A narrative review of the epidemiological and mechanistic associations between ABO blood groups and diseases: focusing on cardiovascular diseases, cancers, diabetes, malaria, COVID-19 and rheumatic diseases.

Introduction

The ABO blood group system (including types A, B, AB and O) represents one of the most extensively studied human blood group systems. Since its discovery, the relationship between ABO blood groups and human diseases has attracted considerable research interest. Accumulating evidence indicates that the ABO system is involved in the pathogenesis and progression of various diseases, though the specific mechanisms underlying these associations and their consistency across different populations require further elucidation.

The biological pathways through which ABO blood groups influence disease risk are diverse. One well-characterized mechanism involves ABO genetic variants regulating glycosyltransferase activity, which subsequently modulates plasma levels of von Willebrand factor (VWF) and factor VIII, as well as the expression of endothelial cell adhesion molecules. These molecular alterations promote thrombotic formation and inflammatory responses, ultimately increasing the risk of cardiovascular and other diseases. It should be noted that ABO allele frequencies demonstrate significant variation across ethnic groups, and confounding factors such as race, socioeconomic status and comorbidities may influence research findings. Therefore, cautious interpretation of observed associations is warranted.

Although previous systematic reviews and meta-analyses have consolidated evidence for specific associations (such as the relationship between non-O blood groups and venous thromboembolism and the protective effect of type O against severe malaria), a comprehensive synthesis integrating recent mechanistic advances across multiple disease domains remains lacking. This narrative review aims to summarize epidemiological evidence linking ABO blood groups to six major disease categories—cardiovascular diseases, malignancies, diabetes mellitus, Plasmodium falciparum malaria, COVID-19 and rheumatic diseases—while discussing potential underlying mechanisms. Due to limited available evidence, other areas such as transfusion outcomes and pregnancy complications fall beyond the scope of this review.

The primary objectives of this review are threefold: to summarize epidemiological associations between ABO blood groups and the selected diseases, to explore potential biological mechanisms and to identify consistencies and contradictions in existing evidence, thereby highlighting research gaps and suggesting future directions for strengthening causal inference and exploring translational applications.

Materials and methods

This narrative review provides a critical qualitative synthesis of evidence on associations between ABO blood groups and selected diseases, focusing on epidemiological patterns and potential mechanisms.

Literature search and study selection

A two-phase search was conducted in PubMed and CNKI databases (through September 2025). An initial exploratory search using broad keywords (‘ABO blood group’ combined with ‘susceptibility’, ‘incidence’ or ‘risk’) identified six major domains: cardiovascular diseases, malignancies, diabetes, malaria, COVID-19 and rheumatic diseases. Targeted searches for each domain were then combined with ‘ABO blood group’ and specific disease terms (e.g. ‘gastric cancer’ and ‘venous thromboembolism’). Searches were limited to English and Chinese publications.

Literature screening was performed through title, abstract and full-text review by the authors, with disagreements resolved by discussion. We included epidemiological studies (cohort, case–control or cross-sectional) or meta-analyses examining ABO–disease associations, as well as mechanistic investigations providing biological insights. We excluded studies with weak relevance, major methodological flaws (e.g. unclear case definitions and inappropriate controls) or case reports/small case series lacking a comparison group. Given the limited literature in certain areas (e.g. specific cancer subtypes and rheumatic diseases), we also included methodologically sound studies of moderate sample size. Priority was given to large-scale prospective cohorts and high-quality meta-analyses where available.

No formal systematic quality assessment or risk-of-bias scoring was performed; instead, we critically appraised methodological soundness on a case-by-case basis, considering sample size, study design, confounder adjustment and consistency across populations. Extracted data included study design, sample characteristics, ABO distributions, effect estimates and proposed mechanisms.

Evidence grading framework

To provide a transparent qualitative summary of evidence strength in Tables 2 and 3 , we adopted a three-tier framework. The framework is applied with different evaluative focuses: in Table 2 , it reflects the methodological quality of individual studies; in Table 3 , it reflects the overall consistency and strength of evidence for disease-level associations.

