Association between fetal sex, birthweight percentile and adverse pregnancy outcome

Abstract Introduction The objective was to evaluate the association between fetal sex and adverse pregnancy outcome, while correcting for fetal growth and gestational age at delivery. Material and methods Data from the Netherlands Perinatal Registry (1999‐2010) were used. The study population comprised all white European women with a singleton delivery between 25+0 and 42+6 weeks of gestation. Fetuses with structural or chromosomal abnormalities were excluded. Outcomes were antepartum death, intrapartum/neonatal death (from onset of labor until 28 days after birth), perinatal death (antepartum death or intrapartum/neonatal death), a composite of neonatal morbidity (including infant respiratory distress syndrome, sepsis, necrotizing enterocolitis, meconium aspiration, persistent pulmonary hypertension of the newborn, periventricular leukomalacia, Apgar score <7 at 5 minutes, and intracranial hemorrhage) and a composite adverse neonatal outcome (perinatal death or neonatal morbidity). Outcomes were expressed stratified by birthweight percentile (p90 [large for gestation]) and gestational age at delivery (25+0‐27+6, 28+0‐31+6, 32+0‐36+6, 37+0‐42+6 weeks). The association between fetal sex and outcome was assessed using the fetus at risk approach. Results We studied 1 742 831 pregnant women. We found no increased risk of antepartum, intrapartum/neonatal and perinatal death in normal weight and large‐for‐gestation males born after 28+0 weeks compared with females. We found an increased risk of antepartum death among small‐for‐gestation males born after 28+0 weeks (relative risk [RR] 1.16‐1.40). All males born after 32+0 weeks of gestation suffered more neonatal morbidity than females regardless of birthweight percentile (RR 1.07‐1.34). Infant respiratory distress syndrome, sepsis, persistent pulmonary hypertension of the newborn, Apgar score <7 at 5 minutes, and intracranial hemorrhage all occurred more often in males than in females. Conclusions Small‐for‐gestation males have an increased risk of antepartum death and all males born after 32+0 weeks of gestation have an increased risk of neonatal morbidity compared with females. In contrast to findings in previous studies we found no increased risk of antepartum, intrapartum/neonatal or perinatal death in normal weight and large‐for‐gestation males born after 28+0 weeks.


| INTRODUC TI ON
In pregnancy, fetal sex is known to affect placentation, 1 intrauterine growth, 2 preterm birth [3][4][5][6][7][8][9] and perinatal outcome. [10][11][12][13] Previous studies have suggested a male predominance in miscarriage, 10 antepartum death, [11][12][13] perinatal death, fetal distress, respiratory distress syndrome and low Apgar scores. 14 Although these studies provide valuable information about gender differences, outcomes should be interpreted with care because of methodological weaknesses. First, in some studies, outcomes were adjusted for absolute birthweight and gestational age at delivery while not taking into account that healthy males are on average heavier than healthy females at any given gestational age. 2 This may lead to comparison of small-for-gestation (SGA) males with normal weight females and falsely suggest a higher risk of adverse outcome among males. In other studies, outcomes were based on calculated male to female ratio for adverse outcome, expressed as the number of adverse outcomes among males divided by the number of adverse outcomes among females. Since, historically, more males are born, this comparison may also falsely suggest a higher risk of adverse outcome among males. Secondly, previous studies did not investigate whether there was an association between gestational age at delivery, fetal growth and perinatal outcome.
These methodological limitations can be overcome when birthweight percentiles are used to express growth instead of absolute birthweight, stratification for SGA, normal weight and large-for-gestational-age (LGA) males, females and gestational age at delivery is performed, and the fetus at risk approach is used to rule out bias through unequal numbers of male and female infants.
The objective of this hypothesis-generating study was to evaluate the association between fetal sex and adverse pregnancy outcome, including antepartum death, intrapartum/neonatal death and neonatal morbidity, whilst correcting for fetal growth and gestational age at delivery.

