key: cord-0918727-jdrhu6l1 authors: Backer, A.; Mageswaran, M. title: Double COVID-19 Confirmed Case Fatality Rate in Countries with High Elderly Female Vitamin D Deficiency Prevalence date: 2020-06-16 journal: nan DOI: 10.1101/2020.06.13.20130484 sha: c512d6575c7d49c93eff409ba57f150edf159bfe doc_id: 918727 cord_uid: jdrhu6l1 A number of clues point to a possible role of vitamin D in fighting COVID-19: a reduction in case growth speed with solar zenith angle, higher fatality rate in black people, lower fatality rate in populations that spend more time outdoors. Yet a direct demonstration that vitamin D deficiency is associated with COVID-19 fatalities has remained elusive. We show here in a comparison of 32 countries that countries with high prevalence of vitamin D deficiency among elderly females show a confirmed case fatality rate twice as high as those with low prevalence. We then show that this effect cannot be explained by differences in life expectancy between countries. A mechanistic role for vitamin D in the severity of COVID-19 is proposed. The world is in the grip of the COVID-19 pandemic. Public health measures that can reduce the 25 risk of infection and death in addition to quarantines are desperately needed. The world is undergoing the largest public policy experiment ever conducted. Billions of people are being told to stay at home to avoid COVID-19 contagion. While isolation undoubtedly helps prevent contagion, having people stay at home and closing public parks has side effects that are not fully understood. UV radiation in natural sunlight stimulates vitamin D production in the 30 hours and days following exposure 1 . At least a billion people worldwide are estimated to be vitamin D deficient, mainly because of inadequate exposure to sunlight and inadequate fortification of food with vitamin D 2 . Interventional and observational epidemiological studies provide evidence that vitamin D deficiency may confer increased risk of influenza and respiratory tract infection 3,4 . Vitamin D deficiency is also prevalent among patients infected with 35 another RNA virus, HIV. Cell culture experiments support the thesis that vitamin D has direct anti-viral effects particularly against enveloped viruses. Though vitamin D's anti-viral mechanism has not been fully established, it may be linked to vitamin D's ability to up-regulate the anti-microbial peptides LL-37 and human beta defensin 2 5 . Sunlight exposure improves survival in patients for some diseases 6 . The SARS coronavirus is inactivated under ultraviolet 40 irradiation 7 . Viruses containing single-stranded nucleic acids such as coronaviruses are more sensitive, unable to repair damage in the absence of a complementary strand. Vitamin D deficiency has been associated with increased risk of autoimmune diseases, infectious diseases and cardiovascular disease 8 . Studies from the Netherlands and France suggest that clots arise in 20-30% of critically ill COVID-19 patients 9, 10 . Vitamin D has anticoagulant effects 11,12 . 5 Are governments inadvertently putting populations at risk by locking them away from the sunlight for months at a time? Should public policy to slow the COVID-19 pandemic include both social isolation and sun exposure or vitamin D intake? Portugal, which borders Spain, has a COVID-19 mortality rate 75% lower than Spain. Norway and Denmark have low numbers of COVID-19 deaths, too, despite low winter sunlight levels. 10 Addition of vitamin D to margarine is compulsory in Norway, Denmark, the Netherlands, Belgium and Portugal. COVID 19 is much more lethal to the elderly, and older populations tend to have much higher levels of vitamin D insufficiency 13, 14, 15 . 43% of US COVID 19 deaths have occurred in nursing homes 16 . 84-93% of US nursing home 15 patients have insufficient vitamin D levels 17, 18 . Relevant amounts of vitamin D are found in seafood, egg yolk, butter and some mushrooms. Japan, a country with high fish consumption 19 , has been relatively spared, with less than 3% of the COVID 19 mortality of the US 20 . The synthesis of vitamin D in human skin is a two-stage process that begins with the production 20 of previtamin D after irradiation of 7-dehydrocholesterol by ultraviolet (UV) radiation. A number of factors control the probability of a suitable UV photon reaching a molecule of 7dehydrocholesterol in the skin. These are astronomical factors that govern the solar zenith angle (SZA), and the local state of the atmosphere, determining the available solar UV radiation; skin pigmentation and age, determining competing absorbers of UV radiation and available 7-25 dehydrocholesterol; individual behavior in the local surroundings, determining exposure of unprotected skin to available UV radiation. The only one of these influences that can be determined unequivocally for any situation is the SZA. At noon, SZA is at a minimum and is equal to latitude minus solar declination angle. At large SZAs there is insufficient solar UV radiation to initiate significant vitamin D synthesis 21 . 