J Chin Med Assoc

Original Article

Association between maternal anemia at

admission for delivery and adverse perinatal
Downloaded from http://journals.lww.com/jcma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW

outcomes
Fu-Chieh Chua, Steven Shen-Wen Shaoa, Liang-Ming Loa, T’sang-T’ang Hsieha, Tai-Ho Hunga,b,*
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC1y0abggQZXdgGj2MwlZLeI= on 10/24/2024

a
Department of Obstetrics and Gynecology, Taipei Chang Gung Memorial Hospital, Taipei, Taiwan, ROC; b College of Medicine,
Chang Gung University, Taoyuan, Taiwan, ROC

Abstract
Background: Maternal anemia is a risk factor for poor pregnancy outcomes and threatens maternal or fetal life. Anemia increases
the risk of low birth weight and preterm birth. We aimed to determine the cutoff level of hemoglobin and risk factors for maternal
anemia at admission for delivery and investigate the association between maternal anemia and adverse perinatal outcomes in
contemporary Taiwanese women.
Methods: About 32,234 women admitted to the Taipei Chang Gung Memorial Hospital from 2001 to 2016 were enrolled in this
retrospective observational cohort study. The prevalence of pre-delivery maternal anemia in Taiwan and the maternal demographic
and perinatal outcomes associated with maternal anemia was assessed.
Results: The 10th and 5th percentile hemoglobin levels of the test cohort (2001–2008, n = 15,602) were 10.8 g/dL and 9.9 g/dL,
respectively. In the study cohort (2009–2016, n = 13,026), women who were multiparous, who were aged >34 years, with history
of cesarean delivery, and with history of uterine fibroids had higher prevalence of anemia. Anemic women were at increased risk
of cesarean delivery, primary cesarean delivery, premature rupture of membranes, early preterm birth <34 weeks, having very low
birth weight infants (<1,500 g), having large for gestational age infants, and neonatal intensive care center transfer, but at lower risk
of having small for gestational age infants.
Conclusion: Maternal anemia at delivery is a risk factor for primary cesarean delivery and adverse maternal and neonatal out-
comes. Furthermore, we hypothesize that maternal anemia might increase fetoplacental vasculogenesis and angiogenesis as an
adaptive response.
Keywords: Anemia; Cesarean section; Maternal mortality; Pregnancy outcome

1. INTRODUCTION risk for maternal and neonatal mortality, infectious diseases,4

Anemia is a condition in which the number of red blood cells and impaired fetal neurodevelopment.5 Postpartum anemia is
or their oxygen-carrying capacity is insufficient to meet the also associated with poorer quality of life, including increased
physiologic needs of the body. According to the World Health tiredness, breathlessness, palpitations and infections,6 and
Organization, anemia in pregnancy is defined as a hemoglobin greater stress and depression.7,8
(Hb) level <11 g/dL, with a rate of 23% in industrialized coun- However, the cutoff value of 11 g/dL recommended by the
tries and a global prevalence of 38.2% in 2011.1 Several stud- World Health Organization was derived from studies con-
ies have shown that women with anemia during pregnancy ducted in both developed and developing countries, published
are more likely to have adverse pregnancy outcomes. Anemia decades ago, and was regardless of the timing in the period of
in the first or second trimester increases the risk of low birth pregnancy. Moreover, prior classification of the severity of ane-
weight (LBW, defined as a birth weight <2,500 g) and preterm mia (mild, moderate, and severe) is arbitrary, without epide-
birth.2,3 Women with moderate (a Hb level of 7–9 g/dL) or miologic evidence to support its feasibility. Indeed, it has been
severe anemia (a Hb level less than 7 g/dL) are at increased suggested that the standardized Hb cutoff value for maternal
anemia should be adjusted first for factors that decrease Hb
levels, such as race, population, and gestational weeks, and
*Address correspondence. Dr. Tai-Ho Hung, Department of Obstetrics and
then further adjusted for well-known factors contributing to
Gynecology, Taipei Chang Gung Memorial Hospital, 199, Dun-Hua North Road, increasing Hb levels, including altitude of residence and smok-
Taipei, Taiwan, ROC. E-mail address: [email protected] (T.-H. Hung). ing behavior.9 To date, no study has investigated the preva-
Conflicts of interest: The authors declare that they have no conflicts of interest lence of anemia during pregnancy and the association between
related to the subject matter or materials discussed in this article. maternal anemia and adverse perinatal outcomes in Taiwanese
Journal of Chinese Medical Association. (2020) 83: 400-405. population.
Received July 9, 2019; accepted September 11, 2019. Therefore, this study aimed (1) to determine the cutoff level
of Hb and risk factors for maternal anemia at admission for
doi: 10.1097/JCMA.0000000000000215.
delivery and (2) to investigate the association between mater-
Copyright © 2019, the Chinese Medical Association. This is an open access
article under the CC BY-NC-ND license (http://creativecommons.org/licenses/
nal anemia and adverse perinatal outcomes in contemporary
by-nc-nd/4.0/) Taiwanese women.

