The ABO gene variant that creates blood type O by disabling its enzyme, why blood type matters for transfusion, and which disease links (clotting risk, malaria, cholera) are strong versus which (COVID-19, cancer risk) remain modest or contested.
Blood type is one of the oldest and most consequential pieces of genetic information in medicine, discovered by Karl Landsteiner at the turn of the 20th century once doctors realized that mixing blood from different people could be safe or fatal depending on which combination was used. The ABO system sorts people into four groups — A, B, AB and O — based on small sugar molecules called antigens that sit on the surface of red blood cells (and, in most people, in saliva and other secretions as well).
The gene responsible, called ABO, makes an enzyme that attaches a sugar onto a simpler molecule already present on red cells, called the H antigen. The A version of the enzyme attaches one sugar; the B version attaches a slightly different one; people with both versions are blood type AB. The O allele, however, carries a change — a single deleted DNA letter, cataloged as rs8176719 — that shifts the gene's reading frame partway through, so the resulting enzyme is garbled and nonfunctional. People with two O alleles make no functional enzyme at all, so their red cells display only the plain, unmodified H antigen. That single deletion is the molecular basis of blood type O, the most common blood type in much of the world.
Everyone's immune system is naturally primed with antibodies against whichever ABO antigens they don't carry themselves — a type A person naturally carries anti-B antibodies, a type O person carries both anti-A and anti-B, and so on. Transfusing the wrong type triggers an immediate, potentially life-threatening immune reaction as those antibodies attack the donor's red cells. This is why blood type O is often called the "universal donor" for red cells (its cells carry no A or B antigen for a recipient's antibodies to attack) and why type AB is the "universal recipient" (it carries no anti-A or anti-B antibodies of its own). These rules are foundational to blood banking and are separate from Rh factor (the other major blood group), which is determined by an entirely different gene.
Beyond transfusion, ABO blood type has some of the best-replicated disease associations in human genetics. People with a non-O blood type (A, B, or AB) have measurably higher levels of two clotting factors, von Willebrand factor and factor VIII, and a large meta-analysis pooling data across dozens of studies found that non-O blood type roughly doubles the risk of venous thromboembolism (blood clots in the veins, including deep vein thrombosis and pulmonary embolism) compared with blood type O. This is one of the most consistent genetic findings in clotting research, on par with far better-known clotting risk genes.
Blood type also shapes resistance to two very different infections. Blood type O offers real protection against severe malaria caused by Plasmodium falciparum: infected red blood cells clump together with other red cells in a process called rosetting, which is a hallmark of dangerous, severe malaria, and this clumping happens much less easily on type O cells than on A, B or AB cells. Both a field study and a genetic association study focused on this ABO variant found lower odds of severe disease in blood group O, though they disagree sharply on how large the effect is — roughly a two-thirds reduction in the case-control study, against a more modest 15-20% relative difference in the larger genetic association study. The relationship with cholera runs the opposite direction: several studies, including a detailed household study in Bangladesh, found that people with blood type O who become infected with the cholera bacterium are more likely to develop severe, dehydrating disease than people with other blood types, even though blood type O individuals were somewhat less likely to become infected in the first place.
Some other ABO associations are genuine and have shown up in large, well-powered genetic studies, but the effect on any individual's risk is much smaller. Non-O blood types have been linked to a modestly higher risk of pancreatic cancer in genome-wide association studies, a finding that has replicated across cohorts but corresponds to only a small shift in absolute risk for an already uncommon cancer. Early in the COVID-19 pandemic, a large genetic study found that blood type A was associated with somewhat higher odds of severe respiratory failure and blood type O with somewhat lower odds; that signal was real in the original data but has been considerably less consistent and less pronounced in the many follow-up studies conducted since, and it has not translated into any change in clinical care based on blood type. Susceptibility to norovirus, the leading cause of stomach-flu outbreaks, is mostly governed by an entirely different gene (FUT2, which determines "secretor status"), though ABO type has some secondary influence on which norovirus strains bind most easily — a more minor piece of a picture dominated by that other gene. Older, mid-20th-century epidemiological reports linking blood type A to gastric cancer risk exist as well, but this association is weaker and less consistently confirmed in modern genetic studies than the clotting and malaria findings.
The overall lesson is a useful one for reading about genetics generally: not every disease link attached to a well-studied gene carries the same weight. Blood type genuinely matters for some things (transfusion, clotting risk, certain infections) and is, at most, a very minor factor for others.
ABO (9q34.2) encodes a glycosyltransferase that transfers a terminal sugar residue onto the H antigen precursor (itself generated by the FUT1-encoded fucosyltransferase). The A-transferase adds N-acetylgalactosamine; the B-transferase, differing from the A enzyme by a small number of amino acid substitutions that alter donor-sugar specificity, adds galactose. The classical O allele (commonly designated O01) arises from a single-nucleotide deletion in exon 6/7 of the coding sequence, historically described as c.261delG and catalogued as rs8176719, which produces a frameshift and premature stop codon, yielding a truncated, catalytically inactive protein. Ensembl/VEP annotates rs8176719 as a frameshift variant mapped to ABO, consistent with this mechanism. Individuals homozygous for O alleles (genotype OO) express only unmodified H antigen on red cells. A minority of O alleles (on the order of a few percent) arise from other inactivating variants (e.g., the O02 allele, a missense substitution) not tagged by rs8176719, so this SNP, while the dominant marker of the O phenotype, does not capture every O-causing allele.
