Prevalence, risk factors, and adverse maternal and fetal outcomes of anemia during pregnancy: a retrospective study in China, 2012–2020

Abstract

Background

Anemia is a public world health problem among pregnant females especially in developing countries. China is a representatively developing country with vast territory and significant economic and cultural differences in different regions. Currently most studies on anemia in pregnancy ignored economically backward and resource-poor cities or districts and anemia during pregnancy in China was not accurately estimated. The goal of this study was to assess the prevalence, risk factors, and adverse outcomes of anemia among pregnant females across different regions of China using a large, nationally representative dataset from 2012 to 2020.

Method

A large representative multicenter retrospective study collected data from the National Maternal Near Miss Surveillance System (NMNMSS). The study covered 30 provinces (Tibet excluded), eastern, central, and western regions, both rural and urban areas, as well as medical institutions at all levels. A total of 438 medical institutions participated, with 12, 572, 647 singleton pregnant women and 237, 864 multiple pregnant women enrolled. Socio-demographic characteristics, hemoglobin levels, risk factors and pregnancy outcomes were collected via the online NMNMSS platform from all the participants and the study showed the current real situation of anemia during pregnancy in China during 2012–2020.

Results

A total of 12,810,511 pregnant females aged 15 to 49 years were included. The overall prevalence of pregnancy-associated anemia was 21.25%, with 8.44% mild anemia, 12.03% moderate, 0.58% severe, and 0.20% of unknown severity. The anemia prevalence was 21.08% among singleton pregnancies (1,063,104 mild, 1,492,873 moderate, 69,402 severe, 24,922 unknown severity) and 30.31% among multiple pregnancies (18,156 mild, 48,686 moderate, 4,710 severe, 24,922 unknown severity). Notably, moderate anemia accounted for the largest proportion of anemic cases across all pregnancies. Compared with the singleton pregnancy group, the multiple pregnancy group had significantly higher risks of moderate and severe anemia, which were 1.7-fold and 10-fold higher, respectively. Additionally, gestational anemia was significantly associated with maternal age 15–19 years, parity, previous caesarean sections, education status and most maternal complications during pregnancy. Moderate and severe anemia exhibited a significant proportional association with maternal outcomes, including uterine rupture, placenta previa, placental abruption, postpartum hemorrhage, severe postpartum hemorrhage, blood transfusion, intensive care unit (ICU) admission and multiple organ dysfunction syndrome (MODS), and these conditions were further linked to an elevated risk of low birth weight, poor Apgar scores at birth, and stillbirth. No significant elevation in adverse outcome risk was observed in instances of mild anemia.

Conclusion

Based on large-sample nationwide data of 12.8 million pregnant women in China, this study identified a substantial anemia burden (prevalence: 21.25%), with moderate/severe anemia (12.61%) significantly associated with elevated clinical risks. Multiple pregnancy, as a major risk factor for pregnancy-associated anemia, was associated with a higher anemia prevalence (30.31%) and significantly increased risks of moderate and severe anemia compared with singleton pregnancy (21.08%), with aOR = 1.64 (95% CI: 1.52–1.76). Other risk factors included adolescent pregnancy (15–19 years), low education and obstetric complications. Notably, moderate/severe anemia was associated with heightened maternal and neonatal adverse outcomes, whereas mild anemia exhibited no such associations of risk-stratified screening and preventive strategies to reduce the morbidity and adverse outcomes potentially associated with pregnancy-associated anemmia.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12884-026-09159-1.

Background

Anemia during pregnancy remains an important public health issue and is the most common nutritional deficiency disorder worldwide [1]. Because of fetal growth and increased serum iron demand, blood volume, and maternal nutritional requirements, pregnant women are susceptible to anemia [2, 3]. Anemia during pregnancy is associated with a higher risk of cesarean delivery, maternal death, preterm birth, preeclampsia, low birth weight, and perinatal death [4, 5]. It threatens the health of the pregnant woman and fetus and increases the morbidity rate throughout the entire life cycle [6–8].

According to the World Health Organization (WHO) reported, anemia affects approximately 40% of pregnant women worldwide. The prevalence of anemia in high-income countries was about 9%, whereas the highest prevalence of 40%-57% is seen in West African, Middle Eastern, and South Asian countries [9]. As a developing country, the burden of anemia during pregnancy is high in China, with a prevalence ranging from 11.1% to 40% according to different studies [4, 10]. There is a significant gap in the prevalence of anemia in pregnant women between China and developed countries. China has a vast territory and significant economic and cultural differences across different regions. It is challenging to conduct an objective survey on the prevalence of anemia in pregnant women covering all regions. For example, most studies on anemia during pregnancy were conducted in regions characterized by easy access to antenatal care or tertiary hospitals [10, 11], whereas cities or districts with relatively backward-economies and scarce-resources were not included. Therefore, a large-sample study with comprehensive coverage incorporating all regions is lacking.

Previous studies have reported that the prevalence of anemia during pregnancy varies among women with different socio-economic conditions, lifestyles, and health-seeking behaviors in different cultures. Specifically, a higher prevalence of anemia was found among pregnant women with a lower family per-capita family income [11], inaccessibility of healthcare facilities, low pre-pregnancy body mass index (< 18.5 kg/m²) [12], low education level, or multiparity [13, 14]. In South Africa, common risk factors included human immunodeficiency virus (HIV) and other infections [15]. However, with the rapid development of China’s economy and recent reform of China’s universal two-child policy in recent years, the pace of people’s lives has accelerated and lifestyles have changed, and the childbearing age in the entire female population has been delayed [16, 17]. This may affect the prevalence of anemia during pregnancy. The lack of up-to-date information on the risk factors for anemia presents a major obstacle to developing evidence-based recommendations for managing anemia during pregnancy.

Although several small-sample studies have reported adverse maternal and perinatal outcomes associated with anemia during pregnancy [18–20], the findings are controversial and the relationships vary with the severity of anemia in different small-sample studies, especially the relationships between maternal anemia severity and rare adverse outcomes [21]. Therefore, large-sample data are required. Additionally, most current data predominantly focus on the relationship between singleton pregnancies and anemia during pregnancy. Data on anemia in twin or multiple pregnancies remain limited.