400 www.ejcma.org

<zdoi: 10.1097/JCMA.0000000000000215>
 Original Article. (2020) 83:4 J Chin Med Assoc

2. METHODS (n = 127), stillbirths (n = 33), chronic hypertension (n = 33),

The data for this study were obtained from Taipei Chang pre-pregnancy diabetes mellitus (n = 36), hyperthyroidism (n
Gung Memorial Hospital’s computerized obstetrics database. = 66), hypothyroidism (n = 34), autoimmune disease (n = 9),
Demographic characteristics, medical and obstetric histories, and epilepsy (n = 16) were excluded. In addition, women with
and information regarding the course of the index pregnancy and gestational hypertension (n = 76), preeclampsia (n = 147), and
perinatal outcomes were obtained. In our hospital, all women placenta previa (n = 1324) were also excluded as these condi-
underwent routine complete blood cell count upon admission tions may affect the decision regarding the timing and modes
Downloaded from http://journals.lww.com/jcma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW

for vaginal or cesarean delivery. The data were collected by of delivery. As placenta previa is the most common causes for
trained personnel through daily abstraction from medical and severe antepartum hemorrhage from the second trimester, we
delivery records and via postpartum interviews, if necessary, to chose to exclude patients with placenta previa from our data.
collect supplemental information. Audits of these data were rou- For other, less common, causes of antepartum hemorrhage, such
tinely performed every 2 weeks at departmental meetings. The as cervical insufficiency, placental accrete or placental abrup-
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC1y0abggQZXdgGj2MwlZLeI= on 10/24/2024

study was approved by the Institutional Review Board of Chang tion, we wanted to assess the associations between these con-
Gung Memorial Hospital (IRB no. 201900287B0). founding factors and different cutoff levels of Hb for anemia.
As a result, 13,026 deliveries were included in the analysis and
2.1. Study populations defined as the study cohort (Fig. 2).
As elective termination of pregnancy before 24 weeks of gesta- 2.2. Definitions of confounders and adverse perinatal
tion is allowed in Taiwan, we included all deliveries after 24
outcomes
weeks of completed gestation between January 1, 2001 and
December 31, 2016 in the analysis. Our study consisted of two The following maternal demographic and pregnancy character-
steps. First, we analyzed the Hb levels upon admission of women istics were considered as confounders, and their effects on the
who delivered between 2001 and 2008 (n = 16,670), excluding association between maternal anemia and adverse perinatal out-
pregnancies complicated by multiple gestations (n = 634), fetal comes were adjusted in the multiple logistic regression: maternal
chromosomal or structural anomalies (n = 163), and stillbirths age at delivery (categorized as <20, 20–34, and >34 years old);
(n = 77). In addition, women with chronic hypertension (n = 60), primiparity; pre-pregnancy body mass index (BMI, calculated
pre-pregnancy diabetes mellitus (n = 33), hyperthyroidism (n = as weight [kg]/height [m2] and categorized as <18.5, 18.5–24.9,
43), hypothyroidism (n = 12), autoimmune diseases (n = 24), and >24.9 kg/m2); gestational age (estimated based on the first
epilepsy (n = 18), and renal disease (n = 4) were also excluded as day of the mother’s last normal menstrual period or, if this date
these diseases or their associated medication may have an influ- was unknown or uncertain, assigned by ultrasound dating); ges-
ence on the Hb levels. Overall, 15,602 deliveries were included tational weight gain (calculated by subtracting each woman’s
and defined as the test cohort for analysis (Fig. 1). Maternal pregestational weight from her weight at delivery in kilograms);
anemia was defined as a Hb level less than the 10th (anemia 1) Hb level, mean corpuscular volume, and platelet count before
or 5th (anemia 2) percentile of this cohort. delivery; history of cesarean delivery; history of preterm birth;
Next, the feasibility of these two different criteria for ane- history of fetal death; history of spontaneous and induced abor-
mia were tested by investigating their associations with adverse tion; history of uterine fibroids; history of genetic amniocentesis;
perinatal outcomes in women who delivered between 2009 and history of conception assisted by reproductive technology; his-
2016 (n = 15,564). Pregnancies complicated by multiple ges- tory of cigarette smoking during pregnancy; and fetal sex.
tations (n = 637), fetal chromosomal or structural anomalies We studied the following adverse perinatal outcomes:
cesarean delivery, premature rupture of membranes (PROM),

Fig. 1 Test cohort sample selection process. Fig. 2 Study cohort sample selection process.

www.ejcma.org  401
 F.-C. Chu et al. J Chin Med Assoc

oligohydramnios (amniotic fluid index <5 cm), polyhydramnios Table 1.

(amniotic fluid index >24 cm), group B streptococcal coloniza- Differences in characteristics between the test cohort and study
tion at the genitorectal tract, gestational diabetes mellitus (diag- cohort.
nosed by one-step 75-gram, 2-hour oral glucose tolerance test),
postpartum hemorrhage (blood loss >500 mL after vaginal birth Test cohort* Study cohort*
or >1,000 mL after cesarean delivery), operative vaginal deliv- Characteristics (n = 15,602) (n = 13,026) p
ery, placenta accreta, placental abruption, acute chorioamnio- Maternal age (years)
Downloaded from http://journals.lww.com/jcma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW

nitis, early preterm birth (birth before 34 weeks of gestation), <20 65 (0.4%) 21 (0.2%) <0.01
late preterm birth (birth between 34 and 37 weeks of gesta- 20–34 11,178 (71.7%) 8,535 (65.5%) <0.01
tion), very LBWt (VLBW, birth weight <1,500 g), LBW (birth >34 4,359 (27.9%) 4,470 (34.3%) <0.01
weight between 1,500 and 2,500 g), macrosomia (birth weight Gestational age (week) 38.5 ± 1.6 38.4 ± 1.6 <0.01
>4,000 g), small for gestational age (SGA, birth weight below Primipara 8,565 (54.9%) 7,203 (55.3%) 0.50
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC1y0abggQZXdgGj2MwlZLeI= on 10/24/2024

the 10th percentile of the mean weight corrected for gestational Gestational weight gain (kg) 14.5 ± 4.3 13.1 ± 4.3 <0.01
age and fetal sex), and large for gestational age (LGA, birth History of cesarean delivery 2,041 (13.1%) 1,950 (15.0%) <0.01
weight above the 90th percentile of the mean weight corrected History of preterm birth 49 (0.3%) 68 (0.5%) <0.01
for gestational age and fetal sex), meconium stained amniotic History of fetal death 152 (1.0%) 137 (1.1%) 0.51
fluid, transfer to neonatal intensive care units (NICU) after History of induced or spontaneous 5,424 (34.8%) 3,880 (29.8%) <0.01
delivery, Apgar sore <7 at 1 minute and 5 minutes, and fetal and abortion
neonatal death. Uterine fibroids 112 (0.7%) 322 (2.5%) <0.01
Genetic amniocentesis 4,179 (26.8%) 4,811 (36.9%) <0.01
2.3. Statistical analysis Conception by reproductive 161 (1.0%) 243 (1.9%) <0.01
Statistical analyses were performed using MedCalc for technology
Windows, version 15.12.0 (MedCalc Software, Ostend, Smoking during pregnancy 24 (0.2%) 21 (0.2%) 0.56
Male fetus 8,193 (52.5%) 6,715 (51.5%) 0.10
Belgium). Continuous variables were presented as mean ± SD,
whereas categorical variables were calculated as number and * Data presented as number (%) or mean ± SD p value based on independent t test or χ2-test.
rate (%). Comparisons between groups were computed using
independent t test or χ2-test as appropriate. A p value <0.05
was considered significant. Multivariable logistic regression was CI, 1.23-5.17), PROM (adjusted OR: 1.66, 95% CI, 1.19-2.33),
used to control for potential confounding factors when assessing and early preterm birth (adjusted OR: 2.16, 95% CI, 1.54-3.03)
the associations between different cutoff levels of Hb for anemia than those with a Hb level ≥10.8 g/dL at admission for delivery.
and adverse perinatal outcomes. Adjusted odds ratios (ORs) and Neonates of women with anemia 1 were more likely to be VLBW
95% CIs were calculated to estimate relative risk. (adjusted OR: 2.08, 95% CI, 1.22-3.52), LGA (adjusted OR:
1.24, 95% CI, 1.03-1.49), and transferred to NICU (adjusted
3. RESULTS OR: 1.72, 95% CI, 1.31-2.26).
Moreover, in Table 4, women with anemia 2 (Hb level <9,.9 g/
The maternal characteristics of the study participants including dL) had higher rates of cesarean delivery (adjusted OR: 1.39,
the test and study cohort are presented in Table 1. A significant 95% CI, 1.12-1.72), primary cesarean delivery (adjusted OR:
difference was observed between these two groups of women in 1.32, 95% CI, 1.10-1.58), PROM (adjusted OR: 1.83, 95% CI,
terms of maternal age, gestational age, gestational weight gain, 1.10-3.02), and early preterm birth (adjusted OR: 3.01, 95%
history of cesarean delivery, history of preterm birth, history of CI, 1.94-4.69) than those with Hb level ≥9.9 g/dL. Neonates
induced or spontaneous abortion, uterine fibroids, genetic amni- of women with anemia 2 were at increased risk for VLBW
ocentesis, and conception by reproductive technology. (adjusted OR: 2.45, 95% CI, 1.16-5.15) and NICU transfer
In the test cohort, the mean Hb level (± SD) was 12.2 g/dL (± (adjusted OR: 2.15, 95% CI, 1.47-3.15), and a decreased risk
1.1 g/dL). The 10th and 5th percentile Hb levels were 10.8 and for SGA (adjusted OR: 0.56, 95% CI, 0.35-0.89) compared
9.9 g/dL, respectively. Hence, a Hb level below 10.8 g/dL was with women without anemia.
used as the cutoff value for anemia 1, while a Hb level below
9.9 g/dL as the cutoff value for anemia 2. As shown in Table 2,
compared with women with a Hb level ≥10.8 g/dL, those with 4. DISCUSSION
anemia 1 were more likely to be multiparous, aged >34 years,
had previous history of cesarean delivery, has previous history In this homogeneous contemporary Taiwanese population, we
of uterine fibroids, and had less weight gain during pregnancy. found that the mean Hb level at admission for delivery was
Women with anemia 2 who had a Hb level ≥9.9 g/dL were more 12.2 g/dL, with the 10th and 5th percentile levels of 10.8 and
likely to be multiparous, had previous history of cesarean deliv- 9.9 g/dL, respectively. When these two measurements were used
ery, and had less weight gain during pregnancy. to define anemia, we found that anemic women were more likely
A logistic regression model was performed to identify the fac- to be multiparous, aged >34 years, had previous history of cesar-
tors affecting maternal anemia. After adjusting for other vari- ean delivery, and had less weight gain during pregnancy com-
ables, multiparous, maternal age >34 years, previous history of pared with women without anemia. Furthermore, women with a
cesarean delivery, and previous history of uterine fibroids were Hb level <10.8 g/dL (the 10th percentile level) had 1.5 to 2-fold
found as the predictors of anemia among the pregnant women. increased risk for adverse perinatal outcomes including cesarean
Next, we investigated the association between these two dif- delivery, primary cesarean delivery, polyhydramnios, PROM,
ferent criteria for anemia and adverse perinatal outcomes after early preterm birth, having infants of VLBW, and NICU admis-
adjusting for the confounders by multiple logistic regression. As sion than those who had a Hb level ≥10.8 g/dL. The strength of
shown in Table 3, women with anemia 1 (Hb level <10.8 g/dL) the association between maternal anemia and adverse perinatal
had higher rates of cesarean delivery (adjusted OR: 1.40, 95% outcomes increased slightly to 2.0- to 2.5-fold when the Hb level
CI, 1.22-1.59), primary cesarean delivery (adjusted OR: 1.12, of 9.9 g/dL (the 5th percentile) was used as the cutoff level for
95% CI, 1.01-1.26), polyhydramnios (adjusted OR: 2.53, 95% anemia.

402 www.ejcma.org
 Original Article. (2020) 83:4 J Chin Med Assoc

Table 2.
Differences in maternal characteristics between women with different cutoff levels of hemoglobin for anemia and women without
anemia in the study cohort.

Characteristics Anemia 1 Anemia 2

Hb <10.8 g/dL* Hb ≥10.8 g/dL* p Hb <9.9 g/dL* Hb ≥9.9 g/dL* p
Maternal age (years)
Downloaded from http://journals.lww.com/jcma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW

<20 6 (0.3%) 15 (0.1%) 0.06 1 (0.2%) 20 (0.2%) 0.98

20–34 1,114 (62.1%) 7,421 (66.1) <0.01 408 (65.0%) 8,127 (65.5%) 0.79
>34 675 (37.6%) 3,795 (33.8%) <0.01 219 (34.8%) 4,251 (34.3%) 0.78
Gestational age (week) 38.2 ± 1.9 38.5 ± 1.6 <0.01 38.1 ± 2.1 38.5 ± 1.6 <0.01
Primiparity 826 (46.0%) 6,377 (56.8%) <0.01 275 (43.7%) 6,928 (55.9%) <0.01
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC1y0abggQZXdgGj2MwlZLeI= on 10/24/2024

Pre-pregnancy body mass index (kg/m2)