30 One of us has previously shown SZA is the best predictor we know for COVID-19 case growth speed 22 . And yet, the topic remains controversial, and recent reviews maintain there is no evidence of a link between vitamin D and COVID-19 23,24,25,26 . To address this question, we compared the prevalence of severe vitamin D deficiency (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) hydroxyvitamin D serum concentration below 12 ng/ ml (30 nmol/l)) and intake levels in different countries 27, 28, 29, 30, 31, 32, 33, 34 , with COVID-19 case growth rates and confirmed case fatality rates (cCFR). COVID-19 case growth speed was not significantly correlated with the prevalence of subclinical vitamin D deficiency in the female elderly population. This result is consistent with a recent 40 study of the UK Biobank 35 . . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 16, 2020. . https://doi.org/10.1101/2020.06.13.20130484 doi: medRxiv preprint A multiple linear regression model showed a significant correlation between both vitamin D deficiency in the female, but not male, elderly population and cCFR (Fig.1, p<0 .05). The significant association with vitamin D deficiency prevalence is particularly notable given that the seasonality of vitamin D levels is likely to have introduced noise in the data. Furthermore, the fact that the effect is significant even with national averages, which mask the effect in 5 individuals who are vitamin D deficient with others who are not, suggests the effect is likely much stronger when comparing individuals with different vitamin D levels. Noticing that some of the countries with high cCFR yet low vitamin D deficiency, such as Brazil, had a large black population, and given our previous result showing COVID-19 mortality is higher in Black populations in low irradiance locations 63 , we carried out the same analysis correlating COVID-19 cCFR with the percentage of the population that is Black, and found a 15 significant positive correlation (Fig. 2, p<0 .05). To disambiguate the impact of each of these effects, we developed a multivariate linear model to estimate the significance of vitamin D deficiency prevalence and black population % when evaluated jointly. Both factors were significantly associated with COVID-19 cCFR (Table 1) . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted June 16, 2020. . https://doi.org/10.1101/2020.06.13.20130484 doi: medRxiv preprint The prevalence of vitamin D deficiency among elderly females was not significantly correlated 5 with the % of Blacks in the population. Vitamin D intake was not significantly correlated with case growth rate or with cCFR. The elderly are particularly susceptible to COVID-19. Even in low latitudes where severe vitamin D deficiency is rare, there is frequent occurrence of subclinical vitamin D deficiency, especially in elderly people 36 . 10 Kumar and Srivastava recently claimed that average life expectancy could explain the apparent impact of vitamin D deficiency prevalence on COVID 19 37 . However, that study calculated cCFR with no lag between deaths and cases, which produces an inaccurate estimate of fatality. It also used a relatively small number of countries. We addressed both of those issues. We estimated confirmed case fatality rate as the average ratio between the number of new observed 15 deaths each day and the number of new deaths that would have been expected given the number of new cases per day convolved by the diagnosis to death lag curve, as described in the literature 38 CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted June 16, 2020. . https://doi.org/10.1101/2020.06.13.20130484 doi: medRxiv preprint significantly correlated with COVID-19 cCFR (Table 2; p<0.05). Interestingly, we were able to replicate Kumar and Srivastava's result when cCFR was measured the way they did, with constant zero lag between cases and deaths. Yet such an estimate fails to account for the significantly lower of cases that existed one lag before in an exponentially growing epidemic, and/or the number of cases yet to die. There was no significant correlation between each 5 country's cCFR and life expectancy, or between life expectancy and vitamin D deficiency prevalence. An analysis of global COVID-19 data using Causal Inference, constructing two contrasting models for vitamin D's role in COVID-19 fatality, one causal and one acausal, setting out clearly 20 multiple predictions made by each model, showed that observed data strongly matched predictions made by the causal model 42 . One in two fatal cases of COVID-19 