Therefore, in this study we used the National Maternal Near Miss Surveillance System (NMNMSS) in China, which covers a large population base, a wide geographical distribution, and samples from different institutional levels. Our focus was on the prevalence and risk factors of anemia during pregnancy as well as the association between the severity of anemia during pregnancy and risk of adverse maternal and fetal outcomes, including postpartum hemorrhage, cesarean section, preterm birth, low birth weight, blood transfusion, admission to intensive care unit, and stillbirth. Therefore, we designed a large-sample, multi-center, cross-sectional, retrospective cohort analysis of singleton and multiple pregnant women in China. Our study aimed to investigate the current status of anemia during pregnancy in China and to improve the anemia status of pregnant women by providing more representative, timely, up-to-date, high-quality evidence.

Method

Data resource and study participants

Data were extracted from the NMNMSS between January 2012 and December 2020. The NMNMSS covers 438 medical institutions in 30 provinces (except Tibet) in mainland China, of which 215 are in urban areas, and 223 in rural areas. There are 109 medical institutions in Eastern China, 173 in Central China, and 156 in Western China. Each participating facility delivered more than 1000 babies annually, of which 142 were tertiary hospitals, 234 were secondary hospitals, and 62 were below class II hospitals. The grade of hospitals in China is based on the hospital’s scale, service provision, medical technology, equipment, medical research, and so on.

Data for this retrospective cohort study were derived from the hospital’s electronic medical records. The source data comprised sociodemographic and obstetric information that was prospectively recorded for all pregnant and postpartum women during their inpatient stay. Obstetricians were responsible for completing the maternal questionnaire based on inpatient medical records via the online NMNMSS platform. An associate director of the obstetric department verified and reviewed the data at the primary level, investigated missing reports, and aggregated the data and reports into the NMNMSS. Data quality was monitored and verified through stratified random sampling at the secondary, tertiary, and quaternary levels in county-, prefecture-, and provincial-level maternal and child healthcare hospitals. Finally, the National Office for Maternal and Child Health Surveillance System is responsible for data quality control at the quinary level. Incomplete data were supplemented until requirements were met. All the data we obtained were de-identified.

As is shown in Fig. 1, the initial cohort consisted of 13,473,704 females with pregnancy outcomes during the study period (2012–2020) recorded in the National Maternal Near-Miss Surveillance System (NMNMSS). Participants were excluded if they had missing data regarding anemia (n = 25,102), were aged 15 years or younger (n = 2,647), or had unknown numbers of fetuses (n = 635,444), resulting in 12,572,647 singleton pregnancies and 237,864 multiple pregnancies included in the epidemiological and risk factor analysis. Participants aged 15 years or younger were excluded from the primary analysis to ensure a more homogeneous adult study population, as the physiological and clinical characteristics of pregnancy in this youngest adolescent group differ substantially from those of older adolescents and adults, with potential impacts on analysis reliability and ethical compliance. The multiple pregnancies included 235,261 twin, 2,504 triplet, and 99 quadruple pregnancies.

Fig. 1.

Flowchart of study participants

Further exclusions for the pregnancy outcome analysis included spontaneous abortion in early pregnancy, ectopic pregnancy, induction of labor due to fetal malformation before 28 weeks gestation, second trimester abortion and pregnancies at ≥ 28 weeks of gestation that were not terminated during the study period. The final dataset for pregnancy outcome analysis comprised 11,977,684 singleton pregnancies and 221,444 multiple pregnancies.

Definition of anemia

The diagnostic criteria for anemia in pregnancy varied between 2012 and 2020, and prenatal anemia was defined as a hemoglobin concentration below 100 g/L before 2014 and below 110 g/L after 2014. However, for the purpose of ensuring a consistent diagnostic standard across the entire study cohort in this analysis, prenatal anemia was uniformly defined as a hemoglobin (Hb) concentration below 110 g/L for all participants, regardless of the year of admission. Anemia during pregnancy (excluding the puerperium) was classified into three categories according to the relevant International Statistical Classification of Diseases and Related Health Problems, 10th Edition, (ICD-10) codes: (i) mild anemia (hemoglobin concentration, 100–109 g/L), (ii) moderate anemia (hemoglobin concentration, 70–99 g/L), and (iii) severe anemia (hemoglobin concentration, < 70 g/L) [22]. Pregnant women with missing information on hemoglobin concentrations were classified as having anemia of unknown severity. This approach was adopted to maintain epidemiological rigor and full transparency. The lowest hemoglobin concentration in each pregnant woman before delivery and after admission was used to define the degree of anemia during pregnancy.

Risk factors

Maternal age, gravity, parity, number of antenatal visits, previous caesarean sections, educational level, and maternal complications during hospitalization (including hypertensive disorders, diabetes, heart diseases, embolism, liver diseases, kidney diseases, lung diseases, HIV infection, connective tissue diseases, and cancer diagnosed before pregnancy or during pregnancy) were listed as risk factors to be explored in this study. Complications during pregnancy, which might constitute confounders for the association between maternal anemia (exposure) and adverse maternal and fetal outcomes (endpoint), were identified as confounding variables if the time sequence relative to the onset of anemia was unknown. Of course, scarred uterus, placenta accreta spectrum, placenta previa, and abnormal placenta were not listed as complications but were used as confounding elements for adjusted odds ratios (OR). All diagnoses were coded using the ICD-10. Procedures and surgeries were coded by certified professional medical coders in each member hospital using the ninth Clinical Version of the International Classification of Diseases (ICD-9-CM).

Maternal and fetal outcomes

Maternal outcomes identified using ICD-10 codes included cesarean section, uterine rupture, placental abruption, postpartum hemorrhage, blood transfusion, preterm birth (delivery time between 28^+ 0 and 36^+ 6 gestational weeks), admission to the intensive care unit (ICU), shock, multiple organ dysfunction syndrome (MODS; defined as acute and potentially reversible dysfunction of two or more organ systems triggered by multiple different, clinically-diverse factors), and maternal death during hospitalization.

For fetal outcome analysis, the estimated variables were low birth weight, poor birth condition with low Apgar score at 1–5 min (we divided the births into two categories: i) Apgar score < 7 points at 1 min or < 7 points at 5 min and ii) Apgar score < 3 points at 1 min or < 5 points at 5 min; the second was considered the worse status) [23], and stillbirth (death before or during delivery after 28 weeks of pregnancy). Based on previous evidence and clinical knowledge, relevant maternal complications were considered as potential confounding factors for fetal outcomes. Postpartum hemorrhage was not included as a confounder because it has been documented to have minimal direct impact on newborn outcomes. Notably, intermediate variables (e.g., blood transfusion) were not treated as confounders, so that the total effect of anemia on fetal outcomes could be estimated.

