor subjects. Curiously, 100% of ICU patients less than 75 years old had VDI. In addition, the authors demonstrated that lymphocytopenia was present in 92.3% of the Covid-19 patients admitted to ICU. Among the clinical and demographic characteristics of SARS-Cov-19 ICU patients of this study we can highlight: 84% of the subjects were African American, hypertension and diabetes was presented in 76.9% and 46.2% ICU and non-ICU patients respectively, the body mass index of 100% of the patients was considered overweight or obesity, and all had high levels of lactate dehydrogenase.
-----------------------------------------------------------------------------
I wonder about results from producing vitamin D from being out in the sun versus vitamin D supplements. For now I think it is good to do both until I review more.
-----------------------------------------------------------------------------
Vitamin D insufficiency is prevalent in severe COVID-19
Frank H. Lau, MD, FACS1, Rinku Majumder, PhD2, Radbeh Torabi, MD1, Fouad Saeg, BS3, Ryan Hoffman, BS4, Jeffrey D. Cirillo, PhD5, Patrick Greiffenstein, MD, FACS1
Affiliations:
1. Department of Surgery, Louisiana State University Health Sciences Center New Orleans (LSUHSC-NO) 2. Department of Biochemistry, LSUHSC-NO 3. Tulane School of Medicine 4. LSUHSC-NO School of Medicine 5. Department of Microbial Pathogenesis and Immunology, Texas A&M College of Medicine
Corresponding Author:
Dr. Frank H. Lau, MD, FACS; Department of Surgery, Louisiana State University Health Sciences Center New Orleans, 1542 Tulane Ave, 7th Floor, New Orleans, LA 70112; email:
[email protected]; phone: (504) 412-1240
All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprintthis version posted April 28, 2020. .
https://doi.org/10.1101/2020.04.24.20075838doi: medRxiv preprint
Introduction
The novel SARS-CoV-2 virus causes COVID-19 and has resulted in 2.8 million confirmed cases and more than 196,000 deaths. Strikingly, 80-85% of patients are asymptomatic or have self-limiting disease.1 The
remaining require major hospital resources and threaten to collapse our healthcare system. The mechanisms
underlying divergent COVID-19 outcomes are unknown.
Emerging health disparities data are potentially illuminating. In Louisiana, African Americans account for 70% of COVID-19 deaths despite representing only 32% of the population.2 In a Boston homeless shelter, 100% of 147 COVID-19 positive subjects were asymptomatic.3 The mechanisms underlying severe
COVID-19 should account for both these findings, as well as other COVID-19 mortality risk factors:
hypertension, obesity, male sex, advanced age, concentration in northern climates, and COVID-19 associated coagulopathy (CAC).4,5
Vitamin D insufficiency (VDI) meets every one of the above criteria. VDI affects 80-90% of the African
American population. In contrast, homeless persons generally have poorer health and nutrition, but can have greater exposure to sunlight, the source of 80-90% of the body’s vitamin D.6 VDI causes essential hypertension and is associated with every COVID-19 mortality risk factor.7,8 Hydroxychloroquine raises plasma vitamin D levels.9 Lastly, VDI induces a prothrombotic state and adversely impacts both innate and
adaptive immune responses. To better define the VDI-COVID-19 link, we determined the prevalence of
VDI among our COVID-19 intensive care unit (ICU) patients.
Methods
In an Institutional Review Board approved study performed at a single, tertiary care academic medical
center, the medical records of COVID-19 patients between March 27, 2020 and April 21, 2020 were
retrospectively reviewed. Subjects were included for whom serum 25-hydroxycholecalcifoerol (25OHD)
levels were determined. COVID-19-relevant data were compiled and analyzed. The 25OHD assay was
performed in-house, using a UniCel DxI 600 Access Immunoassay System (Beckman Coulter); the laboratory undergoes recertification every six months. VDI was defined as serum 25OHD < 30 ng/mL.10
Results
Twenty COVID-19 patients with serum 25OHD levels were identified; 13 (65.0%) required ICU admission.
Overall, few significant differences were identified between ICU and floor patients (Table 1) but statistical
analysis was limited by the small number of subjects. Lactate dehydrogenase on admission was significantly
higher among ICU patients (441.8 vs. 223.0, p=0.001), consistent with previous reports. No patients were
diagnosed with stroke, myocardial infarction, or pulmonary embolus. Two patients (10%) died during the
study period.
