Relationship between maternal ABO blood groups and pregnancy outcomes: a retrospective cohort study in Dongguan, China

Abstract The purpose of this study was to study the relationship between maternal ABO blood groups and pregnancy outcomes. A total of 29,658 couples in Dongguan were selected as the research subjects. We obtained data on ABO blood groups and pregnancy outcomes and explored the relationship between them through log binomial regression and survival analysis. Compared to mothers with type B blood, the RR of foetal stillbirth in mothers with type A blood was 2.87 (95% CI: 1.70, 4.85), and compared to mothers with type O blood, the RR was 1.72 (95% CI: 1.16, 2.55). Compared with foetuses of other three blood type mothers, foetuses of A blood type mothers have a higher median birth weight (P = 0.011). Other pregnancy outcomes, including preterm birth, macrosomia, caesarean section, multiple births, birth defects, low birth weight, foetal sex, gestational days, birth length, and APGAR score, were not significantly different. The relationship between maternal ABO blood type and pregnancy outcomes was not affected by paternal blood type. More studies are needed to confirm these results. IMPACT STATEMENT What is already known on this subject? The relationship between blood type and disease is being increasingly studied. With regard to the relationship between maternal blood type and pregnancy outcomes, some studies have focused on people undergoing in vitro fertilisation. There are few reports on healthy women. What do the results of this study add? Compared to mothers with type B blood, the RR of foetal stillbirth in mothers with type A blood was 2.87 (95% CI: 1.70, 4.85), and compared to mothers with type O blood, the RR was 1.72 (95% CI: 1.16, 2.55). Compared with foetuses of other three blood type mothers, foetuses of A blood type mothers have a higher median birth weight (P = 0.011). What are the implications of these findings for clinical practice and/or further research? This study is the first to explore the relationship between blood type and pregnancy outcomes in healthy women.These results can provide some clues for the study of the mechanism of pregnancy outcomes.


Introduction
The ABO blood groups, discovered by Karl Landsteiner in 1900, include four major 'ABO' phenotypes-'A', 'B', 'O' and 'AB' (Huang et al. 2017).ABO blood types are the most important blood groups in medicine and are defined by the carbohydrate moiety displayed on the surface of red blood cells and attached to the precursor H substance (Huang et al. 2017).In addition to red blood cells, blood group antigens can also be found on white blood cells, some tissues, plasma proteins, platelets and various cell surface enzymes, and even in body fluids (Abegaz 2021).Correlations between blood type and some diseases may exist because of the presence of blood group antigens.In 1953, it was first reported that there is a statistically significant association between blood type A and the risk of gastric cancer (Aird et al. 1953).Since then, research on the relationship between blood types and diseases has been continuous.Various studies on the correlation between specific blood types and various infectious and non-infectious diseases have been reported (Chen et al. 2016;Abegaz 2021).
The relevance of blood types and reproductive wellbeing has been studied.The reviews by Qiu-He Chen et al. (Chen et al. 2022) and Franchini et al. (Franchini et al. 2016) suggested that pregnant women without blood type O have an increased incidence of venous thromboembolism compared with pregnant women with blood type O. Binder (Binder et al. 2008) described an association between blood type A and early-onset ovarian hyperstimulation syndrome.Some studies have demonstrated that blood type A is significantly associated with the risk of ovarian cancer (Poole et al. 2012;Cozzi et al. 2017).Regarding the relationship between CONTACT Wei-Chao He mailforyvonne@163.comDongguan Maternal and Child Health Care Hospital, No. 99 Zhenxing Road, Dongcheng District, Dongguan City, Guangdong Province, 523112, China.
Supplemental data for this article can be accessed online at https://doi.org/10.1080/01443615.2023.2243508.
maternal blood type and pregnancy outcomes, some studies have focused on the population undergoing in vitro fertilisation (IVF) (Goldsammler et al. 2015;Pereira et al. 2017;Di Nisio et al. 2018;Zhao et al. 2021), while there are few reports on healthy women.In this research, we explored the relationship between maternal blood types and pregnancy outcomes through a cohort study using regression analysis.We hope that our findings can provide clues for basic research on adverse pregnancy outcomes.Mothers undergo B-ultrasound examinations of the reproductive system.There are special funds for the follow-up of pregnancy outcomes for at least one year.Since October 2012, the Dongguan government has provided the NFPHEP for women of childbearing age.This retrospective cohort study included couples who participated in the NFPHEP in Dongguan from October 2012 to June 2019 as the study subjects.These couples had completed all the items of the NFPHEP, including basic situation investigations, physical examinations, laboratory examinations, gynecological examinations and health education.They were followed by medical staff until pregnancy results were entered into the system.For more information about the NFPHEP in Dongguan, we referred to two articles published previously (Jiang et al. 2020(Jiang et al. , 2022)).We divided the subjects into four groups according to their blood types and then compared the differences in various pregnancy outcome rates among the four groups.

