Lack of association between ABO blood groups and susceptibility to SARS‐CoV‐2 infection

Abstract

Studies in China, the USA and Turkey, summarized by Focosi [1], described an association between risk for SARS-CoV-2 infection and ABO blood groups, with the results converging to a higher rate of infection among type A and lower among type O subjects. These data are reinforced by an elegant study employing genome-wide association, which identified a region on chromosome 9 corresponding to the ABO locus, with frequency misbalance between severe COVID-19 patients and healthy blood donors from Italy and Spain [2]. The biological mechanism behind the increased susceptibility of type A and resistance by type O individuals to SARS-CoV-2 infection has been related to anti-A isohaemagglutinin, which may prevent the binding of the viral spike protein to the cellular receptor ACE2 [3]. We sought to evaluate the relationship between blood group types and SARS-CoV-2 laboratory results from our database. Up to 22 June 2020, 196 897 and 256 471 real-time PCR (RT-PCR) and serological tests respectively were performed at Dasa, the largest clinical pathology laboratory in Brazil. Among patients submitted to SARS-CoV-2 testing, we identified 6457 who had a concomitant ABO blood group typing result, being 4353 tested by RT-PCR and 2275 for COVID-19 antibodies, while 171 did both (Table 1). Most patients submitted to RT-PCR were hospitalized, as testing due to reagent shortage was prioritized to them, while serology did not require a prescription and was performed by asymptomatic individuals and patients presenting a broader range of clinical manifestations. However, results did not differ when analysing in separate RT-PCR and serology patients (data not shown). We also retrieved historical ABO blood group typing from our records, representing 1 813 237 patients. The program 'R' (R Foundation for Statistical Computing, Vienna, Austria) was used to perform Pearson's chi-squared test. Our data failed to reproduce the skewed frequency of ABO blood group types reported by the studies above. Though we do find a trend for a higher frequency of A and lower of O type comparing SARS-CoV-2 patients with the general population attended by the laboratory, these differences are quite small and did not reach statistical significance. Essentially, 30% of the patients in our cohort had one or more laboratory markers indicating previous or current SARS-CoV-2 infection, irrespective of the blood group (Table 2). The absence of a relationship between ABO blood type and susceptibility to SARS-CoV-2 infection was also reported in another US study [4], which suggested that ethnicity may have biased previous analysis. Noticeably, a higher frequency of type A was found among COVID-19 patients in non-Hispanic White subjects but not in Black or Hispanic subjects [5], who are certainly more related to the genetic background of Brazilians. In conclusion, ABO blood group types do not seem to significantly impact on the risk for SARS-CoV-2 infection among a representative population from Brazil. The authors declare no conflict of interests. None. Data sharing requests should be sent to José Eduardo Levi ([email protected]).

Topics & Concepts