Article Text

Original research
Association of different doses of antenatal corticosteroids exposure with early major outcomes and early weight loss percentage in extremely preterm infants or extremely low birthweight infants: a multicentre cohort study
  1. Shuaijun Li1,2,
  2. Qi Feng2,
  3. Xiaofang Huang2,
  4. Xiuying Tian3,
  5. Ying Zhou4,
  6. Yong Ji5,
  7. Shufen Zhai6,
  8. Wei Guo7,
  9. Rongxiu Zheng8,
  10. Haijun Wang9
  1. 1Department of Maternal and Child Health, School of Public Health, Peking University Health Science Center-Weifang Joint Research Center for Maternal and Child Health, Peking University, Beijing, China
  2. 2Department of Pediatrics, Peking University First Hospital, Beijing, China
  3. 3Department of Neonatology, Tianjin Central Hospital of Obstetrics and Gynecology, Tianjin, China
  4. 4Department of Pediatrics, Peking University Third Hospital, Beijing, China
  5. 5Neonatal Intensive Care Unit, Children’s Hospital of Shanxi, Taiyuan, China
  6. 6Department of Neonatology, Handan Central Hospital, Handan, China
  7. 7Department of Neonatology, Xingtai People’s Hospital, Xingtai, China
  8. 8Department of Pediatrics, Tianjin Medical University General Hospital, Tianjin, China
  9. 9Department of Maternal and Child Health, National Health Commission Key Laboratory of Reproductive Health, Peking University Health Science Center-Weifang Joint Research Center for Maternal and Child Health, Peking University School of Public Health, Beijing, China
  1. Correspondence to Professor Qi Feng; fengqizf{at}126.com

Abstract

Objectives To determine the dose-dependent associations between antenatal corticosteroids (ANS) exposure and the rates of major morbidities, and the early weight loss percentage (EWLP) in hospital among extremely preterm infants (EPI) or extremely low birthweight infants (ELBWI).

Methods A multicentre, retrospective cohort study of EPI or ELBWI born between 2017 and 2018 was conducted. Infants were classified into no ANS, partial ANS and complete ANS exposure group; three subgroups were generated by gestational age and birth weight. Multiple logistic regression and multiple linear regression were performed.

Results There were 725 infants included from 32 centres. Among no ANS, partial ANS and complete ANS exposure, there were significant differences in the proportions of bronchopulmonary dysplasia (BPD) (24.5%, 25.4% and 16.1%), necrotising enterocolitis (NEC) (6.7%, 2.0% and 2.0%) and death (29.6%, 18.5% and 13.5%), and insignificant differences in the proportions of intraventricular haemorrhage (IVH) (12.5%, 13.2% and 12.2%), and extrauterine growth restriction (EUGR) (50.0%, 56.6% and 59.5%). In the logistic regression, compared with no ANS exposure, complete ANS reduced the risk of BPD (OR 0.58, 95% CI 0.37 to 0.91), NEC (OR 0.21, 95% CI 0.08 to 0.57) and death (OR 0.36, 95% CI 0.23 to 0.56), and partial ANS reduced the risk of NEC (OR 0.23, 95% CI 0.07 to 0.72) and death (OR 0.54, 95% CI 0.34 to 0.87). Compared with partial ANS exposure, complete ANS decreased the risk of BPD (OR 0.58, 95% CI 0.37 to 0.91). There were insignificant associations between ANS exposure and IVH, EUGR. In the multiple linear regression, partial and complete ANS exposure increased EWLP only in the ≥28 weeks (w) and <1000 g subgroup (p<0.05).

Conclusions Different doses of ANS (dexamethasone) exposure were protectively associated with BPD, NEC, death in hospital, but not EUGR at discharge among EPI or ELBWI. Beneficial dose-dependent associations between ANS (dexamethasone) exposure and BPD existed. ANS exposure increased EWLP only in the ≥28 w and<1000 g subgroup. ANS administration, especially complete ANS, is encouraged before preterm birth.

Trial registration number NCT06082414.

  • growth
  • neonatology
  • mortality

Data availability statement

Data sharing not applicable as the confidentiality agreement of the study existed.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • Antenatal corticosteroids (ANS) accelerate fetal lung maturation, reduce morbidities and mortality; however, evidence is limited on dose-dependent associations between ANS and major morbidities, especially from China, and data are few regarding the effects of ANS on the early weight loss percentage (EWLP) after birth in extremely preterm infants (EPI) or extremely low birthweight infants (ELBWI).