<19.8 265 (14.8%) 1,585 (14.1%) 0.47 99 (15.7%) 1,751 (14.1%) 0.25
19.8–24.2 1,359 (75.7%) 8,430 (75.1%) 0.58 470 (74.7%) 9,318 (75.2%) 0.80
>24.2 171 (9.5%) 1,216 (10.8%) 0.11 60 (9.6%) 1,328 (10.7%) 0.35
Gestational weight gain (kg) 12.8 ± 4.1 13.2 ± 4.3 <0.01 12.7 ± 4.2 13.1 ± 4.3 0.01
History of cesarean delivery 384 (21.4%) 1,566 (13.9) <0.01 154 (24.5%) 1,796 (14.5%) <0.01
History of preterm birth 12 (0.7%) 56 (0.5%) 0.35 5 (0.8%) 63 (0.5%) 0.33
History of fetal death 20 (1.1%) 117 (1.0%) 0.78 3 (0.5%) 23 (1.0%) 0.15
History of induced or spontaneous abortion 543 (30.2%) 4,874 (34.4) 0.65 202 (32.1%) 3,678 (29.7%) 0.19
Uterine fibroids 59 (3.3%) 263 (2.3%) 0.02 21 (3.3%) 301 (2.4%) 0.15
Genetic amniocentesis 681 (38.0%) 4,130 (36.8%) 0.35 232 (36.9%) 4,579 (37.0) 0.98
Conception by reproductive technology 37 (2.1%) 206 (1.8%) 0.51 9 (1.4%) 234 (1.9%) 0.41
Smoking during pregnancy 4 (0.2%) 17 (0.2%) 0.48 0 (0.0%) 21 (0.2%) 0.30
Male fetus 912 (50.8%) 5,803 (51.7%) 0.48 312 (49.6%) 6,403 (51.6%) 0.32
* Data presented as number (%) or mean ± SD.
p value based on independent t test or χ2-test.

Table 3.
Associations between maternal anemia defined as a Hb <10.8 g/dL and adverse perinatal outcomes.
Variables Anemic* Non-anemic* Unadjusted odds ratio 95% CI Adjusted odds ratio 95% CI p
Total number of pregnancies 1,795 (13.8%) 11,231 (86.2%)
Maternal outcome
Cesarean section 798 (44.4%) 4,099 (36.5%) 1.39 1.25-1.54 1.40 1.22-1.59 <0.01
Primary Cesarean section 453 (25.2) 2,564 (22.7) 1.14 1.02-1.28 1.12 1.01-1.26 <0.01
Postpartum hemorrhage 36 (2.0%) 176 (1.5%) 1.28 0.89-1.84 1.26 0.88-1.82 0.19
Operative vaginal delivery 72 (4.0%) 525 (4.7%) 0.85 0.66-1.09 1.00 0.78-1.29 0.95
Cervical incompetence 7 (0.4%) 34 (0.3%) 1.28 0.57-2.91 1.06 0.46-2.42 0.88
Placental accreta 3 (0.2%) 26 (0.2%) 0.72 0.21-2.38 0.68 0.20-2.27 0.53
Placental abruption 39 (2.2%) 179 (1.6) 1.37 0.96-1.94 1.38 0.97-1.97 0.06
Oligohydramnios 22 (1.2%) 114 (1.1%) 1.11 0.70-1.75 1.23 0.78-1.95 0.36
Polyhydramnios 11 (0.6%) 26 (0.2%) 2.65 1.31-5.38 2.53 1.23-5.17 0.01
GBS infection 291 (16.2%) 1,744 (15.5%) 1.05 0.91-1.20 1.05 0.92-1.21 0.42
Gestational diabetes mellitus 149 (8.3%) 1,077 (9.6%) 0.83 0.71-1.02 0.84 0.70-1.01 0.06
PROM 44 (2.4%) 177 (1.6%) 1.56 1.12-2.19 1.66 1.19-2.33 <0.01
Chorioamnionitis 17 (0.9%) 87 (0.8%) 1.22 0.72-2.06 1.50 0.88-2.54 0.12
Neonatal outcome
Early preterm birth <34 weeks 48 (2.7%) 134 (1.2%) 2.27 1.63-3.17 2.16 1.54-3.03 <0.01
Late preterm birth 34–37 weeks 129 (7.2%) 663 (5.9%) 1.23 1.01-1.50 1.19 0.98-1.45 0.10
Very low birth weight 19 (1.1%) 56 (0.5%) 2.13 1.26-3.60 2.08 1.22-3.52 <0.01
Low birth weight 116 (6.5%) 637 (5.7%) 1.14 0.93-1.40 0.88 0.69-1.12 0.30
Macrosomia 28 (1.6%) 195 (1.7%) 0.89 0.60-1.33 0.89 0.60-1.33 0.59
Small for gestational age 82 (4.6%) 653 (5.8%) 0.77 0.61-0.98 0.84 0.66-1.06 0.14
Large for gestational age 151 (8.4%) 723 (6.4%) 1.33 1.11-1.60 1.24 1.03-1.49 0.02
Meconium stain 99 (5.5%) 995 (8.9%) 0.60 0.48-0.74 0.65 0.52-0.80 <0.01
NICU transfer 70 (3.9%) 261 (2.3%) 1.70 1.30-2.23 1.72 1.31-2.26 <0.01
Apgar score 1 min <7 24 (1.3%) 116 (1.0%) 1.29 0.83-2.02 1.34 0.86-2.09 0.19
Apgar score 5 min <7 3 (0.2%) 14 (0.2%) 1.34 0.38-4.67 1.34 0.38-4.72 0.64
Neonatal death 1 (0.1%) 6 (0.1%) 1.04 0.12-8.66 1.20 0.14-10.09 0.86
* Data presented as number (%).
p value based on multivariable logistic regression.
NICU = neonatal intensive care unit.