Statistical analysis

Descriptive statistics were calculated for all variables, and categorical variables were expressed as numbers and proportions.

To analyze the epidemiological characteristics of anemia during pregnancy, we estimated the overall prevalence of anemia during pregnancy in the past 9 years and calculated proportions to describe the distributions in different geographic regions and hospital levels. We also examined the incidence of anemia of various severities at different maternal gestational ages. We first performed univariate and multivariate logistic regression analyses to identify factors associated with maternal anemia (as the outcome).

Furthermore, we explored the effect of anemia on maternal and fetal outcomes from two perspectives using multivariate logistic regression models. First, we determined the risk of anemia during pregnancy. Second, we estimated the risk of different classifications of anemia. Non-anemic pregnant women were used as the reference group in all analyses. Maternal age, geographical region, gravity, parity, history of caesarean section, maternal education level, and complications were included as covariates. The time sequence for these complications relative to the onset of anemia was unclear but might have been associated with the outcomes. All analyses were performed using STATA software, version 17.0 (StataCorp, College Station, TX, USA). Statistical significance was set at P < 0.05.

To address the potential confounding effect of the change in clinical diagnostic cutoffs implemented in 2014, sensitivity analyses were conducted by stratifying the dataset into two distinct periods: 2012–2013 and 2015–2020. This approach allowed for an assessment of temporal trends and the robustness of the primary findings by minimizing bias attributable to the alteration in diagnostic criteria.

Results

Prevalence of anemia during pregnancy and distribution by maternal age and region

As shown in Fig. 2A, a total of 12,810,511 pregnant women were enrolled in our study, with the overall prevalence of pregnancy-associated anemia of 21.25%. Stratified by pregnancy type, the anemia prevalence was 21.08% among 12,572,647 singleton pregnancies; among these anemic cases, 1,063,104 (8.46%) were classified as mild anemia, 1,492,873 (11.87%) as moderate anemia, 69,402 (0.55%) as severe anemia, and 24,922 (0.2%) as anemia of unknown severity. In contrast, 237,864 women with multiple pregnancies were included, with an overall prevalence of anemia of 30.31%. The severity distribution of anemia in this group was as follows: 18,156 women (7.63%) with mild anemia, 48,686 (20.47%) with moderate anemia, 4,710 (1.98%) with severe anemia, and 24,922 (0.2%) with anemia of unknown severity. Notably, moderate anemia accounted for the largest proportion of anemic cases across all pregnancies. Compared with the singleton pregnancy group, the multiple pregnancy group had significantly higher risks of moderate and severe anemia, which were 1.7-fold and 10-fold higher, respectively.

Fig. 2.

Incidence, severity of anemia and the association with maternal age in pregnancy. A The overall prevalence of anemia in singleton and multiple pregnancies. B The estimated probability of anemia singleton and multiple pregnancies at the specific age (15-19y, 20-24y, 25-29y, 30-34y, 35-39y, >=40y). C, D&E. In different age groups the proportion of mild, moderate and severe anemia in single and multiple pregnancies.

The overall relationship between the prevalence of anemia during pregnancy and maternal age exhibited a hook-shaped pattern, consistent with findings in singleton pregnancies. To further quantify this nonlinear relationship, we fitted the trend between maternal age and anemia rates using a quadratic model. Based on the median age of each age group and corresponding anemia rates, the quadratic model fitting results showed excellent goodness of fit (R² = 0.982, P < 0.001), with the fitting formula [Y = 35.62–1.02X + 0.018X^2] (Y: Anemia rate (%), X: Maternal median age). The coefficients of the model (coefficient of Age: -1.02, coefficient of Age²: +0.018) further confirmed the hook-shaped nonlinear association. As illustrated in Fig. 2B, the proportion of anemia in singleton pregnancies decreased from 21.54% at 15–19 years to 19.97% at 20–24 years, followed by a gradual increase to 20.76% at 25–29 years, 22.09% at 30–34 years, 23.32% at 35–39 years, and 23.40% among aged 40 years and older. In multiple pregnancies, the prevalence of anemia was 36.80%, 32.99%, 31.09%, 30.03%, 29.74%, and 28.17% across the same age groups (15–19, 20–24, 25–29, 30–34, 35–39, and ≥ 40 years, respectively), showing a steady downward trend.

Furthermore, the proportion of women with mild anemia in each age group was higher among singleton pregnancies than among multiple pregnancies (Fig. 2C). By contrast, the prevalence of moderate and severe anemia was greater in multiple pregnancies compared with singleton pregnancies, with the between-group difference being most prominent among women aged 15–29 years (Fig. 2D and E).

The distribution of enrolled medical institutions across eastern, western and central China was presented in Fig. 3A. The prevalence of anemia was 19.56% in eastern China, while higher prevalence rates (21.92% and 21.94%, respectively) were observed in central and western China (Fig. 3B). The regional distributions of anemia during singleton and multiple pregnancies are further illustrated in Fig. 3C and D, respectively.

Fig. 3.

Incidence and severity of anemia during pregnancy in different regions. A The total prevalence of anemia during pregnancies in different regions. B The prevalence of anemia during singleton pregnancies in different regions. C The prevalence of anemia during multiple pregnancies in different regions of China. (The red marks indicate mild anemia, yellow indicates moderate anemia and blue indicates severe anemia.)