WHAT THIS STUDY ADDS

  • We found the beneficial dose-dependent association between ANS exposure and bronchopulmonary dysplasia.

  • Partial ANS and complete ANS exposure increased EWLP only in the ≥28 weeks (w) and <1000 g subgroup; however, different doses of ANS exposure had no significant effects on extrauterine growth restriction in EPI or ELBWI.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • The findings contribute the evidence from China regarding the effects of different doses of ANS exposure on major morbidities and weight changes in EPI or ELBWI.

  • They also help to promote the administration of ANS, especially complete ANS, before preterm birth in clinical performances.

Introduction

International guidelines recommend that antenatal corticosteroids (ANS) can promote fetal lung maturation in pregnant women at risk of preterm birth at 24–34 weeks of gestation, reduce respiratory complications,1 mortality1 2 and nervous morbidities3 in premature infants. A previous systematic review also found that ANS reduces the incidences of death, respiratory distress syndrome, intraventricular haemorrhage (IVH) and developmental delay during childhood with premature infants.4 Limited data exist on the dose-dependent effect of ANS on outcomes for preterm infants. ANS has a dose-response effect against death5 6 and neurodevelopmental5 impairment in preterm infants. China has the second highest number of preterm infants worldwide.7 However, evidence from China is scarce on the dose-dependent associations between ANS and major morbidities in preterm infants.

ANS has been proven to negatively affect fetal growth. There is a dose-response association between ANS and fetal growth restriction.8 An animal study found that the rate of intrauterine growth restriction, defined as fetal weight below the 10th percentile, increases with an increasing number of ANS doses.9 Birth weights of preterm lambs gradually decreases with an increasing dose of ANS.10 Additionally, ANS is associated with postnatal growth in preterm infants in hospital. Luo et al11 used univariate analysis to report that ANS decreases the changes in weight Z-score at discharge in preterm infants.10 ANS relates to extrauterine growth restriction (EUGR, discharged weight <10th percentile) in preterm infants,12 which associates with long-term growth impairment,13 poor neurodevelopment14 and cardiometabolic alterations.15

The early postnatal weight loss, usually occurring within the first 2 weeks after birth, is associated with morbidities and death in preterm infants. The early postnatal weight loss of <3% or >12% on the third day after birth reduces mortality, IVH16 and >15% in the first week reduces necrotising enterocolitis (NEC) in preterm infants.17 However, data are few regarding the dose-dependent effects of ANS on the early weight loss after birth in preterm infants.

Therefore, the multicentre cohort study in Northern China was designed with the objectives: to determine associations between different doses of ANS exposure and major morbidities, and the early weight loss percentage (EWLP) following birth among extremely preterm infants (EPI, gestational age (GA) <28 weeks (w)) or extremely low birthweight infants (ELBWI, birth weight <1000 grams (g)) in hospital.

Methods

Design

This multicentre retrospective cohort study collected clinical data of EPI or ELBWI admitted to 33 level III neonatal intensive care units (NICUs) from Beijing (12 NICUs, the number of beds is 20–70), Tianjin (4 NICUs, the number of beds is 20–80), Hebei Province (8 NICUs, the number of beds is 20–110), Shanxi Province (3 NICUs, the number of beds is 30–180) and Inner Mongolia Autonomous Region (6 NICUs, the number of beds is 15–40) in Northern China from 1 January 2017 to 31 December 2018. Data on demographic, morbidities, nutrition, ANS and in-hospital weight were extracted from the electronic medical record by trained neonatologists. The detailed process had been described in the previous study.18 19

Inclusion and exclusion criteria of participants

  1. Inclusion criteria: infants with a GA <28 w, or a birth weight (BW)<1000 g, admitted within 72 hours after birth.

  2. Exclusion criteria: infants with severe congenital heart disease, congenital digestive tract or nerve system malformations, inborn errors of metabolism, no information on ANS exposure or minimum weight after birth.

Classification of groups

EPI or ELBWI were classified into three groups by ANS administration: no ANS, partial ANS and complete ANS exposure group. According to GA and BW, three subgroups were generated: <28 w and <1000 g subgroup, <28 w and ≥1000 g subgroup and ≥28 w and <1000 g subgroup.