Definition of pregnancy outcomes
All outcomes included live births, stillbirths and legal pregnancy terminations from 28 weeks of gestation to 7 days postpartum.In particular, for couples who had twin or multiple births, if all babies were delivered normally, normal delivery was considered, and if one or more babies were delivered abnormally, adverse pregnancy outcomes were considered.There were 10 indicators of pregnancy outcomes, including preterm birth (delivery with a gestation period of less than 37 completed weeks), early preterm birth (delivery with a gestation period of less than 34 completed weeks), foetal macrosomia (newborn weight more than 4000 g), low birth weight (newborn birth weight less than 2500 g), neonatal APGAR score, stillbirth, caesarean delivery, multiple births, birth defects and infant sex.

Data collection and quality control
Special funds were provided to track the pregnancy outcomes of the included women.Responsible staff were sent to collect and record the information.The government inspected and examined the NFPHEP system every half year at the county level to prevent missing reports.The diagnosis of adverse pregnancy outcomes was made by professional doctors in county or higher-level medical institutions.In addition, the Chinese government established a full set of birth defect surveillance systems covering the national, provincial, and regional levels in 1986.To ensure the accuracy of pregnancy outcomes, we compared the NFPHEP system with the Dongguan birth defects surveillance system.

Statistical analysis
The incidence of pregnancy outcomes was calculated as the number of cases with pregnancy outcomes divided by the number of births, including foetuses or neonates, or pregnancy termination from 28 weeks of gestation to 7 days postpartum.The chi-square test was used to compare the baseline characteristics among the different ABO blood groups.A log-binomial regression model was used to estimate the relative risk (RR) and its 95% confidence interval for pregnancy outcome incidence in relation to ABO blood type.We used survival analysis to compare the differences among the groups for the measurement data of gestation days, birth weight, birth length and APGAR score.Caesarean birth was defined as deletion, and natural delivery was defined as an end-point event.In addition, we also explored the influence of parental blood group combinations on some pregnancy outcomes.All statistical analyses were performed using SAS software (version 9.3).The forest map was drawn by Excel (version 2007).Statistical significance was defined with a two-tailed threshold of 0.05.

Results
This study ultimately included 29,658 couples who had complete examination data and clear pregnancy outcomes (Figure 1).Han people accounted for 99.6% of the sample.The proportions of the paternal and maternal blood types were similar: 26% for type A, 27% for type B, 7% for type AB and 40% for type O.The average maternal age was29.34yearsold (SD 4.54) with a range of 18-49 years, and the average paternal age was 30.34 years old (SD 5.02) with a range of 21-56 years.We compared the baseline characteristics of the participants according to ABO blood group, including the demographic data and behaviour habits of the mothers and fathers, as well as female reproductive system diseases.We found that the comparison of various data did not show statistical significance.The results are shown in Table 1.
Of 29,658 infants, 1,716 (5.8%) were preterm, 350 (1.2%) were early preterm, 585 (2.0%) had foetal macrosomia, 1,557 (5.2%) had low birth weight, 220 (1.1%) had an APGAR score �7, 132 (0.4%) were stillborn, 10,533 (35.5%) were born by cesarian section, 400 (1.3%) were from multiple pregnancies, 583(2.0%)had birth defects, and 15,616(52.8%)were male.We used the chi-square test to compare the distribution of pregnancy outcomes in the different blood groups.The results indicate that the incidence of stillbirth in foetuses of mothers with type A blood is higher than that of foetuses of mothers with other blood types, and no significant difference was found in the other pregnancy outcomes (Table 2).
In Figure 2 and Supplementary materials 1, we divided 9 scenarios and used a log-binomial regression model to explore the influence of different maternal blood types on pregnancy outcomes.We found that type A blood was a risk factor for stillbirth and that type B blood was a protective factor.Compared to mothers with type B blood, the RR of foetal stillbirth in mothers with type A blood was 2.87 (95% CI: 1.70, 4.85).Compared to mothers with type O blood, the RR of foetal stillbirth in mothers with type A blood was 1.72 (95% CI: 1.16, 2.55).There was no significant difference between foetuses with blood type B and blood type O, and the RR value was 0.60 (95% CI: 0.35, 1.02).We also found that the incidence of preterm birth, early preterm birth, macrosomia, birth defects and other incidence rates were high in foetuses from mothers with type A blood, but the difference was not statistically significant.
We used survival analysis to compare four quantitative variables.The results showed that there were significant differences in birth weight (P ¼ 0.011) but no significant differences in gestational days (P ¼ 0.400), birth length (P ¼ 0.094) or APGAR scores (P ¼ 0.210).The results are shown in Supplementary materials 2. Compared to foetuses of mothers with the other three blood types, foetuses of mothers with type A blood have a higher median birth weight (P ¼ 0.011).In Table 3, we divided the patients into 16 groups for comparison by matching parents' blood types.The results showed that birth defects and preterm birth had no relationship with the father's blood type.The stillbirth rate of foetuses born to mothers with type A blood, regardless of the father's blood type, was higher.