Among ICU subjects, 11 (84.6%) had VDI, vs. 4 (57.1%) of floor subjects. Strikingly, 100% of ICU patients less than 75 years old had VDI (n=11; Table 2). Among these, 64.6% (n=7) had critically low 25OHD (<20 ng/mL) and three had <10 ng/mL. The sepsis-induced coagulopathy score (SIC) was calculable for 8
subjects; 62.5% (n=5) had SIC ≥ 4. Suppressed immune function was prevalent: 92.3% (n=12) were
lymphocytopenic, and 9 were profoundly so (absolute lymphocyte count ≤ 0.4 10^3/uL; normal range 1.105.00).
Discussion
COVID-19 is an emerging disease whose pathogenic mechanisms are not well understood. Despite being
an acute respiratory infection (ARI), its mortality risk factors overlap those of cardiovascular disease:
hypertension, diabetes, obesity, advanced age, and male sex. From a health disparities perspective, notable
features include an over-representation of African Americans among COVID-19 deaths, and a 100%
asymptomatic presentation in a universal survey of a Boston homeless shelter.
Interestingly, VDI and COVID-19 share prevalence patterns for hypertension, diabetes, obesity, advanced
age, and male sex (Table 3). VDI can contribute to our understanding of COVID-19 health disparities: VDI
is highly prevalent in dark-skinned persons (82.1% of African Americans vs. 41.6% overall). In contrast,
although U.S. homeless persons are generally considered to have poor health and decreased access to
micronutrients that confer immune benefits, they usually have more exposure to sunlight, a key source of
vitamin D production. In Europe, COVID-19 has been severe in Italy, Spain and Greece, but much less so
in Scandinavian countries – this precisely recapitulates VDI data showing that Italy, Spain and Greece have VDI rates of 70-90%, vs. 15-30% in Norway and Denmark.11 Scandinavian diets contain more vitamin D due to higher fatty fish intake and dairy products supplementation with vitamin D.11
The baseline prevalence of VDI amongst ICU patients is 30-40%.12 In this study, we found that 84.6% of COVID-19 ICU patients had VDI, vs. 57.1% of floor patients. Strikingly, 100% of ICU patients less than 75 years old had VDI. We also found that 62.5% had CAC, and 92.3% had lymphopenia. Given these data, we hypothesize that VDI enhances COVID-19 severity via 1) its prothrombotic effects and 2) its
derangement of the immune response.
Prothrombosis
CAC is emerging as a key process in severe COVID-19. The American Society of Hematology recommends routine DVT prophylaxis for all admitted COVID-19 patients.13 In Wuhan, CAC was present in 71.4% of non-survivors vs. 0.6% in survivors.14 Non-survivors demonstrated significantly lower fibrinogen and
antithrombin levels on admission, consistent with coagulation factor depletion induced by a
hypercoagulable state. A meta-analysis of 1,779 COVID-19 patients reported that platelet counts were significantly lower in severe COVID-19, and that lower platelet counts were associated with mortality.15
Anticoagulation can lower mortality: in patients with high SIC scores or D-dimer levels >6-fold the upper limit of normal, heparin reduced mortality to 40.0% vs. 64.2% in controls.16 CAC’s role is further evidenced
by the multiorgan, microvascular clots in hospitalized COVID-19 patients, which include deep vein
thromboses/pulmonary emboli, acute renal failure, cerebrovascular events, myocardial injury, ischemic stroke, and ischemic skin changes.4 Microthromboses are found in extrapulmonary organs at a rate greater than in severe acute respiratory syndrome (SARS), another novel coronavirus.17
VDI is prothrombotic, since Vitamin D receptor knockout (VDRKO) mice develop a CAC-like response to injury, with aggravated, multiorgan thrombosis following lipopolysaccharide injection.18 Expression of
antithrombin in the liver and thrombomodulin in the aorta, liver, and kidney were downregulated, whereas
tissue factor expression was upregulated in the liver and kidney. In humans, VDI is associated with increased risk of CVD and death.8 Vitamin D receptor knockout (VDRKO) mice exhibited increased
thrombogenic activity and increased ADP-induced platelet aggregation.