Definitions of variables and outcomes

Partial ANS exposure: an infant’s mother received at least one dose of 5–6 mg of dexamethasone injection by muscular at 1 day to 7 days prior to delivery. Complete ANS: the mother received ≥4 doses of dexamethasone.20 21 Participating centres in this study administered dexamethasone to mothers at risk of preterm birth. Small for gestational age (SGA): BW was below the 10th percentile for GA and sex.22 NEC: ≥Bell stage II.23 Pulmonary surfactant (PS) use: PS was used within 72 hours after birth. Bronchopulmonary dysplasia (BPD): oxygen support was required at 36 weeks’ corrected age. Late-onset sepsis: blood culture or body fluid culture was positive >72 hours after birth. Retinopathy of prematurity (ROP): any stage of ROP. IVH: ≥grade 3.24 Nutrition protocol: the nutrition protocol was used based on the guideline for clinical practice of nutrition support in Chinese neonates. Duration of parenteral nutrition was referred as the duration of amino acid administration by a venous access. Weight loss percentage was calculated as (birth weight–minimal weight)/birth weight)×100%. The minimal weight usually emerged within the first 2 weeks after birth. EUGR was defined as weight below the 10th percentile at discharge.25 Major morbidities in the study mainly referred to BPD, NEC, IVH, EUGR and death in hospital.

Statistical analyses

Continuous variables for normal distribution, such as GA, BW, maternal age and so forth were depicted as mean±SD, with differences among groups analysed by one-way analysis of variance analysis. Categorical variables, such as male infant, SGA, hypertensive disorders of pregnancy (HDP) and so forth were described as percentage, with differences compared using χ2 analysis. P value was adjusted with Bonferroni method for pairwise comparisons, in which p value <0.017 was considered as statistically significant.

Multiple logistic regression was performed to determine the associations between different doses of ANS exposure and BPD, NEC, IVH, EUGR and death, adjusting for confounders (infants’ and mothers’ variables). Multiple linear regression analysis was performed to determine the associations between different doses of ANS exposure and EWLP. I2 was applied to estimate the heterogeneity among subgroups. Data lacking information on ANS exposure or minimal weight were excluded. P<0.05 was considered statistically significant. Statistical analyses were conducted using SPSS (V.23.0). Heterogeneity analysis and forest plot were performed with R software (V.4.2.1).

Results

A total of 952 EPI or ELBWI were admitted to 33 NICUs from 1 January 2017 to 31 December 2018, of which 1 centre and 227 infants were excluded for no information on ANS or minimal weight. Therefore, 725 infants from 32 NICUs were included in this study (figure 1).

Figure 1

Study flow chart. ANS, antenatal corticosteroids; ELBWI, extremely low birth weight infants; EPI, extremely preterm infants; NICUs, neonatal intensive care units.

Table 1 shows the infants in the complete ANS exposure groups had a slightly higher GA respectively compared with those with no ANS and partial ANS exposure (p<0.017). The infants with complete ANS and partial ANS exposure were more likely to delivered by cesarean compared to those with no ANS exposure (p<0.017). Compared with partial ANS exposure, complete ANS exposure had a higher proportion of SGA (p<0.017).

Table 1

Characteristics of infants and mothers categorised by ANS exposure

As shown in table 2, no statistical significance was found among different doses of ANS exposure in weight loss and weight loss percentage. Compared with no ANS exposure, partial ANS prolonged days to regain birth weight, the duration of parenteral nutrition and the duration of hospital stay (p<0.017). Compared with partial ANS exposure, complete ANS decreased the duration of hospital stay (p<0.017). Compared with no ANS exposure, complete ANS decreased the rate of NEC and death in hospital (p<0.017) and partial ANS decreased the mortality (p<0.017). Compared with partial ANS exposure, complete ANS decreased the rate of BPD (p<0.017).

Table 2

In-hospital weight, interventions and major morbidities by ANS exposure

In multiple logistic regression, compared with no ANS exposure, complete ANS reduced the risk of BPD, NEC and death in hospital (p<0.05), and partial ANS reduced the risk of NEC and death (p<0.05). Compared with partial ANS exposure, complete ANS decreased the risk of BPD (OR=0.58, 95%CI 0.37-0.91, p<0.05) (table 3).