experience a cytokine storm, 82% of whom are over the age of 60 43 . In older individuals, NLRP3 may be poised for hyperactivation by SARS-CoV-2 components 25 . Elimination of severe vitamin D deficiency reduces the risk of high CRP levels 25 (odds ratio of 2) which may be used as a surrogate marker of cytokine storm 44 . Vitamin D inhibits NLRP3 activation by impeding its deubiquitination mediated by BRCC3 45 . Vitamin D downregulates NLRP3 and protects against toxin-induced cell death 46 and modulates the NLRP3 inflammasome activation 47 . In addition, likely due to its role in vasodilation and reducing inflammation, ACE2 partially protects against sepsis-induced-and SARS-induced severe acute 30 lung injury in mice 48 . Vitamin D can increase the expressions of ACE2 mRNA and protein levels of ACE2 in rats 49 . Children have higher ACE expression than adults 50 , which could help explain their relative immunity to COVID-19. ACE2 expression is dramatically reduced with aging in rats of both genders: young-adult vs. old P < 0.001 (by 78% in male and 67% in female, respectively) and middle-aged vs. old P < 0.001 (by 71% in male rats and 59% in female rats, 35 respectively). The decrease of ACE2 content is relatively slight between young-adult and middle-aged groups (by 25% in male and 18% in female, respectively). Although there is no gender-related difference of ACE2 in young-adult and middle-aged groups, a significantly higher ACE2 content was detected in old female rats than male. In conclusion, the more elevated ACE2 in young adults as compared to aged groups may contribute to the predominance in SARS and 40 COVID-19 attacks in aged populations 51 . . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted June 16, 2020. . https://doi.org/10.1101/2020.06.13.20130484 doi: medRxiv preprint While men and women have the same COVID-19 prevalence, men with COVID-19 are more at risk for worse outcomes and death, independent of age 52 . Estrogen and testosterone impact immune system response and engagement, resulting in a less robust immunologic response in males and subsequent increased morbidity and mortality from viral respiratory illnesses 53 . Several animal models have demonstrated increased ACE2 activity in the male or 5 ovariectomized model 54 . Males show an upregulation of angiotensin II receptor type 1 (AGTR1), a member of the reninangiotensin system (RAS) that plays a role in angiotensin-converting enzyme 2 (ACE2) activity modulation. In vivo studies have showed that ovariectomized females, in addition to males, usually have higher levels of ACE2 activity compared to non-ovariectomized females. This 10 might indicate a possible role of the sex hormones in regulating ACE2 activity. Many reports highlight the role of human receptor ACE2, which also belongs to the renin-angiotensin system (RAS), in facilitating the binding of the SARS-CoV-2 to the host cells. While AGTR2 showed no significant difference in its expression between both genders, AGTR1 expression levels were higher in tissues obtained from male individuals compared to females 55 . 15 Female gender is associated with lower vitamin D levels (14.5 ± 10.9 vs. 15.9 ± 9.5, p = 0.007) and independently associated with severe vitamin D deficiency (41.9% vs. 30.4%, p < 0.001; adjusted odds ratio (OR) (95% confidence interval (CI)) = 1.42 (1.08-1.87), p = 0.01). Yet note that a French and Canadian study reported elderly males suffered from a serious vitamin D deficiency as compared to elderly females 56 . 20 Vitamin D and vitamin D receptor agonists modulate the immune system against over-reactivity towards tolerance; on this basis, vitamin D has been proposed as a therapeutic candidate to control autoimmune processes. Females retain greater immune reactivity and competence likely due to estrogens, which, at variance with androgens, are associated with a greater resilience to infections but also to a higher risk for autoimmunity 57 . Vitamin D deficiency is considered a risk 25 factor for autoimmune diseases. Vitamin D and its analogues have been proposed as therapeutic tools in autoimmunity considering their exquisite immunoregulatory effect against overreactivity towards tolerance. Beneficial effects against autoimmune processes seem to be allowed by vitamin D acting in synergy with estrogens 58 . These facts may explain why it's vitamin D deficiency among older females that is the stronger 30 predictor of COVID-19 cCFR. DeGroot and Bontrop 59 have pointed out that one of the prominent protein families that senses infection, and that is essential to the innate immune system, comprises the Toll-like receptors (TLR) (Akira and Takeda Whilst males may be more vulnerable to the respiratory symptoms of COVID-19, women may be more likely to show the auto-immune like inflammatory response that appears to be modulated by vitamin D. Vitamin D deficiency and insufficiency occurs in 29.2% and 54.1% of black women, respectively, contrasting with 5% and 42.1% of white women 60 . 