Hostname: page-component-89b8bd64d-ktprf Total loading time: 0 Render date: 2026-05-07T00:09:59.347Z Has data issue: false hasContentIssue false

COVID-19 host genetics and ABO blood group susceptibility

Published online by Cambridge University Press:  10 January 2023

David Ellinghaus*
Affiliation:
Institute of Clinical Molecular Biology, Kiel University and University Medical Center Schleswig-Holstein, Kiel, Germany
*
Author for correspondence: David Ellinghaus, Email: d.ellinghaus@ikmb.uni-kiel.de
Rights & Permissions [Opens in a new window]

Abstract

Twenty-five susceptibility loci for SARS-CoV-2 infection and/or COVID-19 disease severity have been identified in the human genome by genome-wide association studies, and the most frequently replicated genetic findings for susceptibility are genetic variants at the ABO gene locus on chromosome 9q34.2, which is supported by the association between ABO blood group distribution and COVID-19. The ABO blood group effect appears to influence a variety of disease conditions and pathophysiological mechanisms associated with COVID-19. Transmission models for SARS-CoV-2 combined with observational public health and genome-wide data from patients and controls, as well as receptor binding experiments in cell lines and human samples, indicate that there may be a reduction or slowing of infection events by up to 60% in certain ABO blood group constellations of index and contact person in the early phase of a SARS-CoV-2 outbreak. The strength of the ABO blood group effect on reducing infection rates further depends on the distribution of the ABO blood groups in the respective population and the proportion of blood group O in that population. To understand in detail the effect of ABO blood groups on COVID-19, further studies are needed in relation to different demographic characteristics, but also in relation to recent data on reinfection with new viral variants and in the context of the human microbiome.

Topics structure

Information

Type
Review
Creative Commons
Creative Common License - CCCreative Common License - BYCreative Common License - NC
This is an Open Access article, distributed under the terms of the Creative Commons Attribution-NonCommercial licence (http://creativecommons.org/licenses/by-nc/4.0), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original article is properly cited. The written permission of Cambridge University Press must be obtained prior to any commercial use.
Copyright
© The Author(s), 2023. Published by Cambridge University Press
Figure 0

Table 1. Genome-wide significant (P < 5 × 10−8) susceptibility loci for SARS-CoV-2 infection and/or COVID-19 disease severity identified in large-scale (hypothesis-free) genome-wide analyses to date (as of August 2022) (Ellinghaus et al., 2020; COVID-19 Host Genetics Initiative, 2021; Pairo-Castineira et al., 2021; Shelton et al., 2021; Wu et al., 2021; COVID-19 Host Genetics Initiative, 2022; Cruz et al., 2022; Degenhardt et al., 2022; Horowitz et al., 2022; Namkoong et al., 2022; Roberts et al., 2022)

Figure 1

Table 2. Genome-wide significant associations between genetic variants at the ABO gene locus and SARS-CoV-2 infection and/or COVID-19 disease severity (including other important phenotypic associations for the variants)

Figure 2

Figure 1. Summary of results from meta-analysis association studies at the 9q34.2 locus (ABO) conducted by the COVID-19 Host Genetics Initiative (HGI). Meta-analyses of association data show an association of the 9q34.2 locus (ABO) with (i) critical severity of illness, (ii) hospitalization and (iii) infection, as described in COVID-19 Host Genetics Initiative (2022). Upper Manhattan plot: association results for 8,779 critically ill COVID-19 patients versus 1,001,875 population controls. Middle Manhattan plot: association results for 24,274 hospitalized COVID-19 patients versus 2,061,529 population controls. Lower Manhattan plot: association results for 112,612 SARS-CoV-2 infected individuals versus 2,474,079 population controls. X-axis: chromosome positions and gene annotations on human genome build hg38. Y-axis: meta-analysis association p-values (−log10p) of genetic markers. Plots were generated with the COVID-19 Host Genetics Initiative Browser (https://app.covid19hg.org; release 6).

