The main ABO groups and their subtypes.
Cardiovascular diseases (CVD) including coronary heart disease and stroke are leading causes of death and disability globally. Studies of the association between ABO blood groups and CVD have consistently demonstrated an increased risk of coronary heart disease, myocardial infarction, cerebral ischaemic stroke, peripheral arterial disease and venous thromboembolism (VTE) including deep vein thrombosis and pulmonary thromboembolism in patients who possess a non-O blood group type. The most likely mechanism is thought to be the increase in von Willebrand Factor (vWF) and factor VIII levels seen in patients with a non-O blood group. Other postulated mechanisms include elevations in circulating inflammatory markers such as endothelial cell and platelet adhesion molecules in subjects with a non-O blood group. More recently, it has also been recognised that individuals with a non-O blood group type carry a higher risk of SARS-C0V-2 infection and COVID-19 related complications. The increased levels in vWF and factor VIII amongst individuals with a non-O blood group who have contracted SARS-CoV-2 infection may result in an additive thrombophilic effect to that caused by the SARS-CoV-2 virus. Another postulated mechanism is that individuals with an O-blood group are protected by anti-A and B antibodies which possibly inhibit the binding of the SARS-CoV-2 spike protein to lung epithelium angiotensin converting enzyme-2 receptors. There are over 35 minor blood groups on red blood cells, some of which such as Kidd, Lewis and Duffy have been associated with CVD either alone or in combination with a non-O blood group allele(s). However, their role in SARS-CoV-2 infection and mechanism of action for an association with CVD remain unknown. This review explores the relationship between ABO and minor blood groups with CVD and VTE, with a focus on potential mechanisms underlying this relationship and the potential role of ABO blood group types in COVID.
- cardiovascular diseases
- venous thromboembolism
- ABO blood group
- von Willebrand Factor
Cardiovascular diseases (CVD) including coronary heart disease and stroke are the global leading cause of death and a major contributor to disability . Traditional modifiable risk factors include hypertension, dyslipidaemia, diabetes mellitus, current smoking, obesity and physical inactivity and non-modifiable risk factors include age, gender, family history and ethnic background . Amongst the non-modifiable risk factors, genetic variations in conjunction with traditional modifiable risk factors may significantly influence the trajectory of an individual’s CVD risk . Studies on the association between ABO blood group and CVD have consistently demonstrated that possession of the O blood group, the most common phenotype in most populations , confers protection against an individual developing a cardiovascular event [5, 6, 7, 8, 9, 10, 11]. The A and B blood groups are most frequently seen in Caucasian and Asian populations, respectively . However, the magnitude of the association between CVD and ABO blood grouping across different ethnic populations is controversial in part due to the higher population attributable risk of traditional modifiable vascular risk factors [9, 10, 12].
There is also a well-documented interaction between ABO blood group and venous thromboembolism (VTE) [7, 13, 14, 15]. The A2 blood subgroup, which is less common than A1 and rare in Asian populations , has been reported to be associated with a modest VTE risk (Odds Ratio 1.2) whereas the A1 and B subgroups confer a 1.8-fold increased risk . The non-O blood groups are associated with ~25–30% higher plasma levels of factor VIII and von Willebrand Factor (vWF) which are felt to be the major contributing factors to the increased risk of VTE [7, 16]. ABO blood grouping has been reported to influence activated protein C resistance , plasma lipid levels  and markers of inflammation including soluble intercellular adhesion molecule 1 (ICAM1), plasma soluble E-selection and P-selectin and tumour necrosis factor-alpha . An additive effect on VTE risk and ABO blood group has also been described in association with factor V Leiden and prothrombin gene mutations [10, 17]. Finally, ABO blood grouping has more recently been reported to influence susceptibility to SARS-CoV-2 infection and an individual’s propensity to more severe disease [18, 19, 20]. Proposed mechanisms include an additive risk of COVID-19 related thrombophilic complications in patients with a non-O blood group and a protective role of the O-blood group against the binding of the SARS-CoV-2 spike protein to lung epithelium angiotensin converting enzyme-2 receptor [18, 19, 20].
