Association between ABO blood group/genotype and COVID-19 in a Japanese population

An association between coronavirus disease 2019 (COVID-19) and the ABO blood group has been reported. However, such an association has not been studied in the Japanese population on a large scale. Little is known about the association between COVID-19 and ABO genotype. This study investigated the association between COVID-19 and ABO blood group/genotype in a large Japanese population. All Japanese patients diagnosed with COVID-19 were recruited through the Japan COVID-19 Task Force between February 2020 and October 2021. We conducted a retrospective cohort study involving 1790 Japanese COVID-19 patients whose DNA was used for a genome-wide association study. We compared the ABO blood group/genotype in a healthy population (n = 611, control) and COVID-19 patients and then analyzed their associations and clinical outcomes. Blood group A was significantly more prevalent (41.6% vs. 36.8%; P = 0.038), and group O was significantly less prevalent (26.2% vs. 30.8%; P = 0.028) in the COVID-19 group than in the control group. Moreover, genotype OO was significantly less common in the COVID-19 group. Furthermore, blood group AB was identified as an independent risk factor for most severe diseases compared with blood group O [aOR (95% CI) = 1.84 (1.00–3.37)]. In ABO genotype analysis, only genotype AB was an independent risk factor for most severe diseases compared with genotype OO. Blood group O is protective, whereas group A is associated with the risk of infection. Moreover, blood group AB is associated with the risk of the “most” severe disease.


Background
The ABO blood group system has previously been shown to be associated with host susceptibility to various infectious diseases [1,2]. For example, in severe acute respiratory syndrome, a clear correlation between disease transmission and the ABO blood group has been reported, with the finding that group O was less likely to be infected than group non-O [3]. Similarly, since the COVID-19 pandemic, many studies have reported an association between COVID-19 and the ABO blood group [4][5][6][7][8]; however, they showed heterogeneous results. In other infectious diseases, the effect of the ABO blood group varied depending on ethnicity. For example, in the norovirus outbreak, the effect of the ABO blood group was influenced by the outbreak area and ethnicity because of different host receptors in each endemic strain [9,10]. Therefore, we examined whether the association between COVID-19 and the ABO blood group in the Japanese population differs from that in COVID-19 patients from other countries. The phenotypic prevalence of the ABO blood group in the Japanese population has been reported to be 29.25% O, 38.65% A, 22.15% B, and 9.95% AB. However, the prevalence and association between blood groups and severity of COVID- 19 have not been studied on a large scale in the Japanese population. Additionally, recent studies on other diseases have reported the effect of the ABO Extended author information available on the last page of the article genotype on disease transmission. For example, it has been reported that there was a trend toward an increased risk of developing gastric cancer with the addition of the A allele and a decreased risk with the addition of the B allele [11]. Only one study has reported the association between the ABO genotype and COVID-19 [12]; however, the association has not been clarified.
We have conducted a nationwide multi-center consortium to overcome the COVID-19 pandemic in Japan [13,14]. This study aimed to investigate the clinical and genetic characteristics of the association between COVID-19 and the ABO blood group/genotype, and to the best of our knowledge, this is the first such large-scale study in Japan.

Study design and settings
All COVID-19 cases were recruited through the Japan COVID-19 Task Force in this retrospective cohort study [13]. From February 2020 to October 2021, data from consecutive patients aged ≥ 18 years who were diagnosed with COVID-19 based on PCR or antigen test and who agreed to participate in the study were entered into an electronic case record form by the attending physicians at the affiliated research institutions. We excluded patients who met any of the following exclusion criteria: (i) non-Japanese patients, (ii) patients with incomplete medical records, for example, insufficient data to extract critical outcome information, and (iii) patients for whom genome-wide association study (GWAS) genotyping data were not available (Fig. 1). All patients involved in this study provided written or oral consent, and the study design was approved by the ethics committees of Keio University School of Medicine (20200061) and the affiliated research institutions.