Table 2. Summary of ABO blood group associations with COVID-19 susceptibility and severity by region. Region Risk association (vs. reference group) Evidence tier Study design Key findings Japan [ 51 ] A↑, O↓(Infection); AB↑(Severe Disease) Tier 1 (Strong) Large-scale retrospective cohort study utilizing GWAS and genotyping data Supports the classic hypothesis of type A susceptibility and type O protection, while first suggesting in a Japanese population that type AB may increase severe disease risk. Turkey(Diyarbakir) [ 54 ] A ↑; O ↓ Tier 2 (Moderate) Large-sample, single-center cross-sectional study Findings support the classic hypothesis of type A susceptibility and type O protection. China (Xinjiang) [ 50 ] AB ↑; O ↓ Tier 2 (Moderate) Large-sample retrospective study First systematic report from the multi-ethnic population of Xinjiang, China. Supports the protective role of type O, but identifies type AB high risk as a regional characteristic. Saudi Arabia [ 52 ] B ↑; O ↓ Tier 2 (Moderate) Moderate-scale retrospective study with a large control group. Findings indicated increased infection risk for type B, differing from the mainstream finding of type A risk, suggesting regional variation in risk patterns. Pakistan(Peshawar) [ 53 ] B ↑; AB ↓ Tier 2 (Moderate) Large-sample, single-center cross-sectional study No significant association was found for types A or O, implying ethnic/regional variation in ABO associations. Western India(Jodhpur) [ 55 ] No Significant Association Tier 2 (Moderate) Large-sample retrospective study Long-term observation found no stable association between blood type and infection, indicating the ABO effect may vary over time and across populations. Southern Italy [ 49 ] No Significant Association (Infection); A ↑(Symptoms) Tier 3 (Preliminary) Moderate-scale retrospective study Suggests blood type may not affect infection rate, but type A is associated with an increased risk of developing symptoms. Southern Brazil [ 48 ] A↑(Infection); O↑, A↓ (Severe Disease) Tier 3 (Preliminary) Single-center cross-sectional study with limited sample size Findings indicated that type A was associated with higher infection risk, whereas type O had a higher risk of moderate/severe disease and type A was protective against severe outcomes. Note: Evidence tier follows the three-tier framework defined in the Materials and Methods section: Tier 1 (strong), Tier 2 (moderate) and Tier 3 (preliminary).

Summary of ABO blood group associations with COVID-19 susceptibility and severity by region.

Note: Evidence tier follows the three-tier framework defined in the Materials and Methods section: Tier 1 (strong), Tier 2 (moderate) and Tier 3 (preliminary).

Table 3. ABO blood groups and diseases with elevated risk: a summary. ABO blood group Associated disease/outcome Evidence strength Mechanism status Remarks Non-O Cardiovascular Disease (VTE, PE, MI, Coronary Heart Disease) Tier 1: Most Consistent Well-characterized (e.g. elevated VWF/FVIII) Associations are frequently observed across diverse populations. Plasmodium falciparum Malaria Tier 1: Most Consistent Well-characterized (e.g. erythrocyte rosetting) Risk is notably higher compared to blood group O in endemic areas. Pancreatic Cancer Tier 2: Moderately Consistent Proposed (e.g. H. pylori-related pathways) The relative risk magnitude may differ across different blood groups. COVID-19 (Susceptibility & Severity) Tier 3: Limited/Conflicting Proposed (e.g. antigen mimicry) Conflicting associations, population-specific (e.g. A in Japan/Turkey, B in Saudi Arabia and AB in Xinjiang) O Congenital Heart Disease Tier 3: Limited/Conflicting Unclear A modestly elevated risk was reported in a Danish cohort, contrasting with other studies. Specific Cancers (e.g. Thyroid and Endometrial) Tier 3: Limited/Conflicting Unclear Limited/conflicting evidence for elevated risk (thyroid/endometrial cancer; vs non-O); hypothesis-generating Rheumatic diseases (e.g. SLE and systemic sclerosis) Tier 3: Limited/Conflicting Unclear Current findings are conflicting, suggesting population-specific effects. Skin cancer (NMSC, melanoma) Tier 3: Limited/Conflicting Unclear Available studies show contradictory results. A Gastric Cancer Tier 2: Moderately Consistent Proposed (e.g. H. pylori colonization and mucosal immune differences) Modestly increased risk (OR ≈ 1.2), may be related to H. pylori-Le ᵇ -A-Le ᵇ interaction Diabetic complications (e.g. nephropathy and IHD) Tier 3: Limited/Suggestive Unclear Reported in two cohorts (T2DM in China; T1DM in Finland). Hypothesis-generating; requires validation. Head & Neck Cancer Tier 3: Limited/Suggestive Unclear Primarily based on case-control reports; needs confirmation. Skin cancer (NMSC, melanoma) Tier 3: Limited/Conflicting Unclear Available studies show contradictory results. Rheumatic diseases (e.g. RA and rheumatic heart disease) Tier 3: Limited/Conflicting Unclear Available studies show contradictory results. Congenital Heart Disease Tier 3: Limited/Conflicting Unclear Reported with higher frequency in Chinese case series. Effects likely vary by population. B Type 2 Diabetes Tier 3: Limited/Conflicting Unclear Associations have been reported but are not consistent across all groups. SLE (specific clinical manifestations) Tier 3: Limited/Suggestive Unclear Data are primarily from select population studies. Association limited to specific manifestations, not overall disease risk. RA Tier 3: Limited/Conflicting Unclear Current findings are conflicting, suggesting population-specific effects. Endometrial Cancer Tier 3: Limited/Conflicting Unclear Conflicting evidence on risk and prognosis, varying by population and tumor factors. Hypopharyngeal Cancer Tier 3: Limited/Suggestive Unclear Based on a limited number of studies. AB Congenital Heart Disease Tier 3: Limited/Conflicting Unclear Reported with higher frequency in Chinese case series. Data are limited. Note: The same disease may appear under multiple blood group rows to reflect conflicting findings across different studies and populations (e.g. CHD and endometrial cancer). Comparator groups vary across original studies (e.g. O vs. non-O and A vs. non-A); see main text and references for details.