Figure 3

Table 3. Studies of mechanistic and pathophysiological hypotheses of ABO blood group effects as well as clinical findings from COVID-19 patient subgroup studies (not exhaustive) suggest an association between ABO blood groups and SARS-CoV-2 infection and COVID-19 disease severity

Figure 4

Figure 2. Two predominant hypotheses of possible mechanisms involving ABO blood group-related antigens: (A) The ABO-compatibility-dependence model (or ABO-interference) and (B) the ABO-dependent intrinsic model. The ABO-compatibility dependence model was recently modeled by Ellis (2021) under different assumptions and compared with observational healthcare data (Zietz et al., 2020; Zhao et al., 2021) and GWAS data from the Severe COVID-19 GWAS Group (Ellinghaus et al., 2020). Both models have been further evaluated by Boukhari et al. (2021) in a French study population of 666 individuals (333 index persons and their spouses) of known ABO blood type with a high risk of SARS-CoV-2 transmission (hospital employees) as well as receptor-binding domain (RBD) protein binding experiments in cell lines and saliva samples from individuals of known ABO and secretor phenotypes. For the ABO-compatibility-dependence model, ρ represents the relative probability of virus transmission between an infected index person and an ABO-incompatible contact (impeded transmission; pairs denoted with “I”) and was estimated to be 40% on average (between 20 and 55% depending on ABO blood group frequencies and relative risk ratios in different countries) by Ellis (2021). Boukhari et al. estimated a decrease of 41% in ABO-incompatible pairs. The ABO-dependent intrinsic hypothesis remains controversial because of conflicting study results (Boukhari et al., 2021; Wu et al., 2021). n.a., not available. Figure based on Boukhari et al. (2021) and extended.

Author comment: COVID-19 host genetics and ABO blood group susceptibility — R0/PR1

Comments

Dear Editors of Cambridge Prisms: Precision Medicine,

Thank you very much for inviting me to write a review article on "COVID-19 host genetics and ABO blood group susceptibility". Please find attached the review article you requested. I look forward to your response and I am happy to answer any further questions you may have.

Yours sincerely,

David Ellinghaus

Review: COVID-19 host genetics and ABO blood group susceptibility — R0/PR2

Conflict of interest statement

Reviewer declares none.

Comments

Comments to Author: Dr. Ellinghaus, one of the early leaders of COVID-19 genetics research, provides a timely look-back review on the evidence for the role genetics in the infection and clinical sequelae, with particular focus on the ABO blood group antigen variability and its roles.

1) One point the author does not clearly tackle is the extent of clinical applicability. It is difficult to address with surety but given the journal is focused on Precision Medicine it seems relevant to try. It is addressed a bit in the description of modest effect sizes (though larger than the other GWAS variants and more widely replicated). There is I think the point made at various stages that if transmission is widespread this may overwhelm any moderate protective effects. In the clinical context comorbid risk factors may also play a much greater role in triage and risk. Thus, perhaps the statement should be made that in the face of these factors there is not strong evidence at this time for a need to apply genetic risk factors in the clinic for SARS-Cov2 at this time? (translation of SNP genetic tests into the clinic has a spotty record even for relatively strong effects or PRS) Perhaps the point is more what the genetics can teach us about the biology, etiology and potential future treatment strategies?

2) Since the majority of the focus of the review is on the role of ABO blood group antigens, I believe it is worthwhile to spend a bit more space reviewing the longer history of ABO antigens and respiratory viruses, as the emergence of this relationship is not new to SARS-Cov2. There may be some scientific parables here that are useful to consider for future viruses that will evolve and face humanity. The evidence was long for the role of ABO in viruses but in some ways the research may have gone quiet in the area for more than 10 years until SARS-Cov2 emerged with its linkage. I suggest you split out a separate paragraph of section covering this history. Since you have only 1 Figure you could also consider including a Figure with a Historical timeline of ABO in relation to viruses (of course focusing most of the timeline on recent developments). Personally, my opinion is that a review paper should provide some longer focus on the history of the science, and sometimes teach us how and when the breadcrumbs presented themselves in the past. Some suggestions

a. You cited Guillon et al. 2008 (PMID18818423), an important work. However, that work was predicated in part on a study of the Hospital Ward in Hong Kong where SARS-Cov1 broke out in March 2003 and was recognized. Cheng et al. 2005 (JAMA, PMID 15784866) studied 45 staff members on that ward for their Cov1 antibody titer and ABO antigen group (based on prior observations of difference in susceptibility to Norwalk virus and differences in H. pylori infection rates). While the sample was small (34 exposed with confirmed antigen and symptoms; 11 non-infected/seronegative after 2 months); they found a significant and large protective effect of Group O antigen (OR 0.18). I believe it is worth citing and discussing that foundational clinical observation

b. More broadly there is additional history of ABO and infections that can be discussed. You do define/cite this briefly in relation to malaria and review papers but it could be worth a brief further exploration. Due to R.A. Fisher defining the strong heritability and it being one of the early measurable polymorphic factors, and its pleiotropic effect in a variety of conditions including thrombosis, there were many studies of ABO even going back to the 1950s and 1960s. McDonald & Zuckerman published in 1962 (BMJ; PMID 20789459) what was really quite a massive study for the time on blood group antigens in 1,685 patients in the Royal Air Force with viral infections and 41,708 controls for allele distribution. They report protective effects of O antigen for Adenovirus but increased risk for Influenza A2. In 1967 Tyrrell, Peto & King showed that arrivals of an isolate island population via Cape Town into Britain has much greater rates of influenza A among A/B/AB groups than the O group (J.Hyg(Lond); PMID 4293359). At the same time there are some seemingly conflicted articles in the 1960s,70s, and 80s mostly centered around influenza virus. Mackenzi & Fimmerl (J Hyg.(Camb) 1978 PMID 621379) discuss some of the conflicting works, and additional data suggesting some B/O differences.

c. The most comprehensive review on ABO and bacterial, viral and fungal infections in earlier periods that lays out the inconsistencies and controversies but argues for the overall importance is Berger, Young and Edbery (Eur J Clin Microbiol Infect Dis 1989 PMID 2506033).

d. Perhaps some of this history and the contradictory results is relevant to highlight and discuss in terms of the remaining questions and controversies of ABO and SARS-Cov2 as you well highlight, and the possible importance of study design to the prior study conclusions? As you highlight the French couple study (compatible/incompatible design) provides an example of an important study design to disentangle some controversies.

3) In terms of the early COVID-19 pandemic and genetic findings I think it is important again to place things correctly in historical perspective and give the appropriate credits due:

a. The first report of ABO group involvement appeared on medRxiv on March 9, 2020 from Fan et al. that reported much lower O group among n=101 fatalities in Wuhan China. The published paper with similar people (Fan is now a middle author) Lin Zhang et al. (Sep 3 2020; Cambridge Press: Epidemiol Infect) ended up being a retrospective on n=134 critically ill patients showing the similar O group protection for survival.

b. The work by Zhao et al. is a different but still quite important large retrospective work on 2,173 patients published on medRxiv later (March 11, 2020) and in the journal online Aug 4, 2020 (and in print 2021).

c. I believe both works are important and should be cited/differentiated with Fan/Zhang being just a few days earlier on preprint.

4) In re: to Table 1, I wonder if the APOE gene region/variants should be included or at least discussed somewhere? It was the first GWAS-significant signal observed for COVID-19 positivity (Kuo et al. 2020 PMID 32623451) in a sample with n>1,000 – whether the UKBB is considered 1 population may be an open question (to meeting your stated Table requirements). However, as Thibord et al. (2022 PMID 35224516) showed the signal of susceptibility in UKBB degraded over time as the pandemic evolved and repeated GWAS are run in UKBB with increasingly younger age cases. Thibord et al. showed a GWAS-significant signal with COVID mortality with APOE, and a significant age-interaction, and persistence of effects after adjusting for Alzheimer’s disease and CVD factors, though no association with severity was observed in the COVID-19 hgi analysis at that time. Overall it presents a puzzling picture and could be due to either confounding factors, or the need to precisely define the clinical population (something likely lost in the broader analyses such as COVID-hgi). I might dismiss it as confounding or a false positive, but recent work (Zhang et al. 2022 Signal Transduct Target Ther PMID 35915083) demonstrates a functional interaction of APOE with ACE2 in SARS Cov-2 viral entry.

5) In Table 2 (given the Title) some associations are not specifically within the context of COVID-19. That may be OK but perhaps it is important to be clearer about this/ For instance the monocyte counts, VTE, F8, vWF etc. are not in COVID patients and mostly predate the pandemic. Perhaps you should rename the Table to include other important associations for the same variants? The Table could be slightly updated for the new VTE GWAS (Thibord et al. 2022 Circulation PMID 36154123) which essentially combined references 66 and 67 with further replication and found strongest association with rs505922 (P<1.55E-1043), though as pointed out correlated with rs687289.

6) The paper is very well written and structured. Minor suggestions/typos:

a. P.8 “Therefore, future model” -> “Therefore, future models”

b. P.8 “Omikron” -> is it the British spelling? I think globally it has more often been spelled as Omicron.

Review: COVID-19 host genetics and ABO blood group susceptibility — R0/PR3

Conflict of interest statement

Reviewer declares none.

Comments

Comments to Author: This is a fine review summarizing existing studies and results for the association between the ABO locus and SARS-CoV-2. Consider rewording subsection titles (i.e. "The ABO blood group system in a nutshell" may appear informal). It would be beneficial to add a figure displaying Manhattan plots across various GWAS that implicated ABO as a candidate locus for infection and severity, respectively. The manuscript showed adequate attention to detail and consideration of pitfalls from reported associations, while also acknowledging consistency and repeatability in the ABO locus across studies.

Recommendation: COVID-19 host genetics and ABO blood group susceptibility — R0/PR4

Comments

Comments to Author: Dear Dr. Ellinghaus,

The manuscript was fully evaluated by two independent peer reviewers. Based on the advice received, I feel that your manuscript could be reconsidered for publication should you be prepared to incorporate minor revisions suggested by reviewers. We ask you to provide a detailed point-by-point response letter to concerns detailed by reviewers, highlighting your responses to the review comments and a description of the changes you have made to the manuscript.

Decision: COVID-19 host genetics and ABO blood group susceptibility — R0/PR5

Comments

No accompanying comment.

Author comment: COVID-19 host genetics and ABO blood group susceptibility — R1/PR6

Comments

Dear Prof. Dominiczak,

dear Editors of Cambridge Prisms: Precision Medicine,

please find enclosed the revised manuscript (with all changes in red) and a further point-by-point response that addresses the reviewers concerns.

I look forward to your decision and I am at your disposal for any further questions.

Yours sincerely,

David Ellinghaus

Review: COVID-19 host genetics and ABO blood group susceptibility — R1/PR7

Conflict of interest statement

Reviewer declares none.

Comments

Comments to Author: Excellent work in making all appropriate changes.

Review: COVID-19 host genetics and ABO blood group susceptibility — R1/PR8

Conflict of interest statement

Reviewer declares none.

Comments

Comments to Author: Not all comments were addressed in the revision, and a point-by-point revision response would have been preferred as a courtesy to the reviewers when revisions were not made to understand the thinking. However, the most salient revisions were made and I don't see the need to belabor minor points.

Recommendation: COVID-19 host genetics and ABO blood group susceptibility — R1/PR9

Comments

Comments to Author: Both independent reviewers have affirmed that the authors have provided sufficient responses to either all of their peer review comments or the most salient comments.

Decision: COVID-19 host genetics and ABO blood group susceptibility — R1/PR10

Comments

No accompanying comment.