20 A study of 660 women in South Los Angeles showed the prevalence of vitamin D deficiency among Hispanic women was 45% lower than among African American women 61 . One of us has previously shown that non-Black Hispanics have a significantly lower COVID-19 cCFR than non-Hispanics 64 . Any benefit of vitamin D needs to be balanced against the risk of toxicity, which is characterized 25 by hypercalcemia. Daily brief, suberythemal exposure of a substantial area of the skin to ultraviolet light, climate allowing, provides adults with a safe, physiologic amount of vitamin D, equivalent to an oral intake of about 10,000 IU vitamin D 3 per day, with the plasma 25- Inaction can be just as dangerous as action in the face of uncertainty. Several converging lines of evidence now suggest a role for sunlight in the prevention and treatment of COVID-19: sunnier places have slower COVID-19 growth 22 , African Americans and Black people in the UK have higher COVID-19 fatality rates in low irradiance locations 63 , US Latinos have higher sunlight exposure and lower COVID-19 fatality rates 64 , vitamin D-deficient populations such as the 40 elderly and those in nursing homes have higher COVID-19 mortality rates, and vitamin D deficiency is associated with significantly higher COVID-19 fatality rates. Although the possibility remains that another variable correlated with vitamin D (such as sunlight or something correlated with sunlight) is behind the observed effects, we propose that . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted June 16, 2020. . https://doi.org/10.1101/2020.06.13.20130484 doi: medRxiv preprint severe COVID-19 shows some characteristics of a virus-induced autoimmune disease, and that both sunlight and vitamin D be tested for its prevention and treatment. Whilst vitamin D supplementation has been causally linked in a randomized, double-blind, placebo controlled trial to the downregulation of interleukin-1 expression in peripheral blood mononuclear cells in humans 65 , and interleukin-1 Beta (IL-1β) is a critical component of lung 5 inflammation during viral infection 66 , it should be noted that no significant correlation was found between vitamin D intake and COVID-19 fatality. Until a clinical trial of the impact of vitamin D supplementation on COVID-19 is complete, when sunlight is available, the safest course of action may be for populations vulnerable to COVID-19 may be to simply take a little daily dose of sunshine. 10 Vitamin D levels are season-dependent. In Chile, for example, 20% of post-menopausal women have serum 25OHD concentrations below 15 ng/ml in summer, but 60% of them show that level of deficiency in winter. To address this issue, where possible, we used vitamin D deficiency 15 during the season that the COVID-19 epidemic broke out in each country. COVID-19 case growth rates were assessed by computing the number of cases on the Nth day after the 100 th case in each location. This method aligns by stage in the epidemic, is robust to later government interventions, and to idiosyncrasies about the first few cases in each location. cCFR were computed by dividing the number of COVID-19 deaths in each country on each day 20 between the start of the COVID-19 pandemic and June 10 th 2020 by the convolution of the number of cases in days 42 to 7 before and the distribution of lags between diagnoses and deaths 67 . https://www.medscape.com/viewarticle/760055 (2012) 19 Hayes, David. "Japan-Japan is the third largest fishery producer in the world and fish consumption is still very high despite the number of fishermen being on decline." World Fishing-Including Supplements 53.8 (2004): 12-15. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 16, 2020. . https://doi.org/10.1101/2020.06.13.20130484 doi: medRxiv preprint Rob Striker, Vitamin D and the anti-viral state More sunlight exposure may improve the overall survival in patients with pancreas cancer Inactivation of the coronavirus that induces severe acute respiratory syndrome, SARS-CoV Sunlight, UV-Radiation, Vitamin D and Skin Cancer: How Much Sunlight Do We Need Incidence of thrombotic complications in critically ill ICU patients with COVID-19 Pulmonary embolism in COVID-19 patients: awareness of an increased prevalence Augmentation of anticoagulant effect with vitamin D: possible therapeutic target for venous thromboembolism The relationship between vitamin D status and idiopathic lower-extremity deep vein thrombosis High prevalence of vitamin D insufficiency in healthy elderly people living at home in Argentina Prevalence of vitamin D insufficiency in elderly ambulatory outpatients in