ABO blood groups and diseases with elevated risk: a summary.

Note: The same disease may appear under multiple blood group rows to reflect conflicting findings across different studies and populations (e.g. CHD and endometrial cancer). Comparator groups vary across original studies (e.g. O vs. non-O and A vs. non-A); see main text and references for details.

Tier 1/Strong: Supported by multiple large-scale prospective cohorts or meta-analyses with consistent findings across populations and plausible experimental mechanisms. Examples: non-O blood groups and cardiovascular disease; type O and severe malaria protection.

Tier 2/Moderate: Supported by several studies but with notable heterogeneity across populations or designs, or incomplete mechanistic evidence. Examples: non-O blood groups and pancreatic cancer; type A and gastric cancer.

Tier 3/Preliminary (or Limited/Conflicting): Reported in few studies with inconsistent or non-replicated results, primarily from small or single-center reports; hypothesis-generating only. Examples: ABO associations with rheumatic diseases, endometrial cancer, thyroid cancer, skin cancer or COVID-19 (due to substantial heterogeneity across populations).

Grades were assigned by author consensus based on critical reading of the literature and are qualitative heuristics, not formal measures of causality. Table 3 was constructed by summarizing the reported elevated-risk associations identified for each blood group across the reviewed literature, organized by blood group to facilitate risk-profile visualization, with entries derived from the evidence summarized in the main text and graded according to the framework above.

Synthesis approach

Final synthesis involved critical analysis of selected literature to extract key findings, association strengths and mechanistic hypotheses across disease domains. Findings were organized thematically by disease category and discussed with reference to the consistency and limitations of the underlying evidence, as well as potential population-specific and methodological factors that may explain inter-study variability. No quantitative meta-analytic pooling was performed.

ABO blood type and cardiovascular disease

Overall cardiovascular risk

Multiple large-scale studies demonstrate a significant association between ABO blood type and cardiovascular disease (CVD). In a Chinese Han cohort of 18,000 participants, individuals with non-O blood types exhibited a 32% higher risk of cardiovascular events (CVEs) after adjustment for traditional risk factors [ 1 ]. Similarly, Capuzzo et al. followed 289 high-risk blood donors (Cardiorisk score ≥ 20) for a median of 8.7 years and found that non-O individuals had nearly a three-fold higher risk of CVEs (OR = 3.3; 95% CI: 1.1–10.1; p  = 0.033) [ 2 ]. These observations suggest that ABO blood type, an immutable genetic trait, may enhance current CVD risk-prediction models.

Thrombosis

Thrombotic events are a major cause of cardiovascular morbidity and mortality. A systematic review concluded that non-O individuals, particularly those with blood types A and B, have a higher incidence of venous thromboembolism (VTE) and pulmonary embolism (PE) than individuals with blood type O [ 3 ]. In a cohort of 7,830 European patients, type A carriers displayed a significantly elevated VTE and PE risk compared with type O carriers [ 4 ]. Moreover, an international pharmacovigilance study involving 2,347 young women corroborated that non-O blood type is an independent risk factor for VTE [ 5 ]. Further extending these observations, a recent registry study specifically associated blood group antigen B (types B and AB) with a significantly increased risk of early coronary stent thrombosis (OR 2.48; 95% CI 1.08–5.69), independent of traditional risk factors [ 6 ]. However, these findings may not be generalizable to all ethnicities or age groups, warranting further validation.

Myocardial infarction

Myocardial infarction (MI) is a severe complication of coronary artery disease (CAD) and is closely linked to coronary thrombosis. Liang et al. reported that non-O patients had a higher prevalence of acute MI than O patients, with the highest risk observed in type A individuals (OR = 1.44; 95% CI: 1.08–1.93) [ 7 ]. Subsequent angiographic studies revealed more extensive atherosclerosis and heightened platelet reactivity in non-O patients, providing mechanistic insight [ 8 ]. A meta-analysis of 17 studies confirmed that non-O blood type is associated with a 30% increased risk of MI [ 9 ].

Coronary heart disease

CAD is the leading cause of global mortality. Coronary angiography data from 986 Chinese patients demonstrated that the prevalence of significant CAD was highest in type A (80.1%) and lowest in type O (68.4%) ( p  < 0.05) [ 10 ]. Multivariable logistic regression identified blood type A as an independent predictor of CAD (OR = 1.52; 95% CI: 1.08–2.13). A meta-analysis of 17 prospective studies further validated a 14% increased CAD risk in type A individuals compared with non-A individuals (OR = 1.14; 95% CI: 1.03–1.26) [ 11 ]. Moreover, a large prospective cohort study likewise showed a higher coronary heart disease risk in non-O blood types (especially AB), with a maximal adjusted HR of 1.23 vs. type O. A meta-analysis further supported an overall 11% increased risk in non-O individuals (RR = 1.11, 95% CI: 1.05–1.18) [ 12 ].

Congenital heart disease

Evidence regarding the association between ABO blood type and congenital heart disease (CHD) remains inconsistent, though studies from China show consistent trends. In Guangdong, a study of 278 children with CHD found a higher prevalence of blood types A and AB compared to the local population, suggesting increased CHD susceptibility in these groups [ 13 ]. This was supported by a larger study in Yunnan involving 3,551 patients, which also reported elevated frequencies of types A and AB among cases [ 14 ]. Neither study linked ABO type to specific CHD subtypes. In contrast, a Danish cohort of 39,042 newborns found type A associated with lower risk of isolated CHD, while type O carried a modestly elevated risk [ 15 ]. Overall, ABO blood type may influence CHD susceptibility, but its effect likely varies by population due to regional, ethnic and genetic differences. Further multi-regional studies are needed to clarify its etiological role.

ABO blood type and cancer

Head and neck cancer

Case–control studies across diverse populations consistently report an increased risk of head and neck cancers in individuals with blood type A. A Greek study observed a 1.52-fold higher risk of head and neck cancers in type A individuals compared with non-A individuals [ 16 ]. Romanian investigators noted an elevated incidence of hypopharyngeal cancer in type B carriers but a protective effect against oral cavity cancers [ 17 ]. An Indian analytical study further supports these associations, identifying blood group A as a potential risk factor for oral, esophageal and salivary gland cancers, while blood group B was linked to an increased risk of laryngeal cancer [ 18 ]. The consistent associations across populations underscore the role of ABO blood groups in cancer risk, highlighting their value for risk stratification and prevention.

As the most common endocrine malignancy in the head and neck region, studies on the association between thyroid cancer and ABO blood groups have yielded inconsistent results. A Turkish study found that type B patients presented with more advanced-stage disease and extrathyroidal extension than non-B patients, whereas type O patients exhibited lower rates of capsular invasion [ 19 ]. Conversely, Lu et al. reported a higher susceptibility to papillary thyroid carcinoma in type O individuals [ 20 ]. Other studies found no association [ 21 ]. The contradictory findings across studies primarily stem from heterogeneity in study design and potential methodological biases. Consequently, larger, multicenter prospective investigations are warranted to resolve these inconsistencies.

Skin cancer

Current evidence suggests an association between ABO blood groups and the risk of skin cancer, although this correlation varies across different types of skin cancer and among different populations. Regarding non-melanoma skin cancer (NMSC), a case-series study of 512 Turkish patients showed that blood type A was more common among NMSC patients, while type O was relatively less frequent [ 22 ]. In contrast, a large prospective cohort study from the United States indicated that non-O blood types (including A, AB and B) may be associated with a reduced risk of NMSC, particularly squamous cell carcinoma and basal cell carcinoma [ 23 ].

The evidence concerning melanoma appears more complex. A case-control study conducted in Northeast China found that blood type A was significantly positively associated with both the incidence and metastasis of cutaneous malignant melanoma [ 24 ]. Meanwhile, a European study reported a significantly elevated risk of malignant melanoma in individuals with blood type O Rh-negative status [ 25 ]. These findings collectively suggest that the relationship between ABO blood groups and melanoma may be population-specific, varying across different geographic and genetic backgrounds.

Gastric cancer

Gastric cancer risk is consistently elevated in individuals with blood type A. A meta-analysis of 24 studies demonstrated a 21% increased risk in type A individuals (OR = 1.21; 95% CI: 1.06–1.55) [ 26 ]. A Guangdong case–control study confirmed that 36.7% of gastric cancer patients were type A compared with 26.0% of controls ( p  < 0.05), yielding a relative risk of 1.65 [ 27 ]. Mechanistically, Helicobacter pylori, as a significant gastric pathogen, is thought to potentially bind to fucosylated blood group antigens (such as Lewis b) on the gastric mucosal surface via the BabA adhesin. However, in individuals with blood type A, the Leᵇ structure is modified at its terminal end by GalNAc to form A-Leᵇ, which may partially interfere with this binding [ 28 ]. This could partly explain the differences in infection susceptibility and subsequent pathological development among individuals with different blood types. Additionally, secretor status is also considered an influencing factor. Low secretors often exhibit more stable mucosal fucosylation patterns, relatively lower H. pylori bacterial loads and milder gastric inflammatory responses, suggesting a potential protective role during infection [ 29 ]. In summary, ABO blood type and secretor status may collectively regulate the mucosal colonization of H. pylori and the host immune response, thereby influencing the development of gastric cancer.