There are at least 35 minor blood group antigens in addition to ABO including Kidd, MNS, Duffy and Lewis  for which some, including Kidd and Lewis, have also been associated with CVD although the mechanisms of the associations are unclear [22, 23, 24].
This review will focus on the relationship between ABO blood types and CVD including coronary artery disease, ischaemic stroke, peripheral arterial disease and VTE (Figure 1). The contemporary relationship between SARS-CoV-2 infection, CVD and ABO blood grouping and the role of minor blood group antigens in the pathogenesis of CVD will also be discussed.
2. The ABO blood group system
The discovery of the ABO blood group system in 1900 by Austrian physician Karl Landsteiner saw him awarded of a Nobel Prize in physiology and medicine thirty years later . The ABO blood group system consists of three main alleles A, B and O with codominant A and B alleles resulting in an inheritance pattern consisting of six genotypes and four major blood types [4, 25]. The ABO locus is found on chromosome 9 (9q34.1-q34.2) and codes for 2 glycosyltransferases A and B that transfer N-acetyl-D-galactosamine and D-galactose to a H antigen acceptor site on red blood cells (RBC) producing A and B surface antigens and blood group types, respectively (Figure 2) . Lack of glycosyltransferase activity results in an unmodified ABO H antigen precursor and an O blood group type (O standing for ‘Ohne’, the German word for ‘without’) [25, 26]. Although ABO blood group antigens are RBC antigens, they are also expressed on human tissue including epithelial and endothelial cells .
The four basic ABO blood groups are O, A, B and AB. The ABO blood phenotypes have multiple subtypes, including A1, A2 and A3, in which 80% of blood group A cases are A1 in subtype, and O1, O2 and O3, in which O1 accounts for 95% of blood group O cases (Table 1) [4, 26]. From an evolutionary perspective, the oldest blood groups are A and O with A1 subtype considered the ancestral blood group [4, 7]. Single nucleotide polymorphisms (SNPs) of the ABO gene define the major haplotypes of European ancestry populations . A substitution from proline to leucine at amino acid 156 results in a change from A1 to the less common A2 allele [5, 7]. Substitutions from glycine to serine at amino acid 235, leucine to methionine at amino acid 266 and glycine to alanine at amino acid 268 results in B allele subtypes [7, 27]. In contrast, the O1 type is a consequence of a frameshift deletion of guanine (cdel261G, p88fs118Stop) which translates to a protein without enzymatic activity [4, 5, 7].
|ABO blood groups |
|A||A1, A2, A3 and other rare types including A4, A5, A6, Z, X, End, Boutu, g and i|
|B||B1, B2, B3 and rare types w, x, v and m|
|Other subtypes||O1, O2, O3 and other types including Yy, Hh, Xx and Bombay|
There are significant geographic and racial variations in the distribution of blood groups across the world [4, 25]. Contributing factors include migration over the time of humankind’s existence and processes of natural selection influenced by environmental factors such as climate and major diseases including malaria for which the O blood group confers protection . Blood group A is predominant in Northern and Central Europe, B in Central Asia and O in Africa, South America, and Australia [4, 5, 25]. However, there are isolated populations within each continent that have a completely different blood group . For example, the O blood group is common in different areas of Europe including Scandinavia and Switzerland [10, 25]. The most recent blood group, AB, appears to have arisen when A blood group populations in Europe migrated and mixed with B blood group populations of Asia .
Due to the association between clotting and ABO blood group, these geographic and racial differences may contribute to ethnic differences seen in rates of CVD.