Discussion
There are currently dozens of risk factors for adverse pregnancy outcomes reported in research, which can be divided into intrauterine factors and extrauterine factors (Della Rosa et al. 2021).To control for the impact of these confounding factors on pregnancy outcomes, we adjusted for over forty variables of the mothers and fathers in subsequent regression analysis.As shown in Table 1, we found that there was no significant difference in the baseline characteristics of participants in the four blood groups, which indicated that the interference factors were evenly distributed among the groups.Therefore, in the later regression analysis, we did not need to adjust for baseline characteristics.This finding seemed inconsistent with previous literature reports.For example, people with type A blood are more likely to suffer from hepatitis B and vascular diseases (Abegaz 2021).Our research focused on the reproductive-age population, excluding elderly individuals and minors, which may be an important reason for our inconsistent results.
The roles of ABO blood type antigens in the development of malignancies, chronic diseases and infectious diseases have been explored for decades.However, to the best of our knowledge, the correlation between blood type and pregnancy outcomes is rarely reported in the natural population.In this study, the proportions of the different blood type populations were similar to that estimated by Huang (Sun et al. 2015) for the Chinese.We analysed the relationships between 14 pregnancy outcomes and maternal blood types, including 10 qualitative variables and 4 quantitative variables.
Among the four blood types, the incidence of preterm birth, early preterm birth and macrosomia was higher in foetuses of participants with type A blood, but the difference was not significant.Survival analysis showed that the median birth weight of foetuses born to mothers with type A blood was higher than that of foetuses with the other blood types.There was no significant difference in the gestation period among the four groups.A study on single embryo transfer and pregnancy outcomes showed no difference in foetal birth weight and gestational age among mothers with different blood types (Pereira et al. 2017).However, a prospective hospital-based study by Poudyal found that maternal type A blood was a risk factor for premature birth (Poudyal et al. 2018).The influence of blood type on gestation period and birth weight needs further study.Among the ten qualitative variables, stillbirth was the only one with a significant difference.The incidence of stillbirth was the highest in the mothers with type A blood group, followed by the type AB, type O, and type B blood groups.Michelle's investigation in the IVF population also found that the live birth rate of foetuses from mothers with type B blood was the highest, which was consistent with our results (Goldsammler et al. 2015).The incidence of birth defects in foetuses of mothers with type A blood may be higher than that of foetuses of mothers with the other blood types, but the difference is not significant.Due to the sample size limitation, we could not perform analyses according to specific defect types.If the sample size is expanded, more meaningful results may be obtained.At present, research has found that infant blood type may be  related to birth defects.A study analysed the blood types of 1,785 premature infants and found that compared with that of infants without type A blood, the incidence rate of patent ductus arteriosus (PDA) and bronchopulmonary dysplasia (BPD) was significantly higher in infants with type A blood (Cakir et al. 2020).We also analysed several indicators, such as the caesarean section rate, infant sex, multiple pregnancy rate, birth length and APGAR score, but there was no difference.Some studies have explored the effects of paternal occupational exposure, cancer treatment, drug exposure and sperm genetics on pregnancy outcomes (Engeland et al. 2013).Therefore, we analysed paternal baseline characteristics to exclude the influence of these factors on the results and ultimately found that the distribution of paternal data was balanced among the groups.A study suggested that in the population undergoing assisted reproductive technology (ART), the combination of paternal and maternal blood types will affect the occurrence of biochemical pregnancies, clinical pregnancies and live births, which may be related to the incompatibility of ABO blood types (Bao et al. 2022).We explored the influence of parents' blood group combinations on premature birth, stillbirth and birth defects and obtained the same results as when the maternal blood group was analysed separately.That is, mothers with type A blood, regardless of the father's blood type, have a higher probability of foetal stillbirth than those with the other blood types.The pregnancy outcomes of the foetus seemed to be independent of the father's blood type.