| MATERIAL AND ME THODS
This study was performed in a nationwide cohort with the use of the Netherlands Perinatal Registry (PERINED). The PERINED consists of population-based data that contain information on pregnancies, deliveries and re-admissions until 28 days after birth. The PERINED database is obtained by a validated linkage of three different registries: the midwifery registry, the obstetrics registry and the neonatology registry of hospital admissions. 15,16 Records are entered in the PERINED registry at the child's level. The coverage of the PERINED registry is approximately 96% of all deliveries in the Netherlands. It contains pregnancies of ≥22 weeks' gestational age and is used primarily for annual assessment of the quality indicators of obstetric care.
We included all white European women who delivered a singleton baby between 25 +0 and 42 +6 weeks of gestation in The Netherlands between 1 January 1999, and 31 December 2010.
We excluded all women who delivered an infant with congenital anomalies. 17 Women of other ethnicities were excluded to avoid bias through differences in optimal weight for gestation in other groups not taken into account in the Dutch birthweight reference curves and to avoid bias through the presence of different risk profiles for adverse pregnancy outcomes among non-white European women.

Conclusions:
Small-for-gestation males have an increased risk of antepartum death and all males born after 32 +0 weeks of gestation have an increased risk of neonatal morbidity compared with females. In contrast to findings in previous studies we found no increased risk of antepartum, intrapartum/neonatal or perinatal death in normal weight and large-for-gestation males born after 28 +0 weeks.  19 and socioeconomic status. Parity was categorized into nulliparous and multiparous women. The socioeconomic status score is based on mean income level, the percentage of households with a low income, the percentage of inhabitants without a paid job and the percentage of households with on average a low education in a postal code area. 20 The continuous socioeconomic status score was categorized into a low, middle and high group, based on percentile ranges (<25th percentile, 25-75th percentile, >75th percentile).
Demographic and obstetric baseline characteristics were compared between males and females using Student's t test and Chisquare test as appropriate. We tested for interaction between fetal sex and gestational age at delivery and between fetal sex and birthweight percentile. These tests were performed separately for all outcome measures. If interaction was found to be present (defined as a P < 0.001), analyses were performed, stratified for gestational age at delivery (25 +0 -27 +6 weeks, 28 +0 -31 +6 weeks, 32 +0 -36 +6 weeks, 37 +0 -42 +6 weeks) and birthweight percentile group (SGA, normal weight and LGA).
The fetus at risk approach was used to determine the association between fetal sex and pregnancy outcome, expressed as relative risk (RR) with 95% confidence intervals (CI) within the strata for birthweight percentile and gestational age at delivery.

| Antepartum death
For a fetus to be at risk of antepartum death (AD) at 36 +0 weeks it is necessary to be alive at 36 +0 weeks. Consequently, the risk of antepartum death was calculated as a proportion of the ongoing pregnancies (OP) at a particular gestation.
Risk of antepartum death at week n = AD n /OP n *100.

| Intrapartum/neonatal death
In concurrence, the risk of intrapartum/neonatal death (IND) at any gestational age is obtained by dividing the number of intrapartum and neonatal deaths at that gestation by the number of neonates at risk of intrapartum/neonatal death at that gestation. The neonates at risk of intrapartum/neonatal death at a certain gestational age include all delivered neonates (DN) that did not die antepartum, meaning all pregnancies with onset of labor.
Risk of intrapartum/neonatal death at week n = IND n /DN n *100.

| Perinatal death
The risk of perinatal death (PND) at any gestational age is obtained by dividing the number of neonates with perinatal death (antepartum or intrapartum/neonatal death) in each stratum by the total number of deliveries in each birthweight category.
Risk of perinatal death at week n = PND n /DN n *100.

| Composite neonatal morbidity
The denominator used for the risk of neonatal morbidity (NM) is all alive fetuses (AF) born at that gestation, since only alive fetuses can suffer morbidity.
Risk of neonatal morbidity at week n = NM n /AF n *100.

