A global analysis of 64 systematic reviews and meta-analyses found that maternal health conditions, low birth weight, and limited antenatal care were among the most consistent factors linked to neonatal death.
Researchers assessed 73 exposure–outcome associations across 54 maternal and perinatal factors. Of these, 24 were supported by probable evidence, 26 by limited-suggestive evidence, and three by limited or non-conclusive evidence.
The strongest risk factors included preterm birth, low birth weight, maternal death, hypertensive and hemorrhagic disorders of pregnancy, and maternal obesity. Infants born preterm or underweight had markedly higher risks of dying within the first 28 days of life. Other probable risk factors included advanced maternal age, maternal anemia, previous stillbirth, prenatal opioid exposure, and arsenic exposure.
Protective associations were also identified. Antenatal care attendance and facility-based delivery were both linked with reduced neonatal mortality. Timely breastfeeding initiation—especially within the first hour rather than delayed beyond 24 hours—was also associated with improved neonatal survival.
Evidence on antenatal care and facility deliveries mainly came from low- and middle-income countries, while research on maternal obesity, smoking, and opioid exposure was largely from high-income settings.
“Priority interventions should focus on preventing preterm birth, promoting timely initiation of breastfeeding, and improving maternal health through the prevention and management of comorbidities, including haemorrhagic disorders, hypertensive disorders of pregnancy, and anaemia,” said Bereket Kefale, MPH, of the Curtin School of Population Health at Curtin University in Perth, Western Australia, and colleagues.
The researchers conducted an umbrella review of systematic reviews and meta-analyses published from 1990 to July 2025. They searched six major databases—Embase, Medline, Global Health, CINAHL, Web of Science, and Scopus—as well as regional and gray literature sources. Eligible reviews included at least two primary studies that evaluated associations between maternal or perinatal factors and all-cause neonatal mortality.
Study quality was assessed using the AMSTAR-2 tool. Associations were categorized as convincing, probable, limited-suