Allelic distribution of ABO gene in Chinese centenarians

Abstract Objective Human ABO blood groups are determined by the alleles A, B, and O (O01 and O02) of the ABO gene and have been linked to the risks for cardiovascular diseases and cancers that affect lifespan. We examined the genetic associations of the ABO gene and blood groups with longevity. Methods We inspected the frequencies of the A, B, O, and O02 alleles in a large Chinese centenarian population (n = 2201) and in middle‐aged controls (n = 2330). The single nucleotide polymorphisms were selected as allele A (rs507666), B (rs8176743, rs8176746, and rs8176749), O (rs687289), and O02 (rs688976, rs549446, and rs512770). Results Supported by allelic and genotypic association studies, the frequencies of blood types A, B, O, and AB in centenarian versus control participants were not statistically different: 0.2821 versus 0.2781 (χ2 = 0.09, P = 0.76), 0.2867 versus 0.3060 (χ2 = 2.03, P = 0.15), 0.3380 versus 0.3159 (χ2 = 2.52, P = 0.11), and 0.0859 versus 0.0910 (χ2 = 0.37, P = 0.54), respectively. Sex had little effect on these distributions. Conclusion Integrated with other previous reports, we conclude from this large Chinese cohort that genetic variants of the ABO gene and blood groups are not associated with longevity.

gene, 2 allowing for the use of biological materials alternative to the blood sample, such as fingernails, hair, saliva, and oral mucous membranes, in the typology of ABO blood groups.
The human ABO gene is located at chromosome 9q34. 1 The ABO blood group system is the most widely used blood group system. 8 It not only plays a role in blood transfusion and transplantation but is also of interest to many researchers for its relation with diseases. For instance, it has been shown that blood type A is a risk factor for gastric cancer [9][10][11] while blood type O is a protective factor for atherosclerosis. [12][13][14] Cardiovascular diseases and cancers impact the lifespan significantly. Thus, the association between ABO blood types and human longevity has naturally been evaluated previously. It was reported as early as the 1960s that individuals with type A lived longer. 15 Later, a number of studies reported that type B and O were associated with longer lifespan or longevity phenotypically and genotypically. [16][17][18][19] However, these findings remain debatable. 20 The debate is possibly caused by small population sizes and stratifications as well as the moderate effects of blood type on longevity. To our knowledge, the largest population to test the genetic association of blood types with human longevity consisted of only 269 centenarians. 16  it has become necessary to evaluate the association of ABO blood groups in large longevity populations.
In order to search for factors that influence healthy aging and longevity, we initiated the Chinese Longitudinal Healthy Longevity Survey (CLHLS) in a large Chinese cohort from 1998 to 2014 and carried out genetic screening, leading to the identification of a number of genes associated with human longevity. [21][22][23][24] Among these studies, datasets of the genome-wide association study that included 2178 centenarians and 2299 middle-aged controls 23,24 were subjected to searches for the genetic associations of the ABO gene and blood groups with longevity.

| Samples and data source
Sampling, population quality, and genotyping on the cohort have been reported. [23][24][25] Samples and data from the CLHLS were randomly selected from half of the counties and cities in 22

| Selection of single nucleotide polymorphisms for ABO alleles, genotypes, and blood types
Eight single nucleotide polymorphisms (SNPs)-including rs507666

| Statistical analysis
The frequency of each SNP was calculated and used to evaluate its departure from Hardy-Weinberg equilibrium by a chi-square test.
Differences in allele, genotype, and blood type distribution between cases (centenarians) and controls (middle-aged individuals) were analyzed using binary logistic regression adjusted for nongenetic covariates under various genetic models. 21 Sex was also analyzed separately. The chi-square test was performed using GraphPad Prism (Version 8.4.2). A Bonferroni method was used for multiple comparison correction. 27 The chi-square values, odds ratios (ORs), 95% confidence intervals (CIs), and P values were presented for all association tests. A P value < 0.05 was considered to be statistically significant.

| Search strategy
We performed a systematic literature search using PubMed, ScienceDirect, Wiley, Oxford Academic, Web of Science, and SinoMed for studies reporting the association between ABO blood groups and longevity, including relevant articles and reviews (up to April 2020). Only studies published in English were considered. Two search themes were combined using the Boolean operator "AND"; the first theme was "ABO AND longevity," and the second theme was "ABO AND lifespan." F I G U R E 1 Flowchart representing the study selection process for the meta-analysis. Abbreviations: NOS, Newcastle-Ottawa Scale; SNPs, single nucleotide polymorphisms.