Pancreatic cancer

Pancreatic cancer is highly malignant and carries an extremely poor prognosis. In the Pancreatic Cancer Cohort Consortium (1,028 cases and 3,871 controls), individuals with blood types A, AB and B exhibited 32%, 51% and 72% increased risks, respectively, compared with type O [ 30 ]. Similarly, a large prospective cohort study by Wolpin et al. found that non-O blood groups (A, AB and B) were significantly associated with elevated pancreatic cancer risk, with hazard ratios of 1.32, 1.51 and 1.72, respectively. The study estimated an attributable risk of approximately 17% for pancreatic cancer associated with non-O blood groups [ 31 ]. A meta-analysis further corroborated that blood group A was consistently linked to increased risk across diverse populations, whereas the risks for blood groups B and AB varied by geographic region and the endemicity of Helicobacter pylori CagA [ 32 ]. Additionally, a Japanese retrospective study reported that patients with blood type O receiving adjuvant chemotherapy had a superior 5-year survival rate (70.7%) compared to non-O patients (46.2%) [ 33 ], suggesting that ABO status may also influence therapeutic response.

Endometrial cancer

Evidence regarding the association between ABO blood type and endometrial cancer remains inconclusive. A recent study confirmed type B as a significant risk factor for endometrial cancer ( p  < 0.05), independent of BMI and other confounding factors [ 34 ]. However, a Chinese study reported that type O was associated with an elevated risk of type I endometrial cancer, whereas no association was observed for tumor stage or grade [ 35 ]. Conversely, a Turkish survival analysis found that type B patients experienced the longest overall survival, although the difference did not reach statistical significance ( p  = 0.17) [ 36 ]. Future studies should further incorporate molecular subtypes and hormonal factors into the analysis, while taking into account the genetic diversity across different populations. Large-scale prospective studies or meta-analyses should be conducted to clarify the potential role of ABO blood type in the occurrence, progression and prognosis of endometrial cancer, thereby providing a more reliable basis for individualized risk assessment.

ABO blood type and diabetes mellitus

Recent studies have shed light on the intricate relationship between the genetic underpinnings of the ABO blood group system and type 1 diabetes risk. The key gene FUT2, which regulates ABO antigen secretion, plays a critical role in this association. Individuals homozygous for the loss-of-function variant (nonsecretors) demonstrate a significantly elevated risk of developing type 1 diabetes (OR = 1.29), highlighting the complex interplay between host defense mechanisms and autoimmune pathogenesis [ 37 ]. These genetically determined biological pathways provide novel insights into how blood group characteristics may influence diabetes susceptibility and related complications.

In contrast, epidemiological evidence regarding the direct association between ABO blood types and type 2 diabetes reveals considerable inconsistencies. A Malaysian study reported a higher prevalence of diabetes in type B individuals (OR = 1.21; 95% CI: 1.05–1.40) [ 38 ], whereas a Chinese hospital-based study found an increased risk in type A individuals [ 39 ]. These discrepancies likely reflect ethnic and environmental heterogeneity.

Beyond disease risk, ABO antigens may influence diabetic complications. In a cross-sectional study of 1,200 Chinese patients with type 2 diabetes, those with blood types A or AB exhibited a significantly higher prevalence of diabetic nephropathy (OR = 1.43 and 1.37, respectively) [ 40 ]. Among type 1 diabetic patients, type A carriers with nephropathy had a markedly elevated risk of ischemic heart disease [ 41 ]. These findings highlight the potential utility of ABO typing in diabetic risk stratification.

ABO blood type and Plasmodium falciparum malaria

Malaria, a serious parasitic disease caused by Plasmodium, remains a major global health threat. Evidence suggests that ABO blood types influence susceptibility to severe malaria, with type O conferring protection. For instance, a study conducted in Ghana found that children with blood type O had a significantly lower risk of developing complicated malaria compared to non-O blood type children [ 42 ]. Similarly, research by Tekeste et al. reported a reduced risk of severe malaria among individuals with blood type O [ 43 ]. In contrast, non-O blood types (A, B and AB) were associated with higher susceptibility to malaria. Data from studies in India and Ethiopia further indicated a significantly increased risk of severe malaria in individuals with blood type B (OR = 4.09) [ 44 ], as well as an overall elevated risk for those with blood types A, B and AB (A/B: OR = 2.5; AB: OR = 3.3) [ 45 ]. Additional studies in Africa and meta-analyses further support the protective role of type O blood [ 46 ]. Although one study in eastern China did not find a broad association with infection risk, type O patients showed lower rates of severe complications [ 47 ].

It is important to note that the observed association exhibits geographic heterogeneity. Variations in findings across different study regions may stem from local differences in circulating Plasmodium strains, host genetic backgrounds and malaria transmission intensity. Therefore, it is essential to consider these regional factors when evaluating the universality of ABO blood groups as biomarkers for malaria risk.

ABO blood type and COVID-19

Susceptibility

To examine the association between ABO blood groups and COVID-19 susceptibility, we searched PubMed and CNKI, prioritizing studies with sound methodology, representative samples and transparent reporting. Eight studies from Asia, Europe, South America and the Middle East were included to minimize bias from limited regional sampling [ 48–55 ].

Integrated analysis revealed a complex relationship. Blood group O showed a protective tendency in several regions (e.g. Japan and Turkey), while group A was often linked to higher infection or severity risk [ 51 , 54 ]. However, this pattern was not universal: studies from Saudi Arabia highlighted group B as a risk factor [ 52 ] and research in Xinjiang, China, associated group AB with elevated risk [ 50 ] ( Table 1 ).