www.ejcma.org  403
 F.-C. Chu et al. J Chin Med Assoc

Table 4.
Associations between maternal anemia defined as a Hb <9.9 g/dL and adverse perinatal outcomes
Unadjusted Adjusted
Variables Anemic* Non-anemic* odds ratio 95% CI odds ratio 95% CI p
Total number of pregnancies 628 (4.8%) 12,398 (95.2%)
Maternal outcome
Downloaded from http://journals.lww.com/jcma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW

Cesarean section 295 (46.9%) 4,602 (37.1%) 1.50 1.27-1.76 1.39 1.12-1.72 <0.01
Primary Cesarean section 176 (28.0%) 2,841 (22.9%) 1.31 1.09-1.56 1.32 1.10-1.58 <0.01
Postpartum hemorrhage 11 (1.7%) 201 (1.6%) 1.08 0.58-1.99 1.07 0.58-1.98 0.81
Operative vaginal delivery 27 (4.3%) 570 (4.6%) 0.93 0.62-1.38 1.13 0.76-1.70 0.52
Cervical incompetence 4 (0.6%) 37 (0.3%) 2.14 0.76-6.02 1.71 0.57-5.08 0.33
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC1y0abggQZXdgGj2MwlZLeI= on 10/24/2024

Placental accreta 1 (0.2%) 28 (0.2%) 0.70 0.09-5.18 0.68 0.09-5.09 0.71

Placental abruption 12 (1.9%) 206 (1.7) 1.15 0.64-2.07 1.20 0.66-2.17 0.53
Oligohydramnios 8 (1.2%) 138 (1.1%) 1.14 0.55-2.34 1.32 0.64-2.72 0.44
Polyhydramnios 2 (0.3%) 35 (0.3%) 1.12 0.27-4.72 1.11 0.26-4.68 0.88
GBS infection 77 (12.2%) 1,958 (15.8%) 0.74 0.58-0.95 0.75 0.59-0.96 0.02
Gestational diabetes mellitus 55 (8.8%) 1,171 (9.4%) 0.92 0.69-1.22 0.91 0.68-1.21 0.54
PROM 17 (2.7%) 204 (1.6%) 1.66 1.00-2.74 1.83 1.10-3.02 0.02
Chorioamnionitis 2 (0.3%) 102 (0.8%) 0.38 0.09-1.56 0.49 0.12-2.00 0.32
Neonatal outcome
Early preterm birth <34 weeks 24 (3.8%) 158 (1.3%) 3.07 1.98-4.76 3.01 1.94-4.69 <0.01
Late preterm birth 34–37 weeks 51 (8.1%) 741 (6.0%) 1.39 1.03-1.86 1.31 0.98-1.77 0.07
Very low birth weight 8 (1.3%) 67 (0.5%) 2.37 1.13-4.96 2.45 1.16-5.15 0.01
Low birth weight 44 (7.0%) 709 (5.7%) 1.24 0.90-1.70 1.27 0.92-1.75 0.13
Macrosomia 11 (1.7%) 212 (1.7%) 1.02 0.55-1.88 1.07 0.58-1.98 0.81
Small for gestational age 19 (3.0%) 716 (5.8%) 0.50 0.32-0.80 0.56 0.35-0.89 0.01
Large for gestational age 48 (7.6%) 826 (6.7%) 1.15 0.85-1.56 1.12 0.83-1.53 0.44
Meconium stain, n (%) 40 (6.4%) 1,054 (8.5%) 0.73 0.52-1.01 0.82 0.59-1.13 0.23
NICU transfer 31 (4.9%) 300 (2.4%) 2.09 1.43-3.05 2.15 1.47-3.15 <0.01
Apgar score 1 min <7 6 (1.0%) 134 (1.1%) 0.88 0.38-2.00 0.93 0.40-2.12 0.87
Apgar score 5 min <7 0 (0.0%) 17 (0.1%)
Neonatal death 1 (0.1%) 6 (0.1%) 3.29 0.39-27.40 3.76 0.45-31.43 0.22
* Data presented as number (%).
p value based on multivariable logistic regression.
NICU = neonatal intensive care unit; PROM = premature rupture of membranes.