ABO antibodies (isohemagglutinins) are naturally occurring IgM antibodies directed against whichever A/B antigen(s) an individual's own red cells lack; they are present from early infancy without prior antigen exposure and can cause acute intravascular hemolysis if incompatible blood is transfused. Type O red cells lack both A and B antigen and are used as the default red-cell product in emergencies when the recipient's type is unknown; type AB plasma, lacking anti-A and anti-B, is used analogously as universal plasma. ABO typing and crossmatching remain foundational to transfusion safety and are performed independently of Rh(D) typing.
Non-O blood group is associated with approximately 25% higher plasma concentrations of von Willebrand factor and factor VIII, attributed to ABO-antigen-dependent glycosylation effects on von Willebrand factor clearance (including altered susceptibility to ADAMTS13-mediated proteolysis). A meta-analysis of 38 studies (Dentali et al., 2012) found a pooled odds ratio of 2.09 (95% CI 1.83-2.38) for venous thromboembolism in non-O versus O blood group, identifying non-O status as the most common inherited/constitutional risk factor for VTE at a population level, exceeding the population attributable risk of factor V Leiden in most studied cohorts. Effect sizes are smaller in some non-European cohorts (e.g., Han Chinese studies report odds ratios closer to 1.3-1.4), suggesting some population-specific modification.
Rowe et al. (2007) demonstrated that blood group O erythrocytes form fewer and smaller rosettes with Plasmodium falciparum-infected erythrocytes than group A, B, or AB cells, and found blood group O associated with reduced odds of severe malaria in a Malian case-control study (OR 0.34, 95% CI 0.19-0.61). Fry et al. (2008), in a genome-wide association study spanning Gambian, Kenyan and Malawian cohorts, directly implicated common variation at the ABO locus (including the rs8176719-tagged O allele) in susceptibility to severe P. falciparum malaria, providing genetic (not only serologic) confirmation. Conversely, Harris et al. (2005), in a household-contact cohort in cholera-endemic Bangladesh, found blood group O individuals were somewhat less likely to acquire Vibrio cholerae O1 infection (OR 0.67) but more likely to develop severe cholera once infected (OR 2.3), an association whose precise mechanism remains incompletely defined.
Amundadottir et al. (2009), in a genome-wide association study of pancreatic cancer (discovery and replication cohorts totaling several thousand cases and controls), identified genome-wide-significant association between non-O ABO alleles and pancreatic cancer susceptibility; the finding has replicated across subsequent studies but corresponds to a modest per-allele odds ratio and a small absolute risk increase given the low baseline incidence of pancreatic cancer. Ellinghaus et al. (2020), in a genome-wide association study of severe COVID-19 with respiratory failure, reported association at the ABO locus (9q34.2), with blood group A associated with higher odds (OR 1.45, 95% CI 1.20-1.75) and blood group O with lower odds (OR 0.65, 95% CI 0.53-0.79) of severe disease relative to other groups; numerous subsequent studies and meta-analyses have found this association to be considerably less consistent in magnitude and significance, and it has not informed clinical risk stratification or management. Historical epidemiological reports (mid-20th century) associating blood group A with gastric cancer risk are of lower current evidentiary strength than the associations above and have not been as robustly confirmed in modern genetic studies. ABO phenotype has a secondary modulating influence on norovirus genotype-specific binding, but the dominant genetic determinant of norovirus susceptibility is FUT2 secretor status, a separate locus.
What a 23andMe/AncestryDNA export or raw VCF can and can't tell you about ABO Blood Group comes down to these specific, well-studied positions — not a diagnosis.
A single gene, ABO, makes an enzyme that attaches a sugar molecule onto red blood cells. The A and B versions of the enzyme attach different sugars; the O version, caused by a small DNA deletion (rs8176719) that scrambles the enzyme, attaches nothing. Your combination of two ABO gene copies determines your type: O requires two non-functional (O) copies.
Your immune system naturally makes antibodies against whichever ABO antigens you don't carry, so receiving the wrong blood type triggers an immediate immune attack on the transfused cells. Type O red cells are considered a safe default in emergencies because they lack the antigens those antibodies target.
Yes, this is one of the better-established associations in this area. People with non-O blood types (A, B, or AB) have higher levels of certain clotting factors and roughly double the risk of venous thromboembolism compared with people with type O, based on a large meta-analysis of dozens of studies.
Yes, and in opposite directions. Blood type O offers real, well-documented protection against severe falciparum malaria by making it harder for infected red cells to clump together. For cholera, the relationship flips: blood type O individuals who become infected are more likely to develop severe disease, even though they may be somewhat less likely to get infected in the first place.
An early, large genetic study found blood type A associated with somewhat higher odds and blood type O with somewhat lower odds of severe COVID-19 with respiratory failure. That signal was real in the original data but has been much less consistent in later, larger studies, and blood type has never been used to guide COVID-19 care.
Free to reuse. This page's text is original writing from freely-available research, licensed CC BY 4.0 — reuse it, including commercially, with attribution to MyGeneLog. It's general research-derived information, not medical advice or a diagnosis — see Terms of Use.