3. ABO blood groups and arterial diseases
There have been numerous retrospective and prospective studies and meta-analyses demonstrating an association between non-O blood groups and CVD events [5, 8, 9, 10, 11, 12, 24]. These studies have consistently demonstrated an association between non-O blood groups and an increased risk of arterial disease including myocardial infarction, coronary heart disease, peripheral arterial disease, and ischaemic stroke (Table 2). Possible mechanisms for an association with CVD include vWF-related thrombosis and modulation of platelet function through other platelet proteins which also express ABO antigens such as glycoprotein IIb . The following two sections discuss the relationship between ABO blood groups and arterial diseases.
|Study and year||Type of study||Subjects||Age range||Country||Findings (non-O versus O)|
|Medalie et al. 1971 ||5-year prospective observational||10,000 males||>40 y||Israel||↑ MI and angina pectoris in A1, B, A1Jka−, BJka−|
|Wu et al. 2007 ||Meta-analysis||45 arterial studies with 9720 subjects||Multiple ages||Multiple countries||↑ MI (OR 1.25), PAD (OR 1.45) & cerebral arterial ischaemia (OR1.14) in non-blood O|
|Wiggins et al. 2009 ||Case control study|
(10 y follow-up)
|1063 MI, 469 ischaemic|
& 91 haemorrhagic stroke vs. 3462 C
|Age range 30–79||United States||↑ MI in blood group A11 & ↑ ischaemic stroke in blood group B compared to O1O1|
|He et al. 2012 ||26-year prospective observational||62,073 female nurses||30–55 y at baseline||United States||↑ coronary heart disease in A, B and AB blood groups|
|He et al. 2012 ||20-year prospective observational||27,428 males||40–75 y at baseline||United States||↑ coronary heart disease in blood groups A, B & AB|
|Zakai et al. 2014 ||Case cohort study (median 5.8 y follow-up)||646 stroke & 989 non-stroke controls (REGARD study)||Mean ages 63.6–66.8 y||United States||↑ stroke risk in blood group AB versus O with hazard ratio greater in those without diabetes mellitus|
|Vasan et al. 2016 ||SCANDAT2 (12.6 y median follow-up)||1112072 blood donors||Mean age 33 y at baseline||Sweden and Denmark||↑ MI & stroke in non-O blood group|
|Capuzzo et al. 2016 ||> 2-year observational|
(median 5.3 y)
|249 blood donors with Cardiorisk score > 20||35–55 y||Italy||↑ clinical or subclinical CVD events including ACS, cerebral ischaemia, cardiac arrhythmia & PAD in non-O group|
|Chen et al. 2016 ||Meta-analysis||17 studies of 225,810 participants||Multiple ranges||Multiple countries||↑ coronary heart disease in blood group A (OR 1.14) vs. O (OR 0.89)|
|Lin et al. 2017 ||Acute observational||1209 ST-elevation MI patients||Mean age 55 y||China||↑ post-MI spontaneous recanalization in O group|
|Fu et al. 2020 ||Case control study||61 HR & 600 randomly selected MI C||75.6 y (HR) v 66.2 y C||China||↑ HR following MI in blood group A|
4. ABO and coronary artery disease
In 1971, Medalie et al. reported the findings of a five-year prospective study of 10,000 Israeli male government employees aged ≥40 years who were born in six different regions (Eastern, Central and South-eastern Europe, Israel, Asia, and North Africa) . The study found that subjects with blood groups A1, B, and A1B tended to have higher rates of myocardial infarction and those with A1 and B had higher rates of angina pectoris when compared to other blood groups . Further, subjects who were negative for the Kidd glycoprotein (JKa−), a red blood cell urea transporter, had the highest rates of myocardial infarction and angina pectoris and adding this group to the ABO system (A1Jka−, BJka− and, particularly, A1BJka−) was associated with very high incidence rates .