Risk factors associated with anemia during pregnancy

Table 1 shows the baseline maternal characteristics in anemic and non-anemic groups. After adjustment for potential confounders, multiple pregnancy was associated with a significantly higher risk of anemia compared with singleton pregnancy (adjusted odds ratio [aOR] = 1.64, 95% confidence interval [CI]: 1.52–1.76). With regard to maternal age, adolescent mothers aged 15–19 years exhibited a 19% higher risk of anemia relative to the reference group aged 25–29 years (aOR = 1.19, 95% CI: 1.11–1.28). No significant associations were detected for other age groups. For gravidity, women with 3–5 previous pregnancies showed a slightly elevated anemia risk (aOR = 1.05, 95% CI: 1.04–1.09), whereas those with more than five pregnancies demonstrated a non-significant trend (aOR = 1.06, 95% CI: 0.99–1.14). A progressive increase in anemia risk was observed with rising parity. Compared with nulliparous women, the aORs were 1.12 (95% CI: 1.07–1.17) for one delivery, 1.26 (95% CI: 1.15–1.37) for two deliveries, 1.41 (95% CI: 1.24–1.61) for three deliveries, and 1.29 (95% CI: 1.10–1.52) for three or more deliveries. A history of cesarean section was also significantly associated with increased anemia risk in a graded manner: aOR = 1.40 (95% CI: 1.34–1.47) for one previous cesarean section, 1.62 (95% CI: 1.48–1.78) for two, 1.51 (95% CI: 1.22–1.86) for three, and 1.48 (95% CI: 1.09–2.01) for four or more cesarean sections. As shown in Supplementary Table 1, this association was mainly present in singleton pregnancies and not observed in multiple pregnancies. Higher educational attainment was associated with a lower risk of anemia. Compared with the reference group (illiterate women), women with a senior high school or technical secondary school diploma (aOR = 0.86, 95% CI: 0.73–1.00) and those with a college degree or higher (aOR = 0.79, 95% CI: 0.0.67–0.930.81) had significantly reduced odds of anemia during pregnancy. Except for heart disease and connective tissue disease, most maternal medical complications were significantly associated with an increased risk of anemia. These included hypertensive disorders (aOR = 1.35, 95% CI: 1.28–1.42), diabetes mellitus (aOR = 1.10, 95% CI: 1.02–1.18), embolism (aOR = 3.20, 95% CI: 2.80–3.66), liver disease (aOR = 1.40, 95% CI: 1.22–1.60), renal disease (aOR = 1.44, 95% CI: 1.15–1.80), lung disease (aOR = 1.79, 95% CI: 1.47–2.17), HIV infection (aOR = 1.33, 95% CI: 1.06–1.68), and cancer (aOR = 1.33, 95% CI: 1.14–1.55).

Table 1.

Risk factors associated with anemia during pregnancy

Variables	Nonanemic (n = 10,550,069)	Anemia (n = 2,895,886)	Crude OR (95%CI)	Adjusted OR (95%CI)
Maternal age, year
15–19	193,285 (1.83%)	53,462 (1.85%)	1.10 (0.99 ~ 1.21)	1.19 (1.11 ~ 1.28)
20–24	1,620,924 (15.36%)	408,046 (14.09%)	0.96 (0.91 ~ 1.02)	1.00 (0.97 ~ 1.34)
25–29	3,969,717 (37.63%)	1,049,136 (36.23%)	ref.	ref.
30–34	2,690,097 (25.50%)	770,228 (26.60%)	1.10 (1.05 ~ 1.13)	0.98 (0.95 ~ 1.01)
35–39	1,022,874 (9.70%)	313,245 (10.82%)	1.18 (1.12 ~ 1.23)	0.97 (0.92 ~ 1.01)
≥ 40	261,131 (2.48%)	79,904 (2.76%)	1.22 (1.15 ~ 1.29)	0.97 (0.91 ~ 1.03)
Unknown	792,041 (7.51%)	221,865 (7.66%)	0.42 (0.34 ~ 0.51)	0.41 (0.33 ~ 0.50)
Gravidity
< 3	7,268,067 (68.89%)	1,860,969 (64.26%)	ref.	ref.
3–5	3,072,853 (29.13%)	970,622 (33.52%)	1.26 (1.21 ~ 1.31)	1.05 (1.04 ~ 1.09)
> 5	190,659 (1.81%)	63,218 (2.18%)	1.37 (1.28 ~ 1.47)	1.06 (0.99 ~ 1.14)
Unknown	18,490 (0.18%)	1,077 (0.04%)	0.20 (0.05 ~ 0.75)	0.63 (0.35 ~ 1.14)
Parity
0	5,828,864 (55.25%)	1,411,553 (48.74%)	ref.	ref.
1	3,962,135 (37.56%)	1,225,510 (42.32%)	1.30 (1.23 ~ 1.37)	1.12 (1.07 ~ 1.17)
2	630,873 (5.98%)	217,442 (7.51%)	1.51 (1.37 ~ 1.66)	1.26 (1.15 ~ 1.37)
3	83,734 (0.79%)	31,300 (1.08%)	1.61 (1.39 ~ 1.87)	1.41 (1.24 ~ 1.61)
> 3	27,507 (0.26%)	9,301 (0.32%)	1.40 (1.18 ~ 1.66)	1.29 (1.10 ~ 1.52)
Unknown	16,956 (0.16%)	780 (0.03%)	0.15 (0.04 ~ 0.60)	1.11 (0.60 ~ 2.06)
Antenatal visits
0	251,721 (2.39%)	37,869 (1.31%)	0.68 (0.54 ~ 0.86)	0.63 (0.51 ~ 0.77)
1–4	1,518,624 (14.39%)	343,090 (11.85%)	0.85 (0.70 ~ 1.02)	0.77 (0.66 ~ 0.90)
5–8	4,298,367 (40.74%)	1,205,355 (41.62%)	0.97 (0.86 ~ 1.10)	0.91 (0.82 ~ 1.01)
> 8	4,143,839 (39.28%)	1,178,555 (40.70%)	ref.	ref.
Unknown	337,518 (3.20%)	131,017 (4.52%)	1.42 (1.03 ~ 1.94)	1.44 (1.05 ~ 1.96)
Previous caesarean sections
0	8,901,974 (84.38%)	2,270,906 (78.42%)	ref.	ref.
1	1,505,814 (14.27%)	568,975 (19.65%)	1.55 (1.47 ~ 1.62)	1.40 (1.34 ~ 1.47)
2	118,042 (1.12%)	53,080 (1.83%)	2.05 (1.86 ~ 2.26)	1.62 (1.48 ~ 1.78)
3	4,330 (0.04%)	2,020 (0.07%)	2.12 (1.72 ~ 2.62)	1.51 (1.22 ~ 1.86)
> 3	245 (0.0%)	98 (0.0%)	1.75 (1.29 ~ 2.36)	1.48 (1.09 ~ 2.01)
Unknown	19,664 (0.19%)	807 (0.03%)	0.14 (0.05 ~ 0.45)	0.30 (0.18 ~ 0.50)
Education status
College and above	4,093,424 (38.80%)	1060,938 (36.64%)	0.82 (0.68 ~ 0.98)	0.79 (0.67 ~ 0.93)
Senior/Technical Secondary School	2,797,581 (26.52%)	782,997 (27.04%)	0.89 (0.75 ~ 1.05)	0.86 (0.73 ~ 1.00)
Junior	3,090,089 (29.29%)	890,961 (30.77%)	0.91 (0.78 ~ 1.06)	0.89 (0.77 ~ 1.03)
Primary	292,331 (2.77%)	92,962 (3.21%)	1.01 (0.90 ~ 1.13)	0.96 (0.86 ~ 1.08)
Illiteracy	52,695 (0.50%)	16,680 (0.58%)	ref.	ref.
Unknown	223,949 (2.12%)	51,348 (1.77%)	0.70 (0.23 ~ 2.17)	0.65 (0.22 ~ 1.96)
Type of pregnancy
Singleton	9,922,346 (94.05%)	2,650,301 (91.52%)	ref.	ref.
Multiple	165,759 (1.57%)	72,105 (2.49%)	1.64 (1.52 ~ 1.79)	1.64 (1.52 ~ 1.76)
Unknown	461,964 (4.38%)	173,480 (5.99%)	0.34 (0.17 ~ 0.68)	0.37 (0.20 ~ 0.67)
Maternal complications during pregnancy
Hypertensive disorders of pregnancy	433,481 (4.11%)	155,590 (5.37%)	1.41 (1.33 ~ 1.49)	1.35 (1.28 ~ 1.42)
Diabetes	784,313 (7.43%)	238,512 (8.24%)	1.17 (1.07 ~ 1.28)	1.10 (1.02 ~ 1.18)
Heart disease	37,568 (0.36%)	11,892 (0.41%)	1.19 (1.01 ~ 1.39)	1.10 (0.95 ~ 1.26)
Embolism	1530 (0.01%)	1491 (0.05%)	3.56 (3.10 ~ 4.08)	3.20 (2.80 ~ 3.66)
Liver disease	87,426 (0.83%)	34,540 (1.19%)	1.50 (1.29 ~ 1.75)	1.40 (1.22 ~ 1.60)
Kidney disease	34,251 (0.32%)	14,695 (0.51%)	1.57 (1.26 ~ 1.96)	1.44 (1.15 ~ 1.80)
Lung disease	3434 (0.03%)	2051 (0.07%)	2.19 (1.76 ~ 2.73)	1.79 (1.47 ~ 2.17)
HIV	2129 (0.02%)	827 (0.03%)	1.46 (1.16 ~ 1.84)	1.33 (1.06 ~ 1.68)
Connective tissue disease	3606 (0.03%)	1319 (0.05%)	1.31 (1.01 ~ 1.68)	1.18 (0.94 ~ 1.48)
Cancer	2681 (0.03%)	1002 (0.03%)	1.44 (1.23 ~ 1.69)	1.33 (1.14 ~ 1.55)