Table 3

Multiple logistic regression analysis for major morbidities and death by ANS exposure

In multiple linear regression, complete ANS exposure was positively associated with EWLP (p=0.001). No statistical association was found between partial ANS and complete ANS exposure (table 4).

Table 4

Multiple linear regression analysis for early weight loss percentage after birth by ANS exposure

Table 5 shows that compared with no ANS exposure, both complete and partial ANS exposure were positively associated with EWLP (p<0.05), and no significant effects existed in the comparison between complete and partial ANS exposure in the ≥28 w and <1000 g subgroup (p>0.05). In the partial ANS exposure group, no significant differences existed when comparisons were performed among the three subgroups (p>0.05) (table 5).

Table 5

Multiple linear regression analysis for early weight loss percentage after birth in GA and BW subgroups by ANS exposure

In the complete ANS exposure group, compared with the <28 w and <1000 g subgroup, the ≥28 w and <1000 g subgroup was close to associating with EWLP (p=0.05, β=2.14, 95% CI 0.00 to 4.28). Compared with the <28 w and ≥1000 g subgroup, the <28 w and <1000 g and ≥28 w and <1000 g subgroups showed no significant effects (p>0.05) (table 5).

In the complete ANS exposure group, no heterogeneity existed between the <28 w and <1000 g, and <28 w and ≥1000 g subgroups (I2=0.00%, p=0.921). Therefore, we combined the <28 w and <1000 g, and <28 w and ≥1000 g subgroups into the <28 w subgroup (<1000 g or ≥1000 g) (figure 2). The heterogeneity between the <28 w subgroup and the ≥28 w subgroup (<1000 g) existed (I2=61.48%, p=0.107) (figure 3).

Figure 2

The heterogeneity analysis and forest plot between the <28 weeks (w) and ≥1000 g, and <28 w and <1000 g subgroups in the complete antenatal corticosteroids exposure group.

Figure 3

The heterogeneity analysis and forest plot between the ≥28 weeks (w) (and <1000 g) subgroup, and <28 w (<1000 g or ≥1000 g) subgroups in the complete antenatal corticosteroids exposure group.

In the partial ANS exposure group, no heterogeneities existed among subgroups (see online supplemental figures 1 and 2).

Supplemental material

Supplemental material

Discussion

In the multicentre cohort study, we found that partial ANS and complete ANS exposure significantly decreased the risk of NEC and death, and a protective dose-dependent association was found between ANS exposure and BPD among EPI or ELBWI. Different doses of ANS did not influence IVH and EUGR at discharge. Compared with no ANS exposure, partial ANS and complete ANS exposure increased EWLP only in the ≥28 w and <1000 g subgroup.

Chawla et al5 found that partial ANS and complete ANS exposure do not significantly reduce the BPD incidence in infants weighing 401–1000 g and/or at 22–27 w compared with no ANS exposure. Carlo et al26 built two groups by ANS exposure and found that compared with no ANS, ANS increases the BPD incidence in infants at 22–25 w (60.3% vs 50.4%, p<0.05). We found that complete ANS was protectively against the risk of BPD. Chawla et al did not adjust for BW and HDP, which are associated with BPD.27 Carlo et al did not explore the dose-dependent associations between ANS exposure and BPD. Different adjusted variables and groups might cause the disparities.

To our knowledge, the effects of ANS exposure on NEC are inconsistent. Travers et al2 found that ANS exposure significantly increases the risk of NEC ≥Bell stage II in infants at 23 w, and decreases the risk in infants at 27 w, without significant effects on infants at 24–26 w and 28–34 w. Chawla et al5 revealed that partial ANS exposure significantly increases the risk of NEC, and complete ANS exhibits no significant effects. In our study, partial ANS and complete ANS decreased the risk of NEC. SGA is a risk factor of NEC.28 SGA is not adjusted for in studies by Travers et al and Chawla et al, which may produce the disparity between our results and theirs. Additionally, the different types of ANS may contribute to the difference, where we used dexamethasone, and studies by Travers et al and Chawla et al used dexamethasone or betamethasone.

Chawla et al5 found that EPI with complete ANS has a significantly lower risk of IVH ≥grade 3 than those with no or partial ANS. In our study, insignificant effects of different doses of ANS on IVH ≥grade 3 were found. The proportions of IVH in different doses of ANS exposure in our study are lower than those in the study by Chawla et al, which may cause the disparity.