Denver, Colorado Prevalence of vitamin D deficiency, insufficiency and secondary hyperparathyroidism in the elderly inpatients and living in the community of the city of São Paulo, Brazil High Prevalence of Vitamin D Deficiency in Elderly Nursing Home Patients Despite Vitamin Supplements Low Vitamin D Raises Mortality Risk in Nursing Home Patients Who, what, where and when-influences on cutaneous vitamin D synthesis Slower COVID-19 Morbidity and Mortality Growth at Higher Solar Irradiance and Elevation Vitamin D and SARS-CoV-2 virus/COVID-19 disease Vitamin D and SARS-CoV-2 virus/COVID-19 disease Vitamin D: A rapid review of the evidence for treatment or prevention in COVID-19 Serum vitamin D concentrations among elderly people in Europe Current vitamin D status in European and Middle East countries and strategies to prevent vitamin D deficiency: a position statement of the European Calcified Tissue Society Prevalence of subclinical vitamin D deficiency in different European countries Is vitamin D deficiency a major global public health problem? High prevalence of vitamin D deficiency in Chilean healthy postmenopausal women with normal sun exposure: additional evidence for a worldwide concern Vitamin D deficiency and causative factors in the population of Tehran Vitamin D inadequacy in postmenopausal women in Eastern Asia Vitamin D deficiency in elderly: Risk factors and drugs impact on vitamin D status Vitamin D concentrations and COVID-19 infection in UK Biobank Vitamin D deficiency: a global perspective Spurious Correlation? A review of the relationship between Vitamin D and Covid-19 infection and mortality Real-time estimation of the risk of death from novel coronavirus (COVID-19) infection: inference using exported cases World Population Prospects: The 2019 Revision Patterns of COVID-19 Mortality and Vitamin D: An Indonesian Study Vitamin D insufficiency is prevalent in severe COVID-19 Evidence Supports a Causal Model for Vitamin D in COVID-19 Outcomes Why does COVID-19 disproportionately affect the elderly? The Role of Vitamin D in Suppressing Cytokine Storm in COVID-19 Patients and Associated Mortality Vitamin D receptor inhibits NLRP3 activation by impeding its deubiquitination mediated by BRCC3 Vitamin D downregulates NLRP3 and protects against Clostridium difficile toxin-induced cell death NEUROPROTECTIVE EFFECTS OF VITAMIN D VIA THE MODULATION OF NLRP3 INFLAMMASOME ACTIVATION Why does COVID-19 disproportionately affect the elderly? Effect of vitamin D on ACE2 and vitamin D receptor expression in rats with LPS-induced acute lung injury Serum angiotensin-converting enzyme in healthy and sarcoidotic children: comparison with the reference interval for adults Age-and gender-related difference of ACE2 expression in rat lung Gender differences in patients with COVID-19: Focus on severity and mortality The science behind "man flu Sex differences in renal angiotensin converting enzyme 2 (ACE2) activity are 17β-oestradiol-dependent and sex chromosome-independent The molecular basis of gender variations in mortality rates associated with the novel coronavirus (COVID-19) outbreak Vitamin D deficiency in elderly: Risk factors and drugs impact on vitamin D status Vitamin D in autoimmune rheumatic diseases: a view inside gender differences Vitamin D: autoimmunity and gender COVID-19 pandemic: is a gender-defined dosage effect responsible for the high mortality rate among males? High prevalence of vitamin D insufficiency in black and white pregnant women residing in the northern United States and their neonates Association of Vitamin D deficiency with breast cancer in African-American and Hispanic women in south Los Angeles [abstract Philadelphia (PA): AACR Vitamin D and cancer mini-symposium: the risk of additional vitamin D Why COVID-19 May Be Disproportionately Killing African Americans: Black Overrepresentation among COVID-19 Mortality Increases with Lower Irradiance, Where Ethnicity Is More Predictive of COVID-19 Infection and Mortality Than Median Income Non-black Latinos Spend More Time Outdoors and Have Lower Available at SSRN The Effects of Vitamin D Supplementation on Withdrawal Symptoms and the Expression of Inflammatory Cytokines and Insulin in Patients Under Methadone Maintenance Treatment: A Randomized, Double-Blind, Placebo-Controlled Trial Induction of interleukin-1 beta (IL-1β) is a critical component of lung inflammation during influenza A (H1N1) virus infection We are grateful to Prof. Ivan Dmochowski of the University of Pennsylvania and Prof. Armin Schwartzman of UCSD for stimulating discussions, and to Mathworks Inc. for financial support. No competing interests to declare. All data are available at https://tinyurl.com/vitamind4covid19. This paper evolved from earlier work published March 23 rd 2020 onwards here, here and here.