Table 1. ABO blood group distribution in COVID-19 patients and control populations across multiple regions. Region Control Populations Cases Controls AB (%) A (%) B (%) O (%) Total AB (%) A (%) B (%) O (%) Total Southern Brazil [ 48 ] Matched non-COVID individuals 1 (0.93) 54 (50.00) 6 (5.56) 47 (43.52) 108 7 (3.07) 83 (36.40) 26 (11.40) 112 (49.12) 228 Southern Italy [ 49 ] SARS-CoV-2-negative subjects 6 (3.59) 77 (46.11) 15 (8.98) 69 (41.32) 167 44 (4.94) 349 (39.17) 116 (13.02) 382 (42.87) 891 China (Xinjiang) [ 50 ] Healthy population (same period) 118 (14.20) 250 (30.08) 257 (30.93) 206 (24.79) 831 295 (10.62) 819 (29.48) 838 (30.17) 826 (29.73) 2778 Japan [ 51 ] Healthy population genotype database 188 (10.50) 745 (41.62) 388 (21.68) 469 (26.20) 1790 58 (9.49) 225 (36.82) 140 (22.91) 188 (30.77) 611 Saudi Arabia [ 52 ] Healthy population 38 (6.33) 180 (30.00) 145 (24.17) 237 (39.50) 600 411 (4.88) 2518 (29.89) 1513 (17.96) 3981 (47.26) 8423 Pakistan(Peshawar) [ 53 ] Healthy blood donors 228 (11.78) 523 (27.03) 694 (35.87) 490 (25.32) 1935 275 (14.21) 557 (28.78) 617 (31.89) 486 (25.12) 1935 Turkey(Diyarbakir) [ 54 ] Hospital-based population registry 156 (9.36) 753 (45.17) 311 (18.66) 447 (26.81) 1667 10,236 (8.05) 50,449 (39.69) 23,678 (18.63) 42,728 (33.62) 127091 Western India(Jodhpur) [ 55 ] Historical population database 91 (7.98) 266 (23.33) 426 (37.37) 357 (31.32) 1140 7,924 (9.48) 18,626 (22.27) 30,497 (36.47) 26,584 (31.79) 83631 Note: Data are n (%). Percentages may not sum to 100% due to rounding. Control definitions vary by study (see second column).

ABO blood group distribution in COVID-19 patients and control populations across multiple regions.

Note: Data are n (%). Percentages may not sum to 100% due to rounding. Control definitions vary by study (see second column).

Our thorough review of the methodologies and findings from the included literature suggests that differences in research approaches may help explain some of the inconsistencies observed in the results. Japanese studies using large-scale genotyping showed strong internal consistency [ 51 ], whereas a long-term study in Western India found no stable ABO–COVID-19 association [ 55 ] ( Table 2 ). These results suggest ABO blood group is not an independent determinant of susceptibility but may interact with genetic, environmental or temporal factors.

In summary, while blood groups O and A show relatively consistent associations across multiple regions, the ABO–COVID-19 relationship is highly diverse and context-dependent. Future studies should account for population structure and environmental confounders to clarify its biological mechanisms and potential health implications.

Disease severity

Data on ABO blood type and COVID-19 severity remain inconsistent. An analysis of 14,112 hospitalized patients in New York showed that type AB was associated with increased mortality, whereas type A exhibited a lower intubation risk [ 56 ]. Another study of 128 critically ill patients found that those with blood types A and AB required mechanical ventilation more frequently (71% vs 52%, p  = 0.041), had higher ICU mortality (29% vs 11%, p  = 0.018) and experienced longer ICU and hospital stays (median 10 vs 7 days, p  = 0.016; 26 vs 17 days, p  = 0.034) compared to those with types B or O [ 57 ]. A study in Ghana observed a similar trend, linking blood group O to asymptomatic infection and non-O blood groups to symptomatic disease, although these associations did not reach statistical significance [ 58 ]. Other investigations failed to confirm these associations [ 59 ]. Discrepancies may arise from differences in variant circulation, vaccination status and population structure.

ABO blood type and rheumatic diseases

Current evidence on the association between ABO blood groups and rheumatic diseases is inconsistent and limited. Some studies report specific links, such as a higher frequency of blood group B or ABH non-secretor status in rheumatoid arthritis (RA) [ 60 ], and a predominance of blood group A and non-secretor status in rheumatic heart disease [ 61 ]. Recent findings also suggest that in systemic lupus erythematosus (SLE), blood group B is associated with autoimmune hemolytic anemia, while Rh positivity correlates with arthritis, though no link has been established between blood type and serological markers in RA [ 62 ]. However, multiple other studies—from Egypt, Finland, Iran and elsewhere—found no significant difference in overall ABO distribution between patients and the general population [ 62–64 ]. A Turkish study further indicated possible disease-specific variation, with A blood group more common in spondyloarthritis, vasculitis, Behçet’s disease and RA, and O blood group more frequent in SLE and systemic sclerosis [ 65 ].