Data collection
The information related to age, sex, height, weight, clinical symptoms and signs, laboratory findings on admission, comorbidities, and clinical outcomes (intensive care unit (ICU) admission, invasive mechanical ventilation (IMV) usage, and survival status) were extracted from the electronic case record form manually. All laboratory tests were performed according to the clinical needs of the patient. Clinical symptoms and signs observed at the time of referral and admission and during hospitalization were recorded. Laboratory and radiographic results were collected within 48 h of the initial visit or admission. A team of respiratory clinicians reviewed the collected data; if core data were missing, the clinician was contacted to collect the data. Disease severity was defined as most severe for patients requiring support using high-flow oxygen devices, IMV, or extracorporeal membrane oxygenation, or incidence of death; severe for patients requiring support using low-flow oxygen devices; mild for symptomatic patients not requiring oxygen support; and asymptomatic for asymptomatic patients not requiring oxygen support [14,15].

Genotype data of COVID-19 patients
We performed GWAS genotyping of 2520 COVID-19 cases using the Infinium Asian Screening Array (Illumina, CA, USA). We applied stringent quality control filters to the samples and variants, as described in a previous study [13]. A total of 2433 COVID-19 cases passed sample quality control. Subsequently, we performed genome-wide genotype imputation; the details are described in a previous study [13]. Out of 2433 COVID-19 cases, 1790 cases had clinical information. The GWAS data from 611 healthy subjects as reference populations were collected by the Department of Statistical Genetics at Osaka University Graduate School of Medicine and the Japan Biological Informatics Consortium and analyzed.

Estimation of ABO genotype
We estimated the ABO genotype of GWAS subjects based on five coding variants of the ABO gene (rs8176747, rs8176746, rs8176743, rs7853989, and rs8176719) [16,17]. We phased the haplotypes of these five variants based on the best-guess genotype obtained by genome-wide imputation and estimated the ABO blood type as described in a previous study [18].

Statistical analysis
For baseline variables, we obtained summary statistics using frequency and proportion for categorical data and mean and standard error for continuous variables. Data were compared among the four groups classified by the ABO blood group system using the chi-square test and analysis of variance as appropriate.
The chi-square test assessed differences in the blood group distribution between COVID-19 patients and Japanese patients without COVID-19.
To assess the association between the ABO blood group/genotype and the clinical outcomes (death, disease severity, IMV usage, and ICU treatment), we performed multivariate logistic regression analyses after adjusting for characteristics reported as predictors of Data show mean or percentage. Data were analyzed by chi-square test or analysis of variance where appropriate. Alb, albumin; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BNP, brain natriuretic peptide; BUN, blood urea nitrogen; Cre, creatinine; CRP, C-reactive protein; CT, computed tomography; CXR, chest X-ray; GGO, ground-glass opacity; Hb, hemoglobin; KL-6, Krebs-von-den-Lungen-6; LDH, lactate dehydrogenase; SpO 2 , saturation of percutaneous oxygen; WBC, white blood cell  [19][20][21][22]. Subjects with the ABO blood group O or genotype OO were considered at baseline risk. Data are presented as adjusted odds ratio (aOR) with 95% confidence interval (CI). Statistical significance was set at p < 0.05. All data were analyzed using the JMP 16 program (SAS Institute Japan Ltd., Tokyo, Japan).

Frequency of ABO blood group or genotype according to COVID-19 infection
The baseline characteristics and clinical features of the patients are summarized in Table 1. Among 1790 COVID-19 patients, 614 patients (34.3%) were women, the mean BMI was 24.4, and the mean age was 56.7 years. The distribution of COVID-19 and control groups according to the ABO blood group and genotype is shown in Table 2. The COVID-19 group had a higher proportion of males than the control group, and the mean age was older. Regarding the ABO blood group, group A was significantly more prevalent (41.6% vs. 36.8%), and group O was significantly less prevalent (26.2% vs. 30.8%) in the COVID-19 group than in the reference population. There were no significant differences in the prevalence of blood groups B and AB between the two groups. Regarding the ABO genotype, type OO was significantly less common (26.2% vs. 30.8%) in the COVID-19 group than in the reference population. There were no significant differences in the presence of other ABO genotypes between the two groups.

