key: cord-0694887-cfzeblcv authors: Sun, Y.; Zhou, J.; Ye, K. title: White blood cells and severe COVID-19: a Mendelian randomization study date: 2020-10-18 journal: nan DOI: 10.1101/2020.10.14.20212993 sha: d7443f7c9a737d01c60bee3e4fb263af038c4704 doc_id: 694887 cord_uid: cfzeblcv Background: The pandemic of coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has rapidly emerged to seriously threaten public health. We aimed to investigate whether white blood cell traits have potential causal effects on severe COVID-19 using Mendelian randomization (MR). Methods: To evaluate the causal associations between various white blood cell traits and severe COVID-19, we conducted a two-sample MR analysis with summary statistics from recent large genome-wide association studies. Results: Our MR results indicated potential causal associations of white blood cell count, myeloid white blood cell count, and granulocyte count with severe COVID-19, with odds ratios (OR) of 0.84 (95% CI: 0.72-0.98), 0.81 (95% CI: 0.70-0.94), and 0.84 (95% CI: 0.71-0.99), respectively. Increasing eosinophil percentage of white blood cells was associated with a higher risk of severe COVID-19 (OR: 1.22, 95% CI: 1.03-1.45). Conclusions: Our results suggest the potential causal effects of lower white blood cell count, lower myeloid white blood cell count, lower granulocyte count, and higher eosinophil percentage of white blood cells on an increased risk of severe COVID-19. associated with the risk of developing severe COVID-19. Even if white blood cells are 67 measured before infection, they are influenced by many exogenous and endogenous 68 factors (e.g., age, gender, disease status, and medications), which will confound the 69 observational associations [19, 20] . No previous research has been able to interrogate 70 the causal role of white blood cells in severe 72 While traditional observational associations often suffer from confounding, reverse 73 causality, and various biases, a complementary approach, Mendelian randomization 74 (MR), utilizes genetic variants as instrumental variables to approximate the lifetime 75 status of exposure and enable causal inference in observational data. The random 76 allocation of the allele at conception and the natural directional effects of genetic 77 variants empower MR estimates to be less plagued by confounding and reverse 78 causality [21, 22] . In this study, we aimed to test the causal effects of 19 white blood cell 79 traits on severe COVID-19 by performing a two-sample MR analysis. 80 . CC-BY-ND 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 October 18, 2020 . . https://doi.org/10.1101 median (WM) [27] , and Mendelian randomization pleiotropy residual sum and outlier 104 (MR-PRESSO) test [28] . The MR-Egger estimates allowed directional or unbalanced 105 horizontal pleiotropic effects. The weighted median method provides robust causal 106 estimates even when up to 50 % SNPs are invalid genetic instruments [27] . The MR-107 PRESSO test was utilized to correct for the presence of specific horizontal pleiotropic 108 outlier variants via detected outlier removal [28] . All MR analyses were conducted in R 109 with the TwoSampleMR package [23] . 110 111 The COVID-19 Host Genetics Initiative: https://www.covid19hg.org/ 113 The IEU OpenGWAS database: https://gwas.mrcieu.ac.uk/ 114 . CC-BY-ND 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 October 18, 2020 . . https://doi.org/10.1101 Results 115 The counts of white blood cell, myeloid white blood cell, and granulocyte are 116 negatively associated with severe COVID-19 117 By applying a two-sample MR approach, we first investigated the causal associations of 118 the counts of white blood cell and its subpopulations with severe COVID-19. A relatively 119 large number of independent SNPs, ranging from 79 for basophil count to 185 for 120 monocyte count, were selected as genetic instruments for each blood cell count 121 (Supplementary Tables 2-11 Table 136 21). Taken together, we demonstrated that white blood cell count, myeloid white blood 137 . CC-BY-ND 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 October 18, 2020 . . https://doi.org/10.1101 cell count, and granulocyte count had consistent, negative effects on the risk of severe 138 The percentage of eosinophil in granulocyte is positively associated with severe 158 . CC-BY-ND 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 October 18, 2020. . . CC-BY-ND 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 October 18, 2020. . https://doi.org/10.1101/2020.10.14.20212993 doi: medRxiv preprint To our knowledge, this is the first MR study evaluating the causal roles of white blood 173 cell traits in severe COVID-19 risk. Overall, our results suggest potential causal 174 protective effects of increasing white blood cell count, myeloid white blood cell count, 175 and granulocyte count on severe COVID-19. Our novel findings also include that 176 genetically predicted higher eosinophil percentage in white blood cells or in 177 granulocytes was associated with a higher risk of severe 179 Previous observational studies have frequently pointed out increased white blood cell 180 count in the severe COVID-19 patients, when compared to healthy controls or mild 181 COVID-19 patients [1, 11, 12, 29] . However, there are also reports that normal or 182 decreased white blood cell count is more common in COVID-19 patients when 183 compared to the reference range or healty controls [10, [15] [16] [17] [30] [31] [32] . Our MR analysis 184 showed that lower white blood cell count, myeloid white blood cell count, and 185 granulocyte count may play a causal role in increasing the risk of severe The mechanism by which they contribute to severe COVID-19 remains unclear. Immune 187 system disorders have been suspected of playing roles in severe COVID-19 risk [33, 188 34] . The complete elucidation of the potential mechanism warrants further investigation. 