Overall, heterogeneity in results likely reflects differences in population genetics, diagnostic criteria, sample sizes and variable attention to secretor status. Current evidence does not support using ABO blood group or secretor status as reliable predictive markers for rheumatic diseases. Further clarification will require large, prospective, multi-population studies incorporating secretor typing and mechanistic research.

Possible mechanisms

Cardiovascular disease

Approximately 30% of the heritable variation in plasma von Willebrand factor (VWF) levels is attributable to ABO locus polymorphisms [ 66 ]. Non-O individuals exhibit 25%– 30 % higher VWF and factor VIII concentrations than type O individuals [ 67–70 ]. A recent multi-omics study further revealed that hypomethylation at the promoter/enhancer region of the ABO gene in type O individuals is associated with reduced VWF and elevated soluble E-selectin levels, suggesting that epigenetic mechanisms partially mediate the influence of ABO blood type on plasma protein expression and thrombotic risk [ 71 ]. These molecular differences help explain the enhanced VWF–platelet interactions and thrombin generation that underlie the increased thrombotic risk observed in non-O carriers. Additionally, ABO glycosylation modulates the clearance of VWF and the expression of endothelial adhesion molecules, further influencing vascular homeostasis [ 72–80 ]. Moreover, a recent prospective cohort study confirmed that elevated plasma levels of ABO glycosyltransferase (BGAT) independently predict the risk of venous thromboembolism (VTE)—particularly unprovoked VTE and deep vein thrombosis—even after adjusting for VWF and FVIII levels, suggesting that the underlying mechanisms extend beyond classical coagulation pathways [ 81 ].

Plasmodium falciparum malaria

The ABO blood type influences susceptibility to malaria through multiple mechanisms, including red blood cell rosette formation, macrophage phagocytosis, alterations in erythrocyte membrane proteins and antigen expression, natural antibody-mediated immune responses and modulation of inflammatory reactions and cytokines [ 82 ]. Specifically, the presence of A and B antigens on the surface of type A and B erythrocytes facilitates the binding of Plasmodium-infected red blood cells to uninfected erythrocytes, forming rosettes that can obstruct microvasculature and increase the risk of severe malaria. In contrast, type O erythrocytes, which lack A and B antigens, form fewer and less stable rosettes. They are also more readily recognized and phagocytosed by macrophages. Additionally, higher levels of natural antibodies and a moderated inflammatory response may contribute to the relative resistance to malaria observed in individuals with blood type O [ 82–84 ]. Together, these mechanisms explain the lower infection rates and improved clinical outcomes among type O individuals in malaria-endemic regions, providing a scientific basis for understanding the relationship between ABO blood types and malaria susceptibility.

COVID-19

In terms of COVID-19 susceptibility, the association between ABO blood groups and SARS-CoV-2 infection risk may stem from interactions between viral surface proteins and host blood group antigens, along with ABO-related immunomodulatory differences. The molecular basis of this association is closely linked to the expression of ABH antigens: in secretor individuals (determined by the FUT2 gene), ABH antigens synthesized by respiratory epithelial cells can be acquired by viruses, forming an ‘antigenic camouflage’.

Patrice Guillon et al. showed that anti-A and anti-B antibodies—both monoclonal and natural human antibodies—inhibit adhesion between A-antigen-expressing hamster ovary cells and ACE2-expressing Vero E6 cells, suggesting that such antibodies may protect by interfering with the SARS-CoV spike protein–ACE2 interaction [ 85 ]. This mechanism may also apply to SARS-CoV-2, which similarly uses ACE2 as its entry receptor [ 86–88 ]. Deleers et al. further confirmed that when the SARS-CoV-2 spike protein is expressed in cells with a complete ABH glycosylation pathway, it can be modified to carry A or B antigens. Additionally, COVID-19 patients exhibited significantly lower levels of anti-A and/or anti-B IgM antibodies compared to healthy controls, indicating that pre-infection antibody titers are a critical factor influencing susceptibility [ 89 ]. Mortensen et al. found that individuals with blood group O, as well as those with high titers of anti-A IgG or anti-B IgM, had a significantly reduced risk of COVID-19 [ 90 ]. Matzhold et al. also confirmed the protective effect of blood group O [ 91 ].

In summary, the expression of ABH antigens on the viral surface, the secretory status of the individual and pre-existing antibody levels collectively constitute a critical interface regulating SARS-CoV-2 infection, providing a mechanistic explanation for the differential susceptibility among individuals with different blood types.

Discussion

This narrative review summarizes the epidemiological associations and potential biological mechanisms between ABO blood groups and six major categories of diseases. Current evidence suggests that the ABO blood group system, as a genetic background, plays a non-negligible role in the development and progression of various diseases, although the magnitude, direction and specific mechanisms of its effects vary across disease types.