Li et al12 reported that both partial and complete ANS exposure significantly reduce the incidence of EUGR in infants <32 w in hospital. Their mean GA is 30 w, and BW is 1300 g. We found no significant associations between partial and complete ANS exposure and EUGR. The disparity may be attributed to the lower GA and BW in our study than those in the study by Li et al.

One cause of ANS affecting EWLP might be partially explained by ANS adjusting water changes. ANS exposure promotes diuresis by increasing the glomerular filtrate rate.29 Dimitriou et al30 found that receiving similar median fluid input, infants with ANS exposure have lower median insensible water volume, higher median urine output than those without ANS on day 1 after birth. The ANS exposure group has more EWLP (−1.05 (95% CI −10.6 to 20.4) vs +1.18 (28.9 to 17.8), p=0.066) on day 2. However, the study by Dimitriou et al had a small sample (n=96), measured the fluid volume and weight change only within 2 days after birth and conducted univariate analysis. Hammarlund et al31 found infants with 25–27 w have more transepidermal water loss than infants with 28–30 w in the first 30 days after birth. We found that partial ANS and complete ANS exposure positively associated with EWLP only in the ≥28 w and <1000 g subgroup. Different GA might result in differences among the subgroups by the heterogeneity analysis. Another cause of ANS affecting EWLP may be that ANS influences the adrenal gland which regulates infants’ thriving in utero,32 and infants’ postnatal growth.33 Level of cortisol in blood transparently reflects the function of the hypothalamic-pituitary-adrenal (HPA) axis. To our knowledge, the clinical study on the effect of cortisol in EPI or ELBWI is few because of the ethics on blood draw from those infants. An earlier study revealed that compared with those with no ANS exposure, infants>35 w with ANS (betamethasone) exhibit the similar level of cortisol at the age of 24 hours by a 2-hour adrenocorticotropic hormone test. However, a small-sample study (n=18) confirmed that ANS (betamethasone) suppresses the HPA axis response to a stressor, resulting in cortisol failing to increase in infants at 33–34 w on days 3–6 of life.34 Moreover, an animal study demonstrated exogenous ANS crosses through the placenta to affect the fetus’s HPA axis,35 and the effect continues into adulthood.36 The detailed causes are unknown yet. Other factors, such as nutrient supply, fluid volume and other exposures may be potential reasons.

ANS reduces mortality in preterm infants. Travers et al6 found that any ANS exposure significantly decreases in-hospital mortality in EPI. A systematic review found that complete ANS exposure reduces neonatal death.4 We also found similar results in the different doses of ANS exposure groups.

The study had some limitations. This observational cohort study demonstrated the associations between ANS exposure and major morbidities and mortality, and weight change in preterm infants, but could not make causal inference. Additionally, ANS exposure possibly affected body length and head circumference changes in preterm infants, but the data were not available. Detailed data on nutrient and fluid supply, which possibly pose positive influences on preterm infants’ weight change, were unavailable. However, most infants (63.6%, 66.3% and 73.7%, p=0.049) started enteral feed in the first 2 days after birth among the no ANS, partial ANS and complete ANS exposure groups, and a small difference in the duration of parenteral nutrition (27±20, 33±17 and 29±17 days, p=0.003) existed among the three groups; moreover, a consistent nutrition protocol was administered among the 32 NICUs. Multiple potential confounders affecting weight changes were also adjusted for. Therefore, it is possibly speculated that infants among three groups had a little difference in nutrients supply.

This study had several advantages. We found that different doses of ANS administration played a significant role in the major morbidities, mortality and early weight changes in EPI or ELBWI, providing scientific evidence from China for neonatologists. Additionally, multicentre and large-sample study results were more widely generalisable. All 32 NICUs applied dexamethasone as ANS, which weakened the unpredictable effects of different types of ANS on the outcomes. Multiple confounders were also adjusted for. Subgroup analysis and heterogeneity analysis were conducted as well. The methods above ensured the credibility of the results.

Conclusions

In our 32-centre cohort study, different doses of ANS (dexamethasone) exposure were protectively associated with BPD, NEC and death, but not EUGR at discharge among EPI or ELBWI. Beneficial dose-dependent associations between ANS (dexamethasone) exposure and BPD existed. ANS (dexamethasone) exposure increased EWLP only in the ≥28 w and <1000 g subgroup. A complete course of ANS is encouraged in clinical performances.

