Systematic review and meta-analysis of neonatal outcomes of COVID-19 vaccination in pregnancy

The study addresses whether maternal COVID-19 vaccination during pregnancy affects neonatal outcomes. Given the global burden of COVID-19 and increased risks observed in pregnant individuals and their neonates, vaccination has been recommended by several health authorities despite initial exclusion of pregnant women from trials. While antibodies are known to cross the placenta, comprehensive evidence on neonatal outcomes following maternal vaccination is limited. The purpose is to systematically synthesize observational evidence comparing neonatal outcomes among vaccinated versus unvaccinated pregnancies to inform safety and potential benefits for newborns.

Early vaccine trials excluded pregnant women, leading to reliance on observational data. Multiple studies have shown COVID-19 vaccination reduces severe disease in pregnancy without increasing adverse perinatal outcomes. Antibody transfer across the placenta and in breastmilk has been demonstrated. Prior systematic reviews have focused largely on maternal outcomes or perinatal outcomes broadly, with fewer analyses specifically targeting neonatal endpoints. Vaccine recommendations for pregnancy vary internationally, and data from randomized controlled trials are scarce due to ethical constraints, highlighting the need for a focused synthesis on neonatal outcomes.

Design: Systematic review and meta-analysis registered in PROSPERO (CRD42022343713) and conducted per PRISMA guidelines. Data sources: PubMed, EMBASE, and WHO COVID-19 Database from inception to July 3, 2022; no language restrictions. Search terms combined MeSH and free-text for pregnancy, newborns, COVID-19, and vaccines. Eligibility: Observational studies reporting neonatal outcomes comparing pregnant individuals vaccinated against COVID-19 versus unvaccinated controls. Main outcomes: preterm birth, small for gestational age (SGA), NICU admission, and 5-minute Apgar score <7. Additional outcomes (when available): preterm <34 weeks, low birth weight, very low birth weight, congenital anomalies, jaundice, 1-minute Apgar <7, meconium aspiration syndrome, mechanical ventilation, seizures, hypoglycemia, sepsis, encephalopathy, intracranial hemorrhage, transient tachypnea of the newborn, respiratory distress syndrome, umbilical cord blood pH <7.1. Exclusions: irrelevant topic; no main outcomes; reviews, editorials, conference papers, case reports, animal studies; studies lacking an unvaccinated control group. Screening and data extraction were performed independently by two reviewers with arbitration by a third as needed. Extracted data included study setting, design, vaccine platform, doses, timing (trimester), counts of vaccinated/unvaccinated pregnancies, COVID-19 infection status, singleton/multiple gestations, and neonatal outcomes. Risk of bias was assessed using ROBINS-I. Statistical analysis: Dichotomous outcomes pooled as odds ratios (ORs) with 95% CIs using fixed-effects models when I2 ≤50% and random-effects when I2 ≥50%. Heterogeneity interpreted as I2 0–25% insignificant, 26–50% low, 51–75% moderate, >75% high. Analyses were performed in RevMan 5.4. Study flow: 399 records identified; 151 duplicates removed; 248 screened; 194 excluded at title/abstract; 54 full texts reviewed; 39 excluded (9 design, 11 outcomes, 19 intervention); 15 studies included (2 case-control, 13 cohort) from Israel, UK, Romania, Canada, US, Sweden, and Norway, totaling 90,443 vaccinated and 265,063 unvaccinated pregnancies. Most vaccinations were mRNA (BNT162b2, mRNA-1273); few viral vector vaccines. Seven studies reported dose details and eight reported trimester of vaccination; most vaccinations occurred in the second/third trimester.

Main outcomes: Preterm birth (11 studies; 38,181 vaccinated vs 56,246 unvaccinated): pooled OR 0.83 (95% CI 0.74–0.95), P=0.004, I2=55%, indicating lower odds with vaccination. Small for gestational age (10 studies; 64,787 vs 180,784): OR 0.93 (95% CI 0.90–0.96), P<0.0001, I2=46%, suggesting reduced odds with vaccination. NICU admission (7 studies; 72,167 vs 213,581): OR 0.93 (95% CI 0.82–1.05), P=0.24, I2=83%, no significant difference. Low 5-minute Apgar <7 (8 studies; 61,322 vs 215,166): OR 0.92 (95% CI 0.86–0.99), P=0.03, I2=0%, indicating reduced odds. Additional outcomes: Generally no significant differences between groups; examples include preterm <34 weeks (3 studies; 4,085 vs 1,922): OR 0.59 (95% CI 0.33–1.05), P=0.07, I2=10%; low birth weight <2500 g (2 studies; 16,878 vs 9,314): OR 0.90 (95% CI 0.79–1.02), P=0.10, I2=0%; meconium aspiration syndrome (3 studies; 4,125 vs 2,052): OR 0.73 (95% CI 0.34–1.56), P=0.40, I2=0%; mechanical ventilation (2 studies; 3,952 vs 1,523): OR 0.51 (95% CI 0.26–1.00), P=0.05, I2=0%. Subgroup analyses by timing: Studies in which >80% of vaccinations occurred in the third trimester showed no effect on preterm birth (OR 0.87; 95% CI 0.55–1.37; I2=0%) or low 5-min Apgar (OR 1.17; 95% CI 0.68–2.02; I2=0%). When <80% occurred in the third trimester (i.e., more earlier gestation vaccination), protective associations were observed for preterm birth (OR 0.80; 95% CI 0.66–0.96; I2=80%) and low 5-min Apgar (OR 0.92; 95% CI 0.86–0.99; I2=0%). Heterogeneity: Preterm and NICU outcomes showed moderate (I2=55%) and high (I2=83%) heterogeneity, respectively. Removing Lipkind et al. reduced preterm I2 to 0% (OR 0.89; 95% CI 0.81–0.96; P=0.005) without changing significance. Removing Fell et al. reduced NICU I2 to 0% (OR 0.99; 95% CI 0.95–1.03; P=0.68) with unchanged non-significance.

Across 15 observational studies encompassing over 355,000 pregnancies, maternal COVID-19 vaccination was not associated with increased adverse neonatal outcomes and was linked to lower odds of preterm birth, SGA, and low 5-minute Apgar scores. No significant differences were observed for NICU admission or a range of additional neonatal morbidities. Limited subgroup evidence suggests that vaccination earlier than the third trimester may confer greater protective effects on preterm birth and Apgar outcomes, though selection and timing biases cannot be excluded. These findings align with real-world data supporting vaccine safety and effectiveness in pregnancy and reinforce professional recommendations encouraging vaccination. Biological plausibility is supported by efficient transplacental antibody transfer and presence of antibodies in breast milk, potentially offering neonatal protection. Nevertheless, gaps remain regarding dose-response, vaccine platforms beyond mRNA, trimester-specific effects, and variant-specific impacts.

Maternal COVID-19 vaccination during pregnancy is not associated with increased adverse neonatal outcomes and is linked to modest reductions in preterm birth, SGA, and low 5-minute Apgar scores. The evidence supports vaccination during pregnancy and may help alleviate hesitancy. Future research should include trimester-specific analyses, evaluation by vaccine type and dose number, inclusion of diverse global settings and vaccine platforms, randomized or high-quality quasi-experimental designs where feasible, and assessment of variant-specific effects on neonatal outcomes.

Primary limitations include the absence of randomized controlled trials; limited reporting by vaccine dose, platform, and precise trimester of exposure; few studies reporting many additional neonatal outcomes; data predominantly from developed countries with mostly mRNA vaccines, limiting global generalizability; inability to assess variant-specific effects; and potential bias in cohorts largely vaccinated in the third trimester.