Severity of anemia during pregnancy and adverse maternal and fetal outcomes

Multivariate-adjusted analyses (Table 2) were performed to evaluate the associations between anemia severity and adverse pregnancy outcomes. Pregnancy anemia was associated with elevated risks of multiple adverse outcomes, and these risks exhibited a clear dose–response relationship with anemia severity: risk estimates increased progressively with worsening anemia.

Table 2.

Adjusted odds ratio of adverse pregnancy outcomes in anemic females*

Pregnant outcome	All	Mild	Moderate	Severe	Unknown
Maternal outcome
Cesarean section	1.10 (1.02 ~ 1.18)	0.96 (0.91 ~ 1.02)	1.22 (1.11 ~ 1.33)	1.03 (0.92 ~ 1.15)	0.73 (0.64 ~ 0.83)
Uterine rupture	1.84 (1.40 ~ 2.40)	1.73 (1.29 ~ 2.33)	1.83 (1.41 ~ 2.38)	3.75 (2.82 ~ 4.99)	1.02 (0.57 ~ 1.83)
Placenta previa	2.43 (2.19 ~ 2.70)	1.41 (1.26 ~ 1.57)	2.95 (2.64 ~ 3.29)	7.69 (6.70 ~ 8.83)	1.24 (1.02 ~ 1.50)
Placental abruption	2.55 (2.36 ~ 2.76)	1.69 (1.53 ~ 1.87)	2.85 (2.63 ~ 3.09)	8.24 (7.28 ~ 9.32)	1.24 (1.00 ~ 1.54)
Postpartum hemorrhage	5.68 (4.60 ~ 7.02)	1.93 (1.53 ~ 2.44)	6.89 (5.51 ~ 8.70)	42.81 (34.18 ~ 53.63)	1.51 (1.00 ~ 2.27)
Severe postpartum hemorrhage	7.40 (6.19 ~ 8.86)	1.85 (1.52 ~ 2.26)	8.84 (7.24 ~ 10.79)	58.05 (46.33 ~ 72.74)	2.02 (1.40 ~ 2.93)
Blood transfusion	11.21 (10.05 ~ 12.49)	1.90 (1.68 ~ 2.15)	11.59 (10.31 ~ 13.03)	209.82 (183.08 ~ 240.47)	3.00 (2.29 ~ 3.92)
Premature birth	1.07 (1.01 ~ 1.13)	0.98 (0.92 ~ 1.04)	1.10 (1.04 ~ 1.16)	1.66 (1.55 ~ 1.77)	0.99 (0.87 ~ 1.13)
ICU admission	4.51 (3.22 ~ 6.31)	2.57 (1.95 ~ 3.39)	4.99 (3.67 ~ 6.78)	25.59 (15.30 ~ 42.81)	1.14 (0.48 ~ 2.72)
Shock	8.65 (7.26 ~ 10.29)	1.12 (0.91 ~ 1.39)	7.19 (6.03 ~ 8.57)	128.92 (105.89 ~ 156.97)	1.61 (0.80 ~ 3.23)
MODS	8.44 (7.25 ~ 9.83)	1.21 (1.00 ~ 1.47)	6.88 (5.91 ~ 8.01)	131.44 (109.52 ~ 157.75)	1.17 (0.58 ~ 2.36)
Maternal mortality	2.17 (1.69 ~ 2.78)	0.45 (0.29 ~ 0.72)	1.59 (1.19 ~ 2.11)	25.64 (18.66 ~ 35.22)	~
Fetal outcome
Low birth weight	1.06 (1.00 ~ 1.13)	0.94 (0.88 ~ 1.01)	1.10 (1.02 ~ 1.17)	2.13 (1.96 ~ 2.31)	1.06 (0.91 ~ 1.23)
Apgar < 7 points at 1 min or < 7 points at 5 min	1.15 (1.00 ~ 1.33)	0.80 (0.67 ~ 0.95)	1.28 (1.12 ~ 1.46)	3.20 (2.81 ~ 3.65)	0.78 (0.64 ~ 0.95)
Apgar score < 3 points at 1 min or < 5 points at 5 min	1.11 (0.95 ~ 1.28)	0.72 (0.59 ~ 0.88)	1.20 (1.04 ~ 1.38)	3.90 (3.26 ~ 4.65)	0.83 (0.59 ~ 1.17)
Stillbirth	0.89 (0.81 ~ 0.97)	0.65 (0.58 ~ 0.73)	0.88 (0.80 ~ 0.96)	3.39 (0.98 ~ 3.87)	1.19 (0.94 ~ 1.51)