In the field of cardiovascular diseases, non-O blood group individuals generally show a tendency for higher thrombosis risk and cardiovascular event rates, an association that has been repeatedly observed across different populations. The underlying mechanisms are relatively well understood, primarily involving ABO gene-regulated increases in plasma levels of VWF and FVIII, alterations in endothelial cell adhesion molecule expression, and epigenetic regulation, providing important insights into the pathophysiology of thrombotic diseases.

For infectious diseases, the protective mechanism of blood group O against severe malaria is relatively clear, including reduced rosette formation of red blood cells, enhanced macrophage clearance efficiency, and modulation of innate immune responses. In contrast, the relationship between ABO blood groups and COVID-19 appears to be more complex: although multiple studies suggest that blood group A may be associated with higher infection risk and disease severity, while blood group O appears to confer some protection, these associations have varied across regions, populations and viral variants. Proposed mechanisms include interactions between viral surface proteins and ABH antigens, differences in secretor status and variations in naturally occurring anti-A/anti-B antibody levels; however, the specific causal pathways still require further experimental validation.

Notably, the role of ABO blood groups in modulating the gut microbiota offers a new perspective on their influence in infection and immunity. Research has found that ABO blood groups participate in the regulation of infection and immunity by shaping the gut microbiota through the modulation of mucosal glycan availability. In secretor individuals, non-O blood types are associated with alterations in bacteria such as Bacteroides, while blood group A is linked to changes in microbial families like Lachnospiraceae [ 92 , 93 ]. Genome-wide association studies and pig model experiments collectively confirm that ABO gene variants affect the abundance of specific bacteria (e.g. Bifidobacterium) and influence bacterial growth by regulating carbon sources such as GalNAc [ 94 , 95 ]. This evidence forms a new microbiota-immune pathway connecting blood groups to disease risk.

In oncology, the association between blood group A and increased gastric cancer risk is relatively consistent, partly potentially explained by Helicobacter pylori colonization characteristics in the gastric mucosa and differences in host immune responses. However, for other cancers such as thyroid and endometrial cancer, study results remain inconsistent, suggesting possible unrecognized confounding factors, population-specific genetic backgrounds or limitations in study design.

Regarding diabetes and rheumatic diseases, the role of ABO blood groups has progressed from epidemiological associations to mechanistic investigation. In diabetes, gene polymorphisms such as FUT2 and secretor status can influence the risk of type 1 diabetes. Studies on rheumatic diseases suggest that specific blood group phenotypes may be associated with conditions such as rheumatoid arthritis and rheumatic heart disease. However, these associations show considerable heterogeneity across different populations. Beyond pathways involving chronic inflammation and immune regulation, ABO-related gut microbiota dysbiosis may contribute to disease pathogenesis through the microbiota-immune axis, offering a new direction for mechanistic studies.

It should be noted that research in this field continues to face common challenges. Most evidence comes from observational epidemiological studies, making it difficult to fully exclude confounding biases; some associations have not been consistently replicated across different ethnicities or regions, suggesting that population genetic background and environmental exposures may be important effect modifiers; moreover, the independent predictive value and clinical utility of ABO blood groups as disease risk markers have not been systematically evaluated in large-scale prospective studies, and there is currently insufficient evidence to support their use in individualized prevention or clinical decision-making.

Conclusion and future directions

In summary, the ABO blood group system demonstrates diverse association patterns across various disease domains through mechanisms such as modulating coagulation factor levels, influencing pathogen-host interactions and regulating immune microenvironments. The strength of evidence exhibits a discernible gradient: its role in predicting the risk of cardiovascular events like venous thromboembolism and myocardial infarction is relatively well-established; its association with the prognosis of severe malaria also garners considerable support; whereas for certain cancers such as gastric and pancreatic cancer, as well as COVID-19 and some autoimmune diseases, the existing evidence remains inconsistent, with effects likely modulated by factors including population genetic background, environmental exposures or pathogen variations. To provide a clearer overview of these associations, we have summarized the high-risk diseases, strength of evidence and current research status on mechanisms corresponding to each blood group in Table 3 . It should be noted that the studies included in this review are predominantly based on observational data and may lack representativeness in certain regions and ethnic groups; thus, the extrapolation of conclusions should be approached with caution. Overall, the ABO blood group, as an important genetic background factor, offers a unique perspective for understanding disease susceptibility and clinical heterogeneity, yet its clinical translation remains exploratory. Future efforts should focus on conducting prospective cohort studies in more diverse populations, integrating multi-omics approaches and advanced typing technologies to systematically elucidate ABO-related molecular pathways and rigorously evaluating its incremental value in risk stratification models, thereby laying a scientific foundation for exploring its potential role in precision health.

Funding Statement

This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.

Disclosure statement

The authors declare that they have no competing interests.

Data availability

No new primary data were generated in the course of this review. The literature extraction forms and summary tables used in the analysis are available from the corresponding author upon reasonable request.

References

Associated Data

Data Availability Statement

No new primary data were generated in the course of this review. The literature extraction forms and summary tables used in the analysis are available from the corresponding author upon reasonable request.