Comparison of clinical characteristics among ABO blood groups
The baseline and clinical characteristics of the patients classified for each ABO blood group are summarized in Tables 3  and 4, respectively. There were no significant differences in demographics, comorbidities, vital signs, laboratory findings, and radiological findings except the clinical sign of shortness of breath among the four groups classified based on the ABO blood group system. Table 5 and Supplementary Table 1 show the multiple logistic regression analysis results of parameters that included the ABO blood groups and previously reported prognostic factors [19][20][21][22]. Compared with individuals with blood group O, after adjustment for age, sex, BMI, and presence of comorbidities, blood group AB was the independent risk factor for most severe disease [aOR (95% CI) = 1.84 (1.00-3.38)] (Table 5). A similar analysis was performed, and no significant differences were observed concerning the incidence of death, IMV usage, and ICU treatment (Supplementary Table 1) between the groups.  Table 2 show multiple logistic regression analysis results of parameters that included the ABO genotypes and previously reported prognostic factors [19][20][21][22]. Compared with individuals with genotype OO, after adjustment for age, sex, BMI, and presence of comorbidities, genotype AB was the independent risk factor for most severe disease [aOR (95% CI) = 1.84 (1.00-3.37)]. A similar analysis was performed, and no significant differences were found concerning the incidence of death, IMV usage, and ICU treatment between the groups. Moreover, there was no significant difference in the risk for the occurrence of the most severe disease in COVID-19 patients by adding the O allele, A allele, or B allele (Fig. 3).