A pooled analysis of two large prospective United States (US) cohort studies, the Nurses’ Health Study (NHS) which included 62,073 women and the Health Professionals Follow-up Study (HPFS) which included 27,428 men, both of which had >20 years follow-up, also found that the ABO blood group was significantly associated with an increased risk of coronary heart disease for men and women . A limitation of these two patient cohorts was the self-reporting of ABO and Rh factor status. However, a validation analysis of a subsample of 98 subjects found a 93% serologically confirmed ABO consistency for NHS and 90% consistency for HPFS. The combined analysis found that those with blood group A, B or AB were more likely to develop coronary heart disease and this risk was independent of age, level of physical activity, alcohol or smoking consumption or diabetes history [5, 11]. A meta-analysis, performed by the same authors, of 7 cohort studies including NHS and HPFS which combined a total of 114,648 individuals and 5,741 coronary heart disease cases found a significant pooled relative risk for coronary heart disease in patients with a non-O blood group of 1.1 (95% CI 1.05–1.18, p = 0.0001). Subjects with an O blood group had a lower risk for coronary heart disease when compared to B or AB with a trend seen for blood group A [5, 11].
ABO blood group status may be clinically relevant in subjects with concurrent cardiovascular risk factors. A study of 289 Italian blood donors with a high cardiovascular risk score (≥20) found that those with a non-O blood group had an increased risk of CVD events (including acute coronary syndrome, cerebral ischaemia, cardiac arrhythmias and supraaortic trunk or iliac artery stenosis) during a median follow-up of 5.3 years . ABO blood group status may also influence CVD-related patient outcomes. In a study of 1209 patients with acute myocardial infarction, Lin et al. found a higher rate of spontaneous recanalization following myocardial infarction in association with the O blood group whereas the rate of spontaneous recanalization was lower in subjects with an A blood group . Blood group A type has also been associated with an increased risk of heart rupture following myocardial infarction . Hence, ABO blood type may not only predispose susceptible individuals to an increased risk of CVD events but may also influence post-myocardial infarction outcomes.
5. ABO, stroke, and arterial disease in general
ABO blood group status has been shown to influence stroke risk. An association of AB blood group with stroke (adjusted Hazards Ratio [aHR] 1.83, 95% CI 1.01–3.30) was found in the (Reasons for Geographic And Racial Differences in Stroke [REGARDS]) study which involved 30,239 US participants followed up over 5.8 years . This finding remained significant after adjustment for age, gender, race region and Framingham stroke risk factors (systolic blood pressure, taking antihypertensive medication, diabetes, current smoking, atrial fibrillation and left ventricular hypertrophy). The association was greater in those participants without diabetes mellitus (aHR 3.33 95% CI, 1.61–6.88). Factor VIII levels accounted for 60% of the AB associated stroke risk . Another study by Wiggins et al. identified an increased risk of ischaemic stroke in subjects with a B blood group (OR 1.59, 95% CI 1.17–2.17) .
Meta-analysis studies have also demonstrated associations between non-O blood groups and arterial diseases in general. Chen et al. performed a meta-analysis of 17 studies involving 225,810 participants and found that blood group A was associated with an increased risk of coronary artery disease (OR 1.14, 95% CI 1.03–1.26, p = 0.01) and blood group O a lower risk (OR 0.85, 95% CI 0.78–0.94, p < 0.001) . These results remained significant after cases of myocardial infarction were excluded. Wu et al. conducted a systematic review and meta-analysis of 59 studies, both retrospective and prospective, reporting associations of ABO blood groups and arterial disease . They found significant ORs of 1.25 (95% CI 1.14–1.36) for myocardial infarction (n = 22 studies), 1.45 (95% CI 1.35–1.46) for peripheral arterial disease (n = 8 studies) and 1.14 (95% CI 1.01–1.27) for cerebral ischaemia of arterial origin (n = 7 studies) in subjects with a non-O blood group .