F I G U R E 2 Forest plots of ABO blood types and longevity: (A) blood type A and longevity, (B) blood type B and longevity, (C) blood type O and longevity, (D) blood type AB and longevity.
there was overlap among the data, we chose the report with more extensive coverage.

| Data extraction and study quality assessment
The following information was extracted from the selected studies: the first author's name, publication year, country, exposure measures, number of participants, case numbers, and control numbers.
The quality of the studies was assessed with the Newcastle-Ottawa Scale (NOS). With a score ranging from 0 to 9, a score ≥7 indicated a high-quality study.

| Data synthesis
Heterogeneity test was conducted before meta-analysis. Cochran Q and I 2 statistics were used to evaluate heterogeneity. An I 2 > 50% was considered to have severe heterogeneity. 28 If I 2 > 50%, a random effect model was used to combine study individual effect estimates accounting for heterogeneity. Otherwise, the fixed-effect model was selected. 29 A comprehensive meta-analysis was performed (Stata 14.0) to analyze the overall ORs and 95% CIs for the association between ABO blood types and longevity.

| Allelic association analysis
To determine the ABO allelic association with longevity, eight SNPs  (Table 1).

| Genotypic association analysis
A comprehensive analysis of these eight SNPs was executed to determine the ABO genotypic association with longevity. We di-  (Table 2).

| Phenotypic association analysis
We incorporated genotypes that represent the same phenotypes-

| Meta-analysis
To compare our findings with the previously published studies, we performed a meta-analysis. further full-text review. Finally, based on the selection criteria, only five studies were included 16,18,[30][31][32] (Figure 1).

| Data extraction and study quality assessment
All of these five articles were case-control studies with an NOS score ≥ 7. One study only analyzed blood type O and non-O. First author name, publication year, country, exposure measures, number of centenarians, case numbers, control numbers, and NOS scores are listed in Table 4. Data from the current study are also listed.

| Association between ABO blood types and longevity by meta-analysis
Since I 2 > 50% in blood type B versus non-B and O versus non-O groups, the random effect model was used. For blood type A versus non-A and AB versus non-AB, the fixed-effect model was used.
No statistically significant difference between ABO blood types and longevity was shown by forest plots (Figure 2).

| D ISCUSS I ON
In this study, we evaluated the association of ABO alleles, genotypes, and blood types with longevity in our large Chinese centenarian population as well as in previously published datasets and found that genetic variants of ABO genes are not associated with the human longevity trait.
Longevity is a complex trait that is affected by both genetic and environmental factors, including diseases and personalities. [33][34][35][36] Over the past decades, huge efforts have been made to evaluate the genetic contribution to human longevity, leading to the identification of several genes or loci associated with centenarians or exceptionally long-lived individuals through a candidate gene approach or genome-wide association study. [21][22][23]37,38 Human ABO blood groups are genetically determined and have been shown to influence diseases and personalities 39 ; therefore, they could possibly influence lifespan, including longevity.
Our study does not support the genetic association of ABO alleles with human longevity, as no statistical differences were found between centenarians and middle-aged controls even without multiple comparison correction, which is in agreement with the previous studies. 20,30 Our meta-analysis showed a consistent conclusion.
We carefully reviewed the studies that had previously reported a positive association between ABO blood groups and longevity. [15][16][17]20,30 As listed in Table 5, small sample size and statement for longevity are two common problems, which are critical influencers for population-based studies as the population is heterogeneous and stratified by many hidden or unnoticed factors. In this view, a larger population should produce a more robust statistical estimation. In this study, 2201 centenarians and 2330 regionally matched middle-aged individuals as controls were included, presenting the largest population ever for estimating the association between ABO and longevity and providing sufficient statistical power for the statement.
The current study does not debate the association between ABO blood groups and specific diseases and risks, such as myocardial infarction and coronary artery disease, 13,40,41 ischemic stroke, 42 or cancers, 9,43 as previously reported (Table 6). Centenarians represent a model of human healthy aging in contrast to those who suffer from diseases or even death after middle age. That ABO genotypes are associated with diseases, even life-threatening diseases, but not with longevity may imply a notion that disease and longevity are two biological processes with distinct pathways. On the other hand, the two biological processes also share some common pathways. For example, we have shown in our previous study that immune response and inflammation, mitogen-activated protein kinase, sucrose, and xenobiotic metabolism significantly contribute to longevity, 23 but these have also been linked to various diseases and the aging process. Longevity is a multifactorial and polygenic trait, and it has a group of influencers, including risks and diseases, which are intermediate phenotypes that contribute to the longevity trait in a more complicated way than expected.
Since all of our subjects are Han Chinese, our study has a population limitation. It is necessary to validate our findings in other populations. In summary, our study shows that genetic variants of the ABO gene are not associated with the human longevity trait.

CONFLICTS OF INTEREST
Nothing to disclose.