Discussion
To our knowledge, this is the first large-scale study in Japan to present an association between COVID-19 and ABO blood groups/genotypes. Our study provided two novel findings of clinical relevance. First, we identified blood group AB as an independent risk factor for clinical outcomes in Japanese COVID-19 patients. Second, this study is the first to analyze the association between the ABO genotype and outcomes such as the incidence of death, IMV usage, ICU treatment, and risk for the occurrence of most severe disease in COVID-19 patients. It has been reported that differences in host genomes among populations may affect clinical outcomes, and Asians, including Japanese, may have a lower mortality risk [23,24]. Thus, because of possible racial and population differences in clinical outcomes in COVID-19 patients, the association between COVID-19 and ABO blood groups, which has already been reported by several largescale studies in other countries [4][5][6][7][8], was investigated in this large-scale study involving Japanese population.    [25]. Consistent with our findings, previous studies in other countries have reported that blood group A has a higher risk of infection and group O has a lower risk of infection [4,26,27]. The COVID-19 group had a higher proportion of males and an older age than the control group, which is consistent with previous reports that showed that male gender and old age are risk factors for COVID-19 severity [19][20][21][22]. There is a concern that the difference in age between the COVID-19 and control groups may have affected the distribution of the ABO blood group/genotype; however, this effect is expected to be small, considering that the frequencies of ABO blood group and alleles in this study and those in the 1978 report [25] were compatible. However, the results of the study on the association between clinical outcomes and the ABO blood group in COVID-19 patients differed from that of previous studies in other countries, which concluded that group O is associated with a reduced risk of death or severe illness [6], and group A is associated with an increased risk of death [28,29], respiratory failure [30], and pneumonia [31]. In this study, we showed that blood group AB is an independent Fig. 2 Results of the multiple logistic regression analysis. Forest plot of the multiple logistic regression analysis using the ABO genotype and COVID-19 prognostic factors to assess the associa-tion between the ABO genotype and most severe disease. CI, confidence interval; OR, odds ratio Fig. 3 Multivariable-adjusted odds ratio for most severe disease in COVID-19. Forest plot of multivariable-adjusted odds ratio for most severe disease in COVID-19 patients with the addition of the ABO allele. CI, confidence interval; OR, odds ratio risk factor for the occurrence of most severe disease that requires support using high-flow oxygen devices, IMV, extracorporeal membrane oxygenation, or leads to death, despite adjustment for COVID-19-related risk factors, such as age, gender, and comorbidities. Compared with previous reports in Europe [7,32] and the USA [5,33], the present study enrolled a larger number of patients with blood group AB; therefore, our study is reliable in identifying group AB as an independent risk factor for most severe disease. The GATC haplotype of the four polymorphisms of the ABO gene (rs8176746, rs8176740, rs495828, and rs12683493), which is more common in group non-O patients, is positively associated with angiotensin-converting enzyme (ACE) activity, suggesting that patients with blood group O have lower ACE levels and benefit from ACE2 [34]. On the contrary, the A allele is associated with an increased risk of developing cardiovascular disease; the A antigen promotes leukocyte binding to P-selectin and intercellular cell adhesion molecule 1 on the vessel wall, contributing to increased inflammation [34]. It has been suggested that the anti-A antibodies from group O COVID-19 patients are more protective than those from group B COVID-19 patients [35]. Group AB has been reported to be more susceptible to diverse infectious diseases [36] and associated with a higher risk for stroke [37] and thrombosis [38]. In group AB patients, the same mechanisms contributing to the risk of infection and thrombosis may also contribute to the clinical outcomes of COVID-19; however, further studies are warranted.
Previous reports on the association between ABO blood genotype and COVID-19 have analyzed the risk of infection, respiratory symptoms, and need for hospitalization [12]. Few studies have reported an increased risk of developing gastric cancer with the A allele, a decreased risk with the B allele [11], and an increased risk of developing pancreatic cancer with both A and B alleles [39], with the analysis of the ABO allele being significant. In this study, we analyzed the association between the ABO blood genotype and clinical outcomes. The overall analysis showed that only genotype AB was an independent risk factor for most severe diseases compared with genotype OO, similar to the analysis for the ABO blood group. To examine the differences between the AA and AO genotypes in blood group A and the BB and BO genotypes in blood group B, we also analyzed the risk for the occurrence of the most severe disease based on the number of ABO alleles. It has been suggested that the O allele may be protective and the B allele a risk factor for most severe diseases in COVID-19 patients, but no significant differences were reported in the analysis of other diseases.
Our study has two main limitations. First, the number of patients in the analysis of the ABO genotype is insufficient. In this study, there were only 55 (3.1%) study participants with genotype BB. The prevalence is equivalent to that in the previous report of Fujita et al. [25], obtained based on the Hardy-Weinberg equilibrium. However, the number of participants with the BB genotype might be insufficient, and no significant differences between this genotype and COVID-19 severity were observed, even though our results suggest that this genotype may be a risk factor for most severe diseases. Further addition of cases may provide new insights into the association between the ABO genotype and COVID-19. Second, this study included only Japanese patients; we could not examine the impact of racial differences on the association between the ABO blood group/ genotype and COVID-19. It is widely known that the distribution of blood groups and genotypes by race is different. The association between blood groups and severe acute respiratory syndrome coronavirus 2 infection and clinical outcomes have been reported differently in different countries [40], and the possibility that racial factors contribute to this difference cannot be ruled out.

Conclusions
Our study revealed a clinical association between the ABO blood group/genotype and COVID-19 in a Japanese population. Blood group AB was significantly associated with disease severity in COVID-19 patients in the Japanese population. Further studies to elucidate the mechanisms underlying the association are warranted to discuss the clinical impact of the ABO blood group/genotype in COVID-19 patients.
Data availability All data generated or analyzed during this study are included in this published article and its supplementary information files.

Declarations
Ethics approval and consent to participate All patients involved in this study provided written or oral consent, and the study design was approved by the ethics committees of Keio University School of Medicine (20200061) and the affiliated research institutions.

Consent for publication Not applicable
Competing interests The authors declare no competing interests.