Data availability statement

Data sharing not applicable as the confidentiality agreement of the study existed.

Ethics statements

Patient consent for publication

Ethics approval

This study was approved by the Ethics Committee of Peking University First Hospital (approval number 2019(12)).

Acknowledgments

We thank the following neonatologists for contribution to acquisition of data, execution of the study: Wanxian Zhang (Department of Neonatology, Tianjin Central Hospital of Gynecology Obstetrics); Xiaomei Tong, Yanan Jiang (Department of Pediatrics, Peking University Third Hospital); Jianfang Gai (Neonatal Intensive Care Unit, Children’s Hospital of Shanxi); Xiaoxue Zhang (Department of Neonatology, Handan Central Hospital); Yanguo Zhao (Department of Neonatology, Xingtai People’s Hospital); Ying Zhang (Department of Pediatrics, Tianjin Medical University General Hospital); Haiying He, Haiyan Jiang, Ping Zhang (Department of Pediatrics, Baogang Third Hospital of HongCi Group); Xia Liu, Chunyan Guo (Department of Neonatology, Affiliated Hospital of Chengde Medical University); Junyi Wang, Shasha Fan (Department of Pediatrics, Department of Pediatrics, The First Hospital of Tsinghua University); Chaomei Zeng, Yamei Huangshan (Department of Pediatrics, Peking University People’s Hospital); Yuemei Li, Xiaoli Hao (Department of Neonatology, The Second Hospital of Hebei Medical University); Hua Mei, Qiaoyan Du (Department of Neonatology, The Affiliated Hospital of Inner Mongolia Medical University); Pingping Zhang, Jinyan Zhang (Department of Neonatology, Tianjin First Central Hospital); Danhua Wang, Jinyu Li (Department of Pediatrics, Peking Union Medical College Hospital); Lihua Li, Xiaoxiang Li (Department of Pediatrics, Beijing Luhe Hospital, Capital Medical University); Hongyun Wang, Meiyan Guo (Department of Neonatology, Inner Mongolia Maternal and Child Health Hospital); Hong Cui, Yajing Li (Department of Pediatrics, Beijing Friendship Hospital, Capital Medical University); Shulan Yang, Liang Zhang (Department of Neonatology, Chifeng Municipal Hospital); Mingyan Hei, Lu Chen (Neonatal Center, Beijing Children’s Hospital, Capital Medical University); Lixia Sun, Haixia Cheng (Department of Neonatology, Taiyuan Maternal and Child Health Hospital); Xiaohong Gu, Yali Yang (Department of Neonatology, Zhangjiakou Maternal and Child Health Hospital); Wenli Zhao, Honghong Qin (Department of Gynecology and Pediatrics, PLA Rocket Force Characteristic Medical Center); Xiaoying Wang, Ling Xiao (Department of Neonatology, Children’s Hospital Capital Institute of Pediatrics); Shengshun Que (Department of Neonatology, The Second Hospital of Tianjin Medical University); Yanju Hu, Haiyan Bao (Department of Neonatology, Xing An Meng Hospital of Inner Mongolia); Qiuyan Ma, Xiaoxian Yan (Department of Neonatology, Fenyang Hospital of Shanxi); Haijuan Wang, Mei Sun (Department of Neonatology, Baoding Maternal and Child Health Hospital); Hua Xie, Xiaohong Fu (Department of Neonatology, Affiliated Hospital of Chifeng University); Li Ma, Min Sun (Department of Neonatology, Hebei Children’s Hospital); Jiuye Guo (Department of Pediatrics, Chaoyang Maternal and Child Health Hospital of Beijing); Ming Yang, Jiao Zhang (Neonatal intensive Care Unit, Beijing United Family Hospital).

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • QF and HW contributed equally.

  • Contributors Study design: SL and QF. Acquisition of data: all authors. Integration, analysis and interpretation of data: all authors. Drafting of the manuscript: SL. Important revision of the manuscript: all authors. Statistical analysis: SL, QF, XH and HW. Obtained funding: SL and HW. Study supervision: all authors. Guarantor: QF.

  • Funding The study is funded by the research fund of the Peking University Health Science Center-Weifang Joint Research Center for Maternal and Child Health.

  • Competing interests None.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.