6. ABO blood groups and venous thromboembolism
The relationship between the ABO blood groups and VTE is most probably stronger than that seen in arterial diseases . The association between blood group subtypes and VTE has been well documented in a few genome-wide association (GWAS) [5, 31, 32], meta-analyses and/or case–control studies [7, 8, 10, 13] (Table 3). A French GWAS study involving 419 patients with early age onset of first deep vein thrombosis (DVT) who were compared to 1228 controls found that participants with blood type O had a 67% lower risk of VTE compared to n those with a non-O blood group [5, 31]. Relative to other non-O groups, subjects with the uncommon A2 subtype had a 47% lower risk [5, 31]. In the same study, Factor V Leiden mutations were also associated with increased risk of VTE although the authors did not investigate the potential for an additive risk in those having both a non-O blood group and a Factor V Leiden mutation . A similar association of non-O blood group with VTE was found in another GWAS study comparing 1503 VTE subjects to 1459 age and gender matched controls . In this study, the population attributable risk for VTE was highest for the non-O blood group, followed by blood type A, Factor V Leiden, and prothrombin G20210A .
|Study and year||Type of study||Subjects||Age range||Country||Findings (non-O versus O)|
|Morelli et al. 2005 ||Case control study|
|471 patients & 471 C||Not reported||Nether-|
|↑ VTE for all non-O blood group except A2A2 or any A2O combination|
|Wu et al. 2007 ||Meta-analysis||21 VTE studies with 6720 subjects||Multiple||Multiple||↑ VTE in A1A2/A1B/BB (OR 2.44) & A!O/BO/A2B (OR 2.11) blood groups|
|Lima et al. 2009 ||Case control study||65 VTE patients & 51 C||Mean age 34 y|
(range 6–67 y)
|Brazil||↑ VTE for FVL (OR 10.1) and double ↑ risk if also AB (OR 22.3)|
|Trégouët et al. 2009 ||Case control study (GWAS screening)||419 VTE patients & 1228 healthy C||Age at onset<50 y||France||↑ VTE in non-O blood group, relative lower risk VTE of A2 amongst non-O subjects and additive risk of FVL mutation|
|Wiggins et al. 2009 ||Case control study|
(10 y follow-up)
|Peri/post-MP women 504 VTE & 2172 C||Age range 30–89 y||United States||↑ VTE in B (OR 1.82), AB (OR 2.7) & A11 (OR 1.79) compared to O1O1 blood group|
|Heit et al. 2012 ||Case control study (GWAS analysis)||1503 VTE & 1459 age & gender matched C||Mean age for VTE & C 55 y||United States||↑ VTE population attributable risk for non-O blood group followed by A blood group, FVL & prothrombin G20210A|
|Vasan et al. 2016 ||SCANDAT2 (12.6 y median follow-up)||1112072 blood donors||Mean age 33 y at baseline||Sweden and Denmark||↑ VTE including pregnancy-related VTE & DVT in non-O blood group|
|Sun et al. 2017 ||Retrospective observational (7 y)||1412 VTE & 199,248 C||Mean 57.3 y VTE & 47.5 C||China||↑ VTE in non-O blood group (OR 1.35)|
|Goumidi et al. 2021 ||Pooled analysis of 6 studies||5425 VTE & 8445 C||Not reported||Multiple countries||↑ VTE for A2 (OR 1.2), A1 & B (OR 1.8) & ↓ VTE for O2 (OR 0.8) compared to O1|
There is evidence that the A2 subtype of blood group A is associated with a lower VTE risk when compared to the A1 allele [5, 33]. The A2 allele possesses a single base deletion near its carboxyl terminal (1061delC) which results in 30 to 50-fold less A-transferase activity than its A1 counterpart which suggests a correlation between the degree of H antigen glycosylation and VTE risk . Data from 2 population-based case control studies that included 504 post-menopausal women with non-fatal VTE found that the B and AB blood groups were both associated with an increased risk of VTE (OR 1.82, 95% CI 1.29–2.57, and OR 2.7, 95% CI 1.73–4.21, respectively) when compared to O1O1 subjects . Participants with A11, a subtype of the A1 allele, also carried a 79% increased risk of VTE (OR 1.79, 95% CI 1.41–2.26) . An increased VTE risk was also identified in a population-based case control study of venous thrombosis (Leiden Thrombophilia Study) . In this study, an increased thrombotic risk was found for non-O blood groups apart from those with genotypes homozygous to A2 or possession of any A2/O combination. Subjects with A1B/A2B and A1A1/A1A2 blood group genotypes had a 90–110% increased risk and those with BB/BO1/B02 genotypes had a 60% increased risk when compared to the OO genotype .