ICU intensive care unit, MODS multiple organ dysfunction syndrome

* Adjusted for appropriate confounders selected based on clinical knowledge and prior evidence, including maternal age, education level, parity, residence, geographic region, and baseline maternal health conditions. The reference group for all groups was non~anemic pregnant women

~: Indicates that the outcome event did not occur and the model could not fit

For a series of maternal outcomes including uterine rupture, placenta previa, placental abruption, postpartum hemorrhage, severe postpartum hemorrhage, blood transfusion, ICU admission, and multiple organ dysfunction syndrome (MODS), the risk increased in proportion to anemia severity. Compared with non-anemic women, those with mild anemia already demonstrated increased risks, with further incremental elevations observed in the moderate and severe anemia groups. As an example, the adjusted odds ratio (aOR) for uterine rupture was 1.73 (95% CI: 1.29–2.33) in mild anemia, 1.83 (95% CI: 1.41–2.38) in moderate anemia, and 3.75 (95% CI: 2.82–4.99) in severe anemia. Similar graded patterns were observed for placenta previa, placental abruption, and other related outcomes.

In addition, moderate and severe anemia were associated with higher risks of preterm birth, shock, maternal mortality, low birth weight, low Apgar score, and very low Apgar score. For preterm birth, the aOR was 1.10 (95% CI: 1.04–1.16) in the moderate anemia group and 1.66 (95% CI: 1.55–1.77) in the severe anemia group. Collectively, these findings indicate that greater anemia severity during pregnancy is associated with higher risks of adverse maternal and neonatal outcomes.

Stratified analyses by singleton and multiple pregnancies (Supplementary Table 1) further explored these associations. Notably, with respect to neonatal outcomes, anemia in singleton pregnancies was significantly associated with increased risks of low birth weight, low Apgar score, and stillbirth, whereas such significant associations were not observed in multiple pregnancies.

Sensitivity analysis across study periods

To assess the robustness of the main findings, we performed a sensitivity analysis by stratifying the study period into two phases: 2012–2013 and 2015–2020. Associations between anemia severity and adverse pregnancy outcomes, with non-anemic women as the reference group, are presented in Supplementary Table 2.

Results across both periods consistently supported a strong and stable association between severe anemia and elevated risks of most adverse maternal and fetal outcomes. For maternal outcomes, severely anemic women showed markedly increased aORs in both periods, including for uterine rupture (aOR 6.19 in 2012–2013 vs. 3.24 in 2015–2020), placenta previa (8.54 vs. 7.40), placental abruption (8.49 vs. 7.82), and severe postpartum hemorrhage (16.67 vs. 10.33). Particularly strong and consistent associations were observed for critical outcomes, such as blood transfusion (209.55 vs. 217.34), shock (109.01 vs. 161.28), multiple organ dysfunction syndrome (MODS) (99.84 vs. 154.20), and maternal mortality (22.21 vs. 28.35).

For fetal outcomes, severe anemia was also consistently associated with higher risks in both periods, including low birth weight (2.06 vs. 2.05), low Apgar score (2.07 vs. 3.01), very low Apgar score (3.41 vs. 3.30), and stillbirth (3.18 vs. 3.27).

We observed a modest temporal trend: the magnitude of aORs generally attenuated from the 2012–2013 period to 2015–2020 for several outcomes, including cesarean section, uterine rupture, and postpartum hemorrhage. Although point estimates decreased, some confidence intervals (e.g., for postpartum hemorrhage in 2015–2020) crossed the null value. By contrast, associations for mild and moderate anemia were less consistent, especially for fetal outcomes such as Apgar scores, which were often non-significant in the later period. The “unknown anemia severity” group consistently showed aORs close to 1, suggesting no meaningful excess risk.

Discussion

Our study covered 30 provinces (excluding Tibet), eastern/central/western regions, rural/urban districts, and medical institutions of all levels (including levels I, II, and III), and the enrolled population accounted for 8%–10% of the total national births from 2012 to 2020. In this large cross-sectional study, we found that the prevalence of anemia during pregnancy in China is 21.25%, which is a moderate public health issue according to WHO criteria, although the incidence of anemia during pregnancy is lower than that in other developing and relatively poor-resource countries [9]. Overall, mild, moderate, and severe anemia were significantly proportionally associated with uterine rupture, placenta previa, placental abruption, severe postpartum hemorrhage, blood transfusion, and ICU admission. Moderate and severe anemia was significantly associated with an increased risk of cesarean section, postpartum hemorrhage, premature birth, shock, MODS, maternal mortality, low birth weight, poor Apgar score and stillbirth. Analysis of singleton and multiple pregnancies separately showed that although the prevalence of anemia in multiple pregnancies was higher, anemia was not more seriously associated with maternal and fetal outcomes in multiple pregnancies than in singleton pregnancies, especially in case of premature births, low Apgar scores, and stillbirths.

Given the large sample size and wide geographical coverage, including rural and urban areas, our study revealed wide variations in prevalence across regions, with the highest being in the western and central regions, followed by in the eastern region. This might be associated with the different socioeconomic conditions, lifestyles, dietary habits, and health-seeking behaviors across different cultures. The overall prevalence of anemia during pregnancy and regional differences are inconsistent among different studies. A multi-center study comparing the prevalence of anemia among three big cities in China and showed that the anemia prevalence of 38.8% and 23.9% in Guangzhou and Chengdu, respectively, was significantly higher than that in Beijing (19.3%) [11], which may partially represent the Western, Northern, and Southern regions of China, respectively. Qiao’s work, which included 18,948,443 pregnant women from tertiary hospitals in large cities, was more focused on the urban population with better nutritional and health characteristics and lacked an analysis of regional differences [4]. These studies have illustrated the prevalence of anemia among pregnant women in China from different perspectives. Compared with them, our study was based on a large-scale, multi-center surveillance system with wide geographic coverage (eastern, central, and western regions), standardized data collection, and long-term continuous monitoring. These strengths ensured our findings were more comprehensive and objective.