VDR exist in all major cardiovascular cell types, including cardiomyocytes, arterial wall cells, and immune
cells. Studies have established that vitamin D metabolites are integral to vascular function and disease,
including inflammation and thrombosis. For example, 1,25(OH)2D exerts anticoagulant effects by
upregulating the expression of thrombomodulin (an anticoagulant glycoprotein) and downregulating the
expression of tissue factor (a critical coagulation factor) in monocytes and human aortic smooth muscle cells.19
Deranged Immunity
Lymphocytopenia is a hallmark of severe COVID-19, suggesting a deranged immune response.
Convalescent plasma therapy, a form of passive therapy, can improve the deranged response. Convalescent
therapy may improve the host response by reducing VDI or through a protective humoral response.
Differentiating these two possibilities is critical to identifying strategies for mitigating severe COVID-19.
VDI leads to deranged immune response, including to viral ARIs. In a study of over 14,000 individuals,
VDI was associated with a 58% increase in ARI after controlling for seasonal, demographic, and clinical factors.20 A form of macrophage activation syndrome (MAS), where macrophages are highly activated from
the initial systemic inflammatory response to SARS-CoV, could be responsible for hyperferritinemia and lead to or exacerbate VDI.21 It has been proposed that uncontrolled inflammation commonly leads to hyperferritinemia and likely results in immune dysregulation.22 Hyperferritinemia appears to be due to the
hemophagocytosis and hypercytokinemia observed in MAS. This clinical picture can result in a
prothrombotic state that is consistent with hemophagocytic lymphohistiocytosis and can be observed as a consequence of viral infection.23 Overall, these data suggest that an overly exuberant inflammatory
response leads to VDI, hyperferritinemia and the prothrombotic state observed with COVID-19.
Vitamin D plays an essential role in modulating both the innate and adaptive immune response.24,25 If VDI
correlates with severe COVID-19, it would likely explain the high frequency of severe disease in the >60
year old and African American population. Vitamin D-dependent antimicrobial pathways are induced in response to double-stranded RNA, as produced during SARS-CoV-2 replication.26 These pathways
upregulate antimicrobial peptides, including cathelicidin and -defensin, and autophagy. In macrophages
and endothelial cells, cathelicidin production is modulated in a vitamin D dose-dependent manner.27–29 IFN
is strongly antimicrobial and a key activator of these pathways, particularly for macrophages and other
phagocytic cells, resulting in greater production of reactive oxygen through an oxidative burst and nitrogen
species, in addition to these antimicrobial peptides. The over-utilization of these pathways by the host
response to SARS-CoV-2 designed to control viral replication could be one mechanism by which VDI
arises initially, but once VDI is present, this response becomes ineffective.
VDI prevents the ability of the host to activate these host defensive pathways, but has also been shown to play an important role in macrophage and lymphocyte migration.30 Interestingly, the bioactive form of
vitamin D, 1,25-dihydroxyvitamin D3 prevents experimental autoimmune encyphalomyelitis (EAE),
suggesting that one aspect of the pathology associated with the final stages of deranged immunity observed
in COVID-19 may be EAE through inflammation caused by uncontrolled trafficking of macrophages and
T cells into the CNS. The massive influx of macrophages and T cells into peripheral organs, including the
CNS, may represent the mechanism by which lymphocytopenia arises and could suggest an autoimmune
component to the disease. The concept that COVID-19 may impact the CNS is supported by reports of loss of smell and taste by patients and the elevated incidence of ischemic stroke.31,32
The Case Against VDI
While VDI is associated more frequently with ARIs, Vitamin D supplementation does not consistently show benefit against influenza.33 However, these outbreaks were not marked by coagulopathy. Furthermore,
previous trials did not identify subjects with VDI, thereby introducing a major confounding variable.
Conclusions
This small, retrospective observational study suggests a link between VDI and severe COVID-19.
Anecdotal and observational data indicate that VDI may play a significant role in the progression of the
COVID-19 disease state. Low-risk, high-reward potential therapies that target CAC and VDI merit further
investigation. Prospective, randomized controlled studies that properly risk-stratify subjects should be
performed.
Acknowledgements
We appreciate the University Medical Center of New Orleans for providing access to their patient data.
https://www.medrxiv.org/content/10.1101/2020.04.24.20075838v1.full.pdf