In a pooled analysis of 6 case–control/prospective studies of VTE, the A2 subgroup was found to be associated with a modest increase in VTE risk (OR 1.2) whereas the A1 and B subgroups had a 1.8-fold increased risk compared to the O1 subgroup. In contrast, O2 had a relative protective effect (OR 0.8) . Both the A1 and B subgroups were associated with increased vWF and factor VIII plasma levels whereas only the A1 subgroup was associated with increased ICAM levels . A meta-analysis of 21 VTE studies, 18 of which were retrospective in design, found a pooled OR of 1.79 (95% CI 1.56–2.05) for non-O blood group and VTE . The combination of A1A1/A1B/BB genotypes had an OR of 2.44 (95% CI 1.79–3.33) and A1O/BO/A2B an OR of 2.11 (95% CI 1.66–2.68) for VTE .
The presence of a non-O blood group has been found to correlate with unprovoked PTE, provoked including pregnancy induced and recurrent VTE. In a Scandinavian 25-year follow-up study of >1.6 million blood donors, those with a non-O blood group had higher rates of pregnancy-related VTE events, DVT and pulmonary embolism . The risks of recurrent pulmonary embolism and DVT provoked by comorbid illness were also higher in subjects with a non-O blood group . A large hospital-based retrospective study of 200,660 Han Chinese patients including 1412 VTE subjects (600 with DVT, 441 with pulmonary embolism and 371 with both conditions) conducted between 2010 and 2016 found a significant association of non-O blood group with VTE (OR 1.35, 95% CI 1.21–1.54) . Interestingly, subgroup analysis found a relatively greater non-O blood group risk for those with an unprovoked VTE (OR 1.86) compared to provoked VTE (OR 1.22). The OR for having a non-O blood group also appeared to decrease with age . Finally, a Brazilian case control study comparing 65 subjects with a history of DVT to 51 controls showed a significant increased risk of VTE in the presence of Factor V Leiden mutation (OR 10.1) which doubled in those in whom the AB allele of the ABO blood group was also present (OR 22.3) .
Future research into the role of developing risk stratification models or algorithms, for example, by combining ABO and other genetic variables with patient comorbidity and arterial risk factors to identify individuals at higher risk of VTE is warranted. This may translate to improved cardiovascular disease management including decision making for VTE prophylaxis at the hospital and outpatient setting.
7. Von Willebrand factor, factor VIII and other factors associated with ABO blood group and CVD
The majority of studies that have been presented in this review implicate an increase in plasma levels of vWF and factor VIII by non-O blood group types as the likely mechanism for an increased risk of thromboembolic events [5, 6, 7, 8, 9, 10, 11, 12, 13]. VWF/factor VIII are important in the acute phase response to vessel injury . VWF is a carrier of factor VIII, protects it from inactivation, can also recruit platelets to a site of vessel injury to induce a coagulation cascade responsible for clot formation . VWF can also bind to the platelet receptor glycoprotein Iba, to form a bridge between platelets and the endothelium and participate in a thrombo-inflammatory response (Figure 3) [35, 36].
ABO blood groups may have a direct functional effect on circulating vWF and thereby modulate both vWF and factor VIII levels . The mechanism by which the presence of an N-acetyl-D-galactosamine or D-galactose residue on glycans expressed on the H antigen acceptor site of a RBC influences plasma levels and/or bioactivity of vWF is unclear . A plausible mechanism is that ABO glycans on vWF itself influence its release into plasma and/or its clearance . VWF is derived from a pre-pro-polypeptide (ppvWF) synthesized in endothelial cells and megakaryocytes . The expression of blood antigen A on vWF correlates with decreased ppvWF to vWF ratios as well as longer half-life and increased plasma levels of vWF .