189 190 Persistent eosinopenia after admission was associated with COVID-19 risk in previous 191 retrospective and prospective observational studies [12, 15, 31, 32, [35] [36] [37] . There are 192 also reports that eosinophil remains stable in severe COVID-19 patients [13, 29, 38, 39] . 193 A multiparametric flow cytometry analysis found that high eosinophil count was 194 . CC-BY-ND 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 October 18, 2020. . SNPs at this locus are associated with multiple blood cell traits, suggesting they may 210 have pleiotropic effects and are not suitable to be used as genetic instruments [53] . As 211 more COVID-19-associated SNPs are identified in the future, reverse MR analysis will 212 be valuable to understand the effect of COVID-19 on blood cell traits. 213 One assumption of MR is that the instrumental variable influences severe COVID-19 215 risk only through its effect on a specific white blood cell trait. The Cochran Q statistic did 216 not reveal heterogeneity among our genetic instruments, and the MR-Egger intercept 217 . CC-BY-ND 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 October 18, 2020. . test also indicated no presence of pleiotropic effects except those for basophil count. 218 Our MR study was performed with strong instrumental variables and adequate statistical 219 power, and we have conducted extensive sensitivity analyses. Still, we emphasize that 220 our results should be interpreted with caution and future studies are needed to elucidate 221 the mechanistic roles of white blood cell traits in severe . CC-BY-ND 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 October 18, 2020. . https://doi.org/10.1101/2020.10.14.20212993 doi: medRxiv preprint Our results suggest that lower white blood cell count, lower myeloid white blood cell 224 count, lower granulocyte count, and higher eosinophil percentage of white blood cells 225 are causally associated with an increased risk of severe . CC-BY-ND 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 October 18, 2020. . https://doi.org/10.1101/2020.10.14.20212993 doi: medRxiv preprint . CC-BY-ND 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 October 18, 2020. All data generated or analyzed during this study are included in this published article 248 and this supplementary information files. 249 250 The authors declare that they have no competing interests. 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 October 18, 2020. . https://doi.org/10.1101/2020.10.14.20212993 doi: medRxiv preprint the manuscript. All authors critically read, revised, and approved the final version of the 261 manuscript. 262 263 We would like to express our gratitude to all other Ye lab members for stimulating 265 discussions. 266 267 . CC-BY-ND 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 October 18, 2020. . CC-BY-ND 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 October 18, 2020. . Xie G, Ding . CC-BY-ND 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 October 18, 2020. . Zhou Y, Fu B, Zheng X, Wang D, Zhao C, qi Y, Sun R, Tian Z, Xu X, Wei H: 359 Immunity 2020, 52(6):910-941. 403 . CC-BY-ND 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 October 18, 2020. . Table 1 . CC-BY-ND 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 October 18, 2020 . . https://doi.org/10.1101 . CC-BY-ND 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 October 18, 2020. . CC-BY-ND 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 October 18, 2020. . genetic associations of severe COVID-19. The slopes of each line represent the causal 503 association using different MR methods. The light blue line represents the random-504 effects inverse variance-weighted estimate, the dark blue line represents the MR-Egger 505 estimate, and the green line represents the weighted median estimate. 506 . CC-BY-ND 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 October 18, 2020. . https://doi.org/10.1101/2020.10.14.20212993 doi: medRxiv preprint Tables Table 1. Causal associations between white blood cell traits and severe COVID-19 based on random-effects IVW MR estimation. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint 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 October 18, 2020. . https://doi.org/10.1101/2020.10.14.20212993 doi: medRxiv preprint Significant result Non-significant result . CC-BY-ND 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 October 18, 2020. . The epidemiology and pathogenesis of 404 coronavirus disease (COVID-19) outbreak Clinical Characteristics of 138 Hospitalized Patients With Coronavirus-Infected Pneumonia in Wuhan, China The clinical characteristics of pneumonia 409 patients coinfected with 2019 novel coronavirus and influenza virus in Wuhan, 410 China Hui 412 DSC et al: Clinical Characteristics of Coronavirus Disease 2019 in China Early Clinical and CT 415 Manifestations of Coronavirus Disease 2019 (COVID-19) Pneumonia Clinical characteristics and imaging manifestations of the 2019 novel 419 coronavirus disease (COVID-19):A multi-center study in Wenzhou city Xiong 429 Y et al: Clinical features and treatment of COVID-19 patients in northeast 430 Chongqing Reverse Causality in Cardiovascular Epidemiological 432 Research: More Common Than Imagined? Circulation Altered blood cell traits underlie a major 434 genetic locus of severe COVID-19