Our study indicated that multiparity, maternal age >34 years, It is, therefore, challenging to define whether gestational diabetes
a previous history of cesarean delivery, and a previous history mellitus is the cause or consequence of maternal anemia.
of uterine fibroids were the predictors of anemia among the Unexpectedly, we found that infants of anemic mothers
pregnant women. Previous studies reported a similar associa- were significantly more likely to be LGA and less likely to be
tion,2–4,10–12 which may be due to previous bleeding secondary to SGA. Increased numbers of (LGA) infants were demonstrated
surgery, birth, or heavy menstrual bleeding. Interestingly, some to be associated with maternal anemia with an Hb <10.8 g/dL
studies also indicated that pre-pregnancy BMI <18.5 kg/m2 was (Table 3). In contrast, decreased numbers of (SGA) infants were
a predictor of anemia, which may be due to the inadequate found to be associated with maternal anemia with an Hb <9.9 g/
nutrition during pregnancy.2,13 However, we did not find this dL (Table 4). This trend suggests that maternal anemia is associ-
association in our study. This might have been due to variations ated with more LGA infants or fewer SGA infants. However, the
in the methods and study participants involved. These predic- underlying cause of remains unclear. There are two hypotheses
tors of anemia may provide clinical guidance, and these women that may explain this. Firstly, Drukker et al11 hypothesized that
rapid growth of an LGA fetus and its placenta requires abun-
should increase their nutrient intake and take iron supplements.
dant iron, hence inducing maternal anemia. Indeed, maternal
Similar to previous studies,2,3,11–13 we found that women with
Hb level was inversely correlated with the level of cord serum
anemia were at increased risk for adverse perinatal outcomes
erythropoietin, and fetal serum erythropoietin concentrations
including cesarean delivery, primary cesarean delivery, PROM,
were significantly higher in newborns of anemic women than
early preterm birth, VLBW neonates, and NICU admission after in newborns of mothers without anemia.14,15 From this point of
adjustment for the confounding effects from maternal charac- view, maternal anemia is the result of increasing iron demand
teristics. Increased risk of VLBW neonates and NICU admission from the LGA fetus and its placenta.
may be due to increased prevalence of early preterm birth in ane- Alternatively, the second hypothesis is that maternal ane-
mic women. Since we did not compare the perinatal outcome mia may stimulate an increase in fetal growth. Stangret
according to the indications of cesarean section, we cannot evalu- et al16 hypothesized that maternal anemia may upregulate the
ate whether the effect of anemia on cesarean section rates is via angiogenic properties of vascular endothelial growth factor and
maternal, fetal, or combined mechanisms. According to previous placental growth factor, which are exerted through its tyros-
studies, gestational diabetes mellitus may be causative of mater- ine kinase receptors: VEGFR-1 (Flt-1) and VEGFR-2 (Flk-1/
nal anemia and associated with increased cesarean section rate. KDR).16 Hence, better placental perfusion increases the fetal
However, analysis of our results does not show a clear associa- growth. Moreover, some studies found morphological changes
tions between gestational diabetes mellitus and maternal anemia. in the placentas of anemic women due to of increased fetal

404 www.ejcma.org
 Original Article. (2020) 83:4 J Chin Med Assoc

villous capillarization, greater volume and villous diameter, and and associated adverse pregnancy outcomes of anaemia in Chinese
increased number of capillaries per villus cross section.17,18 pregnant women: a multicentre retrospective study. BMC Pregnancy
Stangret also identified a nearly 2-fold increase in expression of Childbirth 2018;18:111.
3. Haider BA, Olofin I, Wang M, Spiegelman D, Ezzati M, Fawzi WW;
the Flt-1 receptor among women in the second trimester, particu- Nutrition Impact Model Study Group (anaemia). Anaemia, prenatal
larly in those presenting with a mild decrease in Hb level (9.7 ≤ Hb iron use, and risk of adverse pregnancy outcomes: systematic review and
≤ 10.8).16 Increased expression of Flt-1, which binds exclusively to meta-analysis. BMJ 2013;346:f3443.
placental growth factor, increased fetoplacental vasculogenesis and 4. Daru J, Zamora J, Fernández-Félix BM, Vogel J, Oladapo OT, Morisaki
Downloaded from http://journals.lww.com/jcma by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AW

angiogenesis. This finding indicates that infants of women with N, et al. Risk of maternal mortality in women with severe anaemia
anemia 1 (Hb < 10.8, mild anemia) are more likely to be LGA than during pregnancy and post partum: a multilevel analysis. Lancet Glob
those of women with anemia 2 (Hb < 9.9, severe anemia). In con- Health 2018;6:e548–54.
5. Basu S, Kumar D, Anupurba S, Verma A, Kumar A. Effect of mater-
clusion, increased vascular density may be an adaptive response to
nal iron deficiency anemia on fetal neural development. J Perinatol
anemia, which inversely benefits the pregnancy outcome. 2018;38:233–9.
nYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC1y0abggQZXdgGj2MwlZLeI= on 10/24/2024