| Composite adverse perinatal outcome
The numerator is all neonates with perinatal death (PND) or morbidity (NM) in each stratum and the denominator all deliveries in each birthweight category (DN).
Risk of adverse perinatal outcome at week n = (PND n + NM n )/ DN n *100.
A post-hoc analysis for all separate components of morbidity was performed to differentiate further which outcome contributes most to the found difference between male and female neonates.
Data were analyzed with the SAS statistical software package version 9.2 (SAS Institute Inc., Cary, NC, USA).

| Ethical approval
The data in the perinatal registry are anonymous and exempt from ethical approval. The Netherlands Perinatal Registry gave approval for the use of the data for this study (approval number 13.73). Baseline characteristics of the cohort are presented in Table 1.

| RE SULTS
There were more male (n = 895 272 [51.4%]) than female infants (n = 847 559 [48.6%]) in the cohort. There were no statistically significant differences in maternal baseline characteristics between the two groups. However, women with a male fetus were more likely to have a hypertensive complication of pregnancy compared with women with a female fetus (9.3 vs 9.0%, P < 0.001). Women with a male fetus were more likely (all P < 0.001) than women with a female fetus to undergo an emergency cesarean section (8.5 vs 7.1%) or vaginal instrumental delivery (12.2% vs 10.0%). The rate of preterm delivery (<37 weeks of gestation) was higher among males than among females (6.4 vs 5.4%).
Interaction between fetal sex and gestational age at delivery was statistically significant for antepartum death (P < 0.001), intrapartum/neonatal death (P < 0.001) and neonatal morbidity (P < 0.001).

| Composite morbidity
The composite morbidity of males and females is shown in Table 5.
The risk of composite morbidity was significantly increased in males compared with females in most (8/12) strata, with relative risks ranging from 1.07 to 1.36. All males born after 32 +0 weeks of gestation have an increased risk of neonatal morbidity compared with females.
The analysis for all infants (not stratified into SGA, normal weight and LGA) shows that composite morbidity was significantly increased in males compared with females born after 28 +0 weeks, with relative risks ranging from 1.07 to 1.34.

| Composite adverse perinatal outcome
The composite adverse neonatal outcome of males and females is shown in Table 6.
The risk of composite adverse neonatal outcome was significantly increased in males compared with females in most (7/12) The analysis for all infants (not stratified into SGA, normal weight and LGA) shows that composite morbidity was significantly increased in males compared with females born after 28 +0 weeks, with relative risks ranging from 1.04 to 1.28.
The individual components of the composite neonatal morbidity (IRDS, sepsis, NEC, meconium aspiration, PPHN, periventricular leukomalacia, Apgar score <7 at 5 minutes and ICH) were also analyzed separately (Tables S1-S8). IRDS was increased in males compared with females at most gestational ages, regardless of birthweight percentile (Table S1). Sepsis was increased in all term males compared with females regardless of birthweight. Between 32 +0 and 36 +6 weeks the risk of sepsis was only increased in normal weight males (Table S2). PPHN was only increased in normal weight males at term compared with females (Table S5). For periventricular leukomalacia, although in only one stratum, a decreased risk was found for SGA males between 25 +0 and 27 +0 weeks of gestation (Table S6).
The risk of ICH was increased in all normal weight males compared with females regardless of gestational age at delivery and in LGA males born between 28 +0 and 31 +6 weeks of gestation (RR 1.52, 95% CI 1.08-2.13) ( Table S6). The incidence of an Apgar score <7 at 5 minutes was increased in all term males compared with females regardless of birthweight. Furthermore, SGA males between 25 +0 and 27 +0 and 32 +0 and 36 +6 weeks of gestation were at increased risk (Table S7). The incidence of NEC (Table S3) and meconium aspiration (Table S4) was not significantly different between males and females.