It is difficult to understand the underlying mechanisms of the relationship between these ABO genes and pregnancy outcomes due to the limited number of relevant studies.However, previous studies can provide some clues.First, blood-type antigens are involved in cell signalling, cell recognition, and cell adhesion (Ewald and Sumner 2016).The epithelial tissues of female reproductive organs such as the uterus and ovaries all contain blood group antigens (Yamamoto et al. 2012), which may affect the occurrence and development of these organ diseases, thereby affecting pregnancy outcomes (Cakir et al. 2020).Second, the current study found that the risk of some diseases and infections is related to blood type (Ewald and Sumner 2016).This infection may have adverse effects on the health of pregnant women, which in turn can affect pregnancy outcomes.Third, some studies have found that some ABO blood group gene polymorphisms are related to elevated levels of some immune and inflammatory mediators, which are related to early embryo implantation and subsequent placental implantation (Zhao et al. 2021).In addition, scientists have examined possible associations between blood type and personality traits and have obtained some positive results (Tsuchimine et al. 2015), while recent prospective cohort studies have reported that the maternal psychological condition has an impact on several adverse obstetric, foetal and neonatal outcomes (Dayan et al. 2010).Our study found that blood type A is a risk factor for stillbirth, which may be the result of the combined effect of the above four factors.Of course, these are only preliminary explorations of the underlying mechanisms behind this association in this study, and our findings are   worth further investigation through appropriately designed experiments.
In conclusion, this was a rare study analysing the association between blood types and pregnancy outcomes.The large sample size and the selection of natural populations made the study representative.Maternal ABO blood groups may be related to some specific pregnancy outcomes for infants.These research results can provide clues and ideas for the study of the causes of adverse pregnancy outcomes.However, our research has some limitations.Firstly, the study subjects were from a population of individuals participating in NFPHEP at a single centre, which may not represent the entire pregnant population, and there is a lack of literature support.Therefore, these results should be considered preliminary and interpreted with caution.Second, although we compared more than 30 baseline characteristic data points, there was still a lack of data on important potential confounding factors, such as placental conditions, foetal monitoring, amniotic fluid status, and labour processes.Finally, we did not collect data regarding the ABO blood types of the babies, and ABO blood type incompatibility between the foetus and the mother can lead to haemolytic diseases in the foetus.Further studies are required to enhance these results.

Figure 1 .
Figure 1.Flow chart of this study.

Figure 2 .
Figure 2. Forest chart for comparison of the pregnancy outcomes of women with different blood types.
Since April 2010, the Chinese government has implemented the National Free Preconception Health Examination Project (NFPHEP) to reduce the number of infants with birth defects and improve the quality of the population.At present, all provinces in mainland China have implemented the NFPHEP.Millions of couples benefit from the project every year.The inspection items of fathers and mothers include questionnaire investigations, clinical tests, and physical examinations.

Table 1 .
Basic information of participants.

Table 1 .
Continued.The figures in between the brackets are percentage; The Han population accounts for 99.6%; EDC -endocrine disruptor; There was no significant difference in baseline data among the four groups using chi-square test (P > 0.05).

Table 2 .
Pregnancy outcomes of patients with different blood types.
# Some infants did not have APGAR scores; � P ＜ 0.05; The figures in between the brackets are percentage.

Table 3 .
Effect of parental blood type matching on pregnancy outcomes.