It has been postulated that the presence of A and B terminal carbohydrate antigens influence the proteolysis of vWF by its major protease ADAMTTS13 (
Blood group antigens are also associated with elevated plasma levels of markers of inflammation including endothelial cell and platelet-derived adhesion molecules . Elevated plasma levels of adhesion molecules including soluble P-selectin, soluble ICAM-1 and tumour necrosis factor-alpha are associated with ABO genotype, which may result in arterial and venous thrombosis and an increased CVD risk [5, 11]. However, in studies reporting the circulating expression levels of sICAM-1 and soluble P-selectin, the blood group A has paradoxically been found to be associated with lower circulating expression levels of sICAM-1 and soluble P-selectin when compared to the O blood group [38, 39]. This contradictory finding has been described in healthy Chinese populations and a study of Caucasian women without a history of chronic disease [38, 39]. A postulated explanation for why blood group A, in general, may be associated with lower circulating inflammatory markers despite carrying a higher CVD risk is that higher levels of sICAM-1 or soluble P-selectin expression are limited to those with significant CVD risk factors and/or symptomatic CVD. The concentrations of membrane forms of these adhesion molecules may also higher and thereby still mediate leucocyte migration and adhesion at an endothelial level . It is also possible that reduced levels of sICAM-1 increase the adhesion of leukocytes on endothelial surfaces which may result in increased arterial inflammation .
Platelet glycoproteins including GPIIb, and platelet endothelial cell adhesion molecule are known to carry ABO blood group antigens and may be involved in thrombosis through modulation of the GPIIb-GPIIIa fibrinogen receptor complex . A relationship between ABO blood group and angiotensin converting enzyme activity has also been reported which implicates a role for ABO blood group in the regulation of arterial risk factors such as hypertension .
Epidemiological studies have demonstrated an association between an elevated serum cholesterol including low-density lipoprotein cholesterol and non-O blood groups [5, 42]. These findings implicate ABO genotypes in the modulation of plasma lipids. ABO blood groups are also associated with phytosterol levels which have also been reported to modify cardiovascular risk .
8. ABO blood group, COVID-19, and CVD
There is growing evidence that ABO blood groups may play a role in the susceptibility to and severity of SARS-CoV-2 infection [18, 19, 20]. Individuals with blood group O have a lower risk and those with blood group A carry a higher risk of SARS-CoV-2 infection . A systematic review by the International Society of Blood Transfusion (ISBT) COVID-19 working group recently reported that subjects with blood group A had a higher rate of SARS-CoV-2 infection as well as an increased risk of requiring mechanical ventilation, continuous renal replacement therapy and prolonged intensive care unit stay . Postulated mechanisms include an increase in angiotensin converting enzyme-1 levels in blood group A patients and an increased risk of cardiovascular, thromboembolic, and inflammatory complications . It is plausible that possession of a non-O blood group and associated increase in vWF and factor VIII levels have an additive effect to the thrombophilia caused by SARS-CoV-2 and results in an increased risk of COVID-related CVD complications [18, 19, 20]. A recent hypothesis for an association between non-O blood group type and risk of SARS-CoV-2 infection is that anti-A and/or anti-B antibodies, which are present in patients with blood group O bind to a corresponding antigen, for example the angiotensin-converting-enzyme-2-receptor, on the SARS-CoV-2 viral envelope which then inhibits viral entry into lung epithelium (Figure 4) . Although ABH antigen structures are yet to be described on the SARS-CoV-2 protein, the spike protein has been reported to possess N-glycans and N-glycosylation sites which could potentially interact with anti-A and anti-B antigens and thus confer protection against infection in individuals with the blood group O type . The possibility that blood group A patients also have higher rates of underlying comorbidities which significantly contribute to COVID-related complications cannot be excluded .