The strengths of our study were as follows: (1) a large sample 6. Milman N. Postpartum anemia I: definition, prevalence, causes, and con-
size was used, (2) blood samples were processed in a uniform sequences. Ann Hematol 2011;90:1247–53.
manner, and (3) the different cutoff levels of Hb level in the 7. Beard JL, Hendricks MK, Perez EM, Murray-Kolb LE, Berg A, Vernon-
population of Taiwanese women were determined. However, Feagans L, et al. Maternal iron deficiency anemia affects postpartum
our study had the following limitations: (1) the Hb data at emotions and cognition. J Nutr 2005;135:267–72.
admission, not from other trimesters, were used for the diagno- 8. Corwin EJ, Murray-Kolb LE, Beard JL. Low hemoglobin level is a risk
factor for postpartum depression. J Nutr 2003;133:4139–42.
sis of anemia. A serial change of Hb level may reflect physiologi- 9. New S, Wirth M. Anaemia, pregnancy, and maternal mortality:
cal and nutritional changes during pregnancy better. However, the problem with globally standardised haemoglobin cutoffs. Bjog
when compared with Hb measurements from other trimesters, 2015;122:166–9.
we believe that the Hb level obtained at admission is a more 10. Cantor AG, Bougatsos C, Dana T, Blazina I, McDonagh M. Routine
reliable parameter for defining the association between anemia iron supplementation and screening for iron deficiency anemia in preg-
and perinatal outcomes. Anemia diagnosed earlier in gestation nancy: a systematic review for the U.S. Preventive services task force.
may have been resolved by treatment or cessation of physiologic Ann Intern Med 2015;162:566–76.
11. Drukker L, Hants Y, Farkash R, Ruchlemer R, Samueloff A, Grisaru-
anemia. However, lack of data on patient’s intake of multivi-
Granovsky S. Iron deficiency anemia at admission for labor and delivery
tamins and micronutrients including iron and folic acid sup- is associated with an increased risk for cesarean section and adverse
plements during pregnancy could be a limitation. The causal maternal and neonatal outcomes. Transfusion 2015;55:2799–806.
relationship between maternal anemia and adverse perinatal 12. Tunkyi K, Moodley J. Anemia and pregnancy outcomes: a longitudinal
outcomes cannot be determined by this study. It will be more study. J Matern Fetal Neonatal Med 2018;31:2594–8.
informative if serial measurements of Hb levels throughout 13. Patel A, Prakash AA, Das PK, Gupta S, Pusdekar YV, Hibberd PL.
gestation are available to clarify the association between mater- Maternal anemia and underweight as determinants of pregnancy out-
nal anemia and adverse perinatal outcomes. (2) The difference comes: cohort study in eastern rural Maharashtra, India. BMJ Open
2018;8:e021623.
between normocytic and microcytic or macrocytic anemia had
14. Erdem A, Erdem M, Arslan M, Yazici G, Eskandari R, Himmetoglu
not been elucidated in this study. However, we believe that this O. The effect of maternal anemia and iron deficiency on fetal eryth-
does not mitigate the value of the association. (3) Anemia seems ropoiesis: comparison between serum erythropoietin, hemoglobin and
to increase the risk of cesarean section and primary cesarean sec- ferritin levels in mothers and newborns. J Matern Fetal Neonatal Med
tion. However, the perinatal outcome was not compared accord- 2002;11:329–32.
ing to the different indications of cesarean section. Hence, we 15. Beguin Y, Lipscei G, Oris R, Thoumsin H, Fillet G. Serum immunoreac-
cannot evaluate whether the effect of anemia on cesarean sec- tive erythropoietin during pregnancy and in the early postpartum. Br J
tion rates is via maternal, fetal, or combined mechanisms. Haematol 1990;76:545–9.
16. Stangret A, Wnuk A, Szewczyk G, Pyzlak M, Szukiewicz D. Maternal
hemoglobin concentration and hematocrit values may affect fetus devel-
opment by influencing placental angiogenesis. J Matern Fetal Neonatal
REFERENCES Med 2017;30:199–204.
1. WHO. Hemoglobin concentrations for the diagnosis of anemia and 17. Charnock-Jones DS, Burton GJ. Placental vascular morphogenesis. Best
assessment of severity. World Health Organization; 2014. Pract Res Clin Obstet Gynaecol 2000;14:953–68.
2. Lin L, Wei Y, Zhu W, Wang C, Su R, Feng H, et al.; Gestational diabetes 18. Torry DS, Hinrichs M, Torry RJ. Determinants of placental vascularity.
mellitus Prevalence Survey (GPS) study Group. Prevalence, risk factors Am J Reprod Immunol 2004;51:257–68.

www.ejcma.org  405