| D ISCUSS I ON
In our study, we analyzed 1 742 831 singleton deliveries and assessed differences in pregnancy outcomes between males and females after adjustments for differences in fetal growth. SGA males have an increased risk of antepartum death and all males born after 32 +0 weeks of gestation have an increased risk of neonatal morbidity compared with females. Differences in neonatal morbidity are mainly caused by increased risks of sepsis, IRDS and ICH in males compared with females.
Our study has some limitations. These limitations are mainly related to the fact that we performed a database study and consequently had to rely on information that was recorded in the registry.
To assess differences in the incidence of adverse outcome, large numbers are needed. In our opinion, it would not have been feasible to collect enough data using a different study design. Women with a male fetus were more likely to have a hypertensive complication of pregnancy compared with women with a female fetus (9.3 vs 9.0%, P < 0.001). This is in accordance with findings in previous research. 22 The association might be term-dependent, but this is outside the scope of this study. TA B L E 4 Perinatal death rate in males and females by birthweight percentiles and gestational age category length and weight, and placental weight and pathology are not registered in the Dutch Perinatal Registry. We do not expect a systematical bias.
Furthermore, ultrasound growth charts used to detect SGA fetuses are not gender-specific. Therefore it is possible that girls (who are by definition smaller than male fetuses) are more often SGA compared with boys, perhaps leading to earlier intervention (induction) and thereby preventing antepartum death. This possible bias cannot be ruled out and should be investigated in further studies.
Our data did not allow us to make statements about the influence of mode of delivery on neonatal death and morbidity. Table 1 shows differences in induction of labor and instrumental deliveries between males and females. These are often a result of suspected compromised fetal condition rather than a confounder of adverse outcome. We can only hypothesize that the higher rate of vaginal instrumental delivery among males might be associated with the increased risk of ICH among males; another explanation could be that the ICH occurred antenatally and caused fetal distress. This is not within the scope of this paper and should be investigated in future studies. Finally, there might be an association between the higher incidence of cesarean sections among males and the risk of IRDS, although information on timing of the cesarean section and administration of betamethasone is lacking.
Finally, some causes of neonatal morbidity are predominantly a problem in the extreme preterm period, whereas others occur at all gestational ages or mainly at term. We decided to use a composite outcome containing important causes of neonatal morbidity available in the PERINED registry to provide useful information for clinicians. We hypothesize that genetic differences between males and females might underlie the fetal sex-related differences. This study gives no clues on how to decrease neonatal morbidity in male infants.
The results of this study matter because they correct the erroneous idea that all male infants suffer more antepartum, intrapartum/neonatal and perinatal death than female infants. The main Abbreviations: CI, confidence interval; GA, gestational age; LGA, large-for-gestational age; RR, relative risk ratio; SGA, small-for-gestational age. a RR calculated as: the numerator are all neonates with perinatal death (antepartum or intrapartum/ neonatal death) or morbidity in each stratum and the denominator all deliveries in each birthweight category.
TA B L E 6 Composite adverse perinatal outcome rate in males and females by birthweight percentiles and gestational age category implication of this study is awareness of differences between male and female perinatal outcome. Our study has further focused on which domain these differences lie.
Further intervention studies should be aware of sex differences in perinatal outcome and should perform prespecified analyses separately for male and female infants. The importance of this advice has already been shown in a recent intervention study that found differences in effectiveness of allopurinol between male and female fetuses. 33 This could lead to a more tailored approach fit for the individual neonate to improve outcomes. In addition, future research could be aimed at unraveling mechanisms that might play a role in the increased neonatal morbidity in males.

| CON CLUS ION
SGA males have an increased risk of antepartum death and all males born after 32 +0 weeks of gestation have an increased risk of neonatal morbidity compared with females. Differences in neonatal morbidity are mainly caused by increased risks of IRDS, sepsis, PPHN, Apgar score <7 at 5 minutes and ICH in males compared with females. In contrast to findings in previous studies, we found no increased risk of antepartum, intrapartum/neonatal and perinatal death in normal weight and LGA males born after 28 +0 weeks.

ACK N OWLED G M ENTS
We thank all Dutch midwives, obstetricians, neonatologists and other perinatal healthcare providers for the registration of perinatal information and the Foundation of the Netherlands Perinatal Registry (www.perin atreg.nl) for permission (13.73) to use the registry data.