A single-centre study from Bangladesh which evaluated 438 patients with SARS-CoV-2 infection also found a significantly higher rate of blood group A amongst COVID-19 patients compared to the general population . However, ABO blood groups were not associated with type of presentation or recovery from infection . Conversely, an observational study of 14,112 individuals tested for SARS-CoV-2 in the New York Presbyterian hospital system found that risk of intubation was increased for AB and B blood groups but decreased for A when compared to O blood group and risk of death was increased for those with AB blood group and decreased for A and B blood groups . Interestingly, Rhesus status, which is not implicated in CVD risk, correlated with COVID-19 risk . Rhesus-negative subjects had a lower risk of SARS-CoV-2 infection, intubation, and death . The mechanism of the relationships of ABO blood group, Rhesus status and SARS-CoV-2 infection is unknown. Further research into the relationship between blood groups and risk of SARS-CoV-2 infection and COVID-19 related complications is warranted.
9. Minor blood group antigens and CVD
There are over 35 minor blood group antigens on red blood cells , some of which including P and Lewis, are widely distributed in other human cells and body fluids [43, 44]. Minor blood groups have been associated with several diseases ranging from malignancy to peptic ulcer disease, infection and CVD [43, 44]. As discussed previously, subjects with a combination of blood groups A or B and the Kidd antigen Jka− status have been shown to be at increased risk of myocardial infarction . Sialyl-Lex (sLex), an antigen of the Lewis blood group system, is a major ligand for the cellular adhesion molecules E, P and L-selectin which are involved in the adhesion of leucocytes to endothelium . The Duffy blood group glycoprotein is a chemokine receptor on RBCs that is involved in the recruitment of leucocytes to sites of inflammation . Although the exact mechanisms are unclear, these biological characteristics offer explanation why the Lewis and Duffy blood group antigens may be associated with an increased risk of CVD.
The Lewis blood group system, which was first discovered by Mourant in 1946, is classified into four phenotypes [Le(a-b-), Le(a + b-), Le(a-b+), Le(a + b+)] determined by two genetic systems closely related on the short arm of chromosome 19 . In a study of 3385 Danish males, the Le(a-b-) phenotype was associated with an increased risk of mortality from ischaemic heart disease . In another Danish study involving 702 participants (72% male), the Le(a-b-) was associated with self-reported non-fatal stroke . However, a North Indian cross-sectional study that compared 161 patients with angiographically-proven coronary artery disease with 71 control subjects with normal angiography, found that the lack of Lewis antigen expression was associated with coronary artery disease in female subjects only . A lack of Lewis antigen expression has also been associated with a higher body mass index, weight gain over time, a lower level of physical activity, type 2 diabetes mellitus and hypertriglyceridemia all of which associated with an increased risk of CVD [49, 50, 51, 52].
The Duffy antigen, located on the long arm of chromosome 1 , has also been implicated in CVD risk . There are four Duffy phenotypes [Fy(a-b-), Fy(a + b-), Fy(a-b+), Fy(a + b+) . A study of 5301 African American participants found that Duffy negative subjects with a neutrophil: lymphocyte ratio ≥ 1.77 were more likely to have coronary artery disease and stroke . Lack of Duffy expression has also been associated with chronic organ damage, in particular renal dysfunction, in subjects with sickle-cell disease .
To the best of our knowledge, there have been no studies to date addressing the relationship of minor blood groups and COVID-19.
The limitations of studies evaluating the role of minor blood groups in CVD include their observational cohort design, case selection, outcome definitions, cohort sizes and influence of population attributable factors.
Subjects with non-O blood group type have an increased risk of arterial and venous thromboembolism. Blood groups A1, B and AB are at particularly increased risk of CVD events whereas blood group O confers a protective effect. Postulated mechanisms of underlying the relationship between ABO blood group and CVD include vWF and factor VIII activity and elevations in circulating inflammatory markers and plasma lipids. Comorbidities including arterial risk factors and predisposing factors to VTE such as concomitant Factor V Leiden mutations may have an additive effect to thromboembolic risk. Minor blood group types including Kidd, Lewis and Duffy are also been associated with CVD. The relationship of non-O blood group type and SARS-CoV-2 infection warrants further research. Future directions include the development and implementation of risk stratification algorithms for thromboembolic risk such as ABO blood group and other factors associated with arterial or venous disease in a hospital or outpatient setting.