Several risk factors may have contributed to the increased prevalence of anemia during pregnancy. Previous studies have shown that being over the age of 35 or 40 years can increase the risk of anemia during pregnancy [24–26]. However, our study showed a greater risk of anemia during pregnancy among adolescents aged 15–19, and the risk of anemia increased with age after 20 years; however, the difference was not significant. Thus, the association between anemia during pregnancy and maternal age led to a hook-shaped curve in this study, rather than an L-shaped or a reverse J-shaped curve as observed in most studies in China and other limited-resource countries [4, 26, 27]. Therefore, gestational anemia among adolescents aged 15–19 requires special consideration. This may attribute to rapid physiological growth, socio-behavioral factors specific to this life stage, lack awareness of contraception, unavailable access to adolescent-friendly services. On adjusting for multiple confounders in our analysis, we found that the age of 15–19 years was the actual predictor of anemia among women of the childbearing age and being over the age of 35 or 40 was not a direct factor. This could be explained by the fact that the probability of complications increases in women aged > 35 or 40 years, leading to an elevated risk of anemia. While the determinants of this elevated risk warrant further investigation, our findings highlight a critical window for targeted public health intervention.

Furthermore, our study revealed an association between anemia and low education levels, indicating the potential role of socioeconomic and environmental factors. Multiparity, rather than multigravidity, was found to be significantly associated with a higher risk of gestational anemia. These results are consistent with studies conducted in Japan and Sudan, which investigated the same associations and represented different economic development levels [13, 14].

Regarding the association between maternal complications and anemia during pregnancy, it is worth mentioning that the incidence of acquired immunodeficiency syndrome (AIDS) in China has shown a significant increasing trend [28]. In this study, 2643 women with singleton pregnancies (prevalence: 0.021%) and 46 with multiple pregnancies (prevalence: 0.019%) were infected with HIV. Among all participants, the overall prevalence of HIV infection was 0.021%. Furthermore, singleton pregnant women with HIV infection had a significantly higher risk of anemia. This could be associated with the direct effect of HIV on hematopoietic progenitor cells and activation of inflammatory pathways that inhibit erythropoiesis [29]. Additionally, the response to iron therapy in HIV-positive women was slower than in HIV-negative women [30]. Unlike several studies that have concluded that anemia is associated with an increased risk of hypertensive disorders during pregnancy [31, 32], we verified and agree with the statement that inverse association is likely to explain why hypertensive disorders of pregnancy are more frequent in women with anemia. It is possible that low hemoglobin levels represent a symptom or consequence of these conditions, rather than an independent associated or causal factor.

The strength of the association between anemia during pregnancy and adverse maternal outcomes increases with the severity of anemia. Notably, in case of singleton pregnancies, the risk of preterm birth is increased in women with severe anemia, while mild and moderate anemia did not show the same association. In multiple pregnancies anemia did not increase the risk of preterm birth when adjustments were made, possibly due to the fact that the association of multiple pregnancies with preterm birth was significantly greater than that of secondary anemia in twin pregnancies. Regarding maternal mortality, we identified similar adverse outcomes to those in previous studies, which found increased risks in patients with moderate and severe anemia, with no risk associated with mild anemia compared to those with normal hemoglobin concentrations [33, 34]. Propensity score analysis showed that severe anemia was associated with maternal death in a large, multi-country dataset [35]. Another study reported a U-shaped curve for the risk of adverse maternal outcomes in relation to maternal hemoglobin concentrations, but the relationship differed by trimester [36]. The association with adverse outcomes is more evident in case of low hemoglobin concentrations in early pregnancy. In the second or third trimesters, these relationships generally become weaker or nonexistent [37]. However, our cross-sectional study lacked these data. Regarding fetal outcomes, the risk of low birth weight, low Apgar score, and stillbirth was higher among singleton pregnant women with severe anemia than among those with normal hemoglobin levels in our study. This result is supported by that of previous study [38]. A prospective cohort study conducted in India and Pakistan found that severe maternal anemia was associated with a higher risk of poor maternal, fetal, and neonatal outcomes, whereas other degrees of anemia were not [39]. Anemia in multiple pregnancies was only associated with an increased risk of low birth weight and showed no statistically significant association with low Apgar score and stillbirth in our study, which was partly supported by Amit Kosto et al., who found that maternal anemia in twin gestations was not associated with an increased risk of adverse perinatal outcomes [40].

Although anemia prevalence is higher in twin/multiple pregnancies, its association with adverse fetal outcomes—such as low Apgar scores, and stillbirth—is weaker than in singleton pregnancies. The phenomenon in multiple pregnancies may attribute to the more significant physiological expansion of maternal blood volume (hemodilution), differences in placental function and compensatory mechanisms, as well as the fetus’s adaptability to the hypoxic environment. This difference likely reflects the potent inherent drivers of adverse outcomes in twin/multiple pregnancies (e.g., the high baseline risk of iatrogenic preterm birth and stillbirth) coupled with more intensive antenatal surveillance and intervention (e.g., more proactive anemia management and fetal monitoring), which collectively diminish the relative contribution of anemia as an independent risk factor.

The relationship between gestational anemia and adverse outcomes requires careful differentiation between causation and correlation. Indeed, we acknowledge that certain complications (e.g., placenta previa, chronic placental abruption) primarily cause anemia through chronic blood loss. However, our adjusted analyses identified significant associations suggesting that anemia may exacerbate outcomes through two distinct pathways as follows. First, severe anemia (Hb < 80 g/L) is associated with impaired oxygen delivery, potentially affecting placental angiogenesis, increased placental size, and fetal organ development [41], particularly in fetal neural and cardiovascular systems [42, 43]. Second, due to diminished compensatory capacity [44, 45], preexisting anemia was associated with worsened the impact of acute hemorrhage (e.g., placental abruption, postpartum bleeding, uterine rupture), as well as increased transfusion needs and ICU admissions, which was consistent with the conclusions of previous studies [46, 47]. Clinically, this bidirectional relationship underscored the importance of investigating persistent anemia for underlying pathology (e.g., occult placental hemorrhage) and proactive management of pre-delivery anemia to mitigate secondary complications, as even anemia secondary to bleeding independently predicts worse outcomes.

The sensitivity analysis confirms the robustness of our primary findings. The consistent and dose-response relationship between anemia severity—particularly severe anemia —and a wide spectrum of serious maternal and fetal complications across two distinct time periods strengthens the causal inference. The observed trend of attenuated aORs for some outcomes in the later period (2015–2020) may reflect improvements in overall obstetric care over time. Nonetheless, the persistently dramatically high risks associated with severe anemia for life-threatening conditions like shock, MODS, and the need for blood transfusion underscores that severe antenatal anemia remains a critical, independent risk factor for adverse pregnancy outcomes. These results highlight the urgent need for vigorous public health interventions aimed at early detection and aggressive management of anemia, especially its severe forms, in pregnancy.

Recently, some international researchers have proposed reevaluating the definition of pregnancy-induced anemia to distinguish between normal and pathological anemia with low hemoglobin values. Redefining the threshold could prevent unnecessary tests and non-indicative iron therapy in pregnancies with regular low hemoglobin levels, leading to more accurate diagnosis and better treatment of anemia [48]. Based on existing diagnostic criteria, our large-sample study concluded that anemia is associated with definite adverse pregnancy outcomes and therefore, we do not advocate lowering the threshold. This finding provides good evidence to support international research. Future studies should evaluate whether early iron supplementation or transfusion thresholds alter these risks.

Clinical and research implications

China has a long way to go to improve the current status of anemia during pregnancy and narrow the gap in the prevalence of anemia in pregnant women between China and developed countries. Given that the WHO has set a global nutrition target to reduce the prevalence of anemia in women of reproductive age by 50% by 2025, the State Council of China has issued “the National Nutrition Plan of Action (2017–2030)” and has put forward a target to reduce the prevalence of anemia in pregnant women to < 10% by 2030 [49]. The National Nutrition Plan 2017–2030 clearly proposes nutritional interventions for pregnant women using “nutrition kits” based on a rational diet. These kits comprise a type of health care product specifically designed to meet the unique needs of pregnant women, which contains various important nutrients required during pregnancy. Mothers from resource-poor regions with limited access to healthcare may only have access to hemoglobin/hematocrit testing and oral iron supplementation. It is imperative that our primary medical institutions enhance the awareness of pregnancy healthcare services and improve their facilities.

A major strength of our study is that we were able to use data from the National Maternal Near Miss Surveillance System, which covers a large population base, a wide geographical distribution, and samples from different institutional levels, including singleton and multiple pregnant women, covering 30 provinces (excluding Tibet), eastern/central/western regions, rural/urban districts, and medical institutions at all levels (including levels I, II, and III). The enrolled population accounted for 8%–10% of the total national birth ratio from 2012 to 2020. We identified the prevalence of anemia during pregnancy and found that anemia is definitely associated with adverse pregnancy outcomes by providing more nationwide, timely, up-to-date, high-quality evidence. The association between anemia and adverse pregnancy outcomes was assessed on a large national scale with adjustments for relevant confounders.

Limitaions

This study has several limitations that should be acknowledged. First, as a cross-sectional design, it relied solely on the lowest hemoglobin concentration measured at admission before delivery. This approach includes women who may have experienced anemia earlier in pregnancy that was resolved by the time of delivery, precluding an assessment of the impact of transient or chronic anemia on pregnancy outcomes. Consequently, the timing of anemia onset during pregnancy, the specific type of anemia (e.g., thalassemia-related anemia, iron deficiency, folate deficiency), and its progression over the gestational period remained unknown. Second, data on potential confounding factors or associated variables, including maternal body mass index (BMI), conception by in vitro fertilization (IVF), interpregnancy interval, and history of parasitic infection, were not available in our dataset and thus could not be included in the analyses. Third, it was critically important that information on iron and/or folic acid supplementation during pregnancy was not collected. This key variable could not be adjusted for in the multivariable models, potentially introducing confounding bias into the observed associations between admission hemoglobin levels and pregnancy outcomes, as supplementation status is likely associated with both anemia severity and the risk of adverse outcomes. Besides, although classifying missing data as “Unknown” prioritizes transparency, it risks bias if the reason for the omission is systematic. Nonetheless, this strategy is preferable to imputation, as it makes no statistical assumptions and explicitly quantifies the missing data. We believe this enhances the study’s validity and clarity, paving the way for future work with more comprehensive datasets. These limitations, particularly the lack of data on antenatal supplementation, necessitate cautious interpretation of our findings regarding the direct associations between hemoglobin levels at delivery and adverse outcomes. Future prospective studies should incorporate serial hemoglobin measurements throughout pregnancy, detailed characterization of anemia etiology, comprehensive data on supplementation adherence and other relevant covariates (e.g., BMI, IVF status, parasitic infections), and potentially randomized designs to better elucidate the relationships.

Conclusion

In conclusion, this study provides some understanding of the comprehensive current status of anemia in pregnant women in China from 2012 to 2020 and determines how far we are from the 2030 target. The results also provide strong evidence indicating that the anemia status during pregnancy needs to be improved in different regions, populations, and age groups; furthermore, this evidence can aid in developing and implementing national health policies and strategies. A coordinated, multisectoral, and multilevel response is recommended as an intervention for patients with any severity level of anemia in order to prevent potential adverse outcomes.

Abbreviations

ICU

Intensive care unit

MODS

Multiple organ dysfunction syndrome

Authors' contributions

Conceptualization: XL, YX, JZ, JL, RZ. Data curation: JZ, JL. Formal analysis: XL, YX. Funding acquisition: JL, RZ. Investigation: XL, YX, JL, RZ. Methodology: XL, YX, YZ, YM, PC, YW, SZ. Supervision: XL, YX. Original draft: XL. Review and editing: XL, YX, ML, ZL, JL, RZ. All authors read and approved the final manuscript.

Funding

This study was granted by the National Key R&D Program of China (No. 2021YFC2701600; No. 2021YFC2701601; No. 2022YFC2704600 and No. 2022YFC2704605).

Data availability

Though individual-level sensitive data should not directly be shared based on the Chinese data protection legislation. It was available for research upon reasonable requests to the National Maternal and Child Health Surveillance Office of China and necessary permission from the National Health Commission of China.

Declarations

Ethics approval and consent to participate

The study was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. The NMNMSS established by the National Health Commission of China was approved by the Ethics Committee of West China Second University Hospital, Sichuan University, China (Protocol ID: 2012008). Patient consent was obtained and informed of permitted use of collected data for further following studies by the surveillance facilities/hospitals when the patient was admitted. It is important to emphasize that minors (refer to citizens younger than 18) seeking medical treatment shall be accompanied by their parents or legal guardians in China. Therefore, the consent of both the minor and their parents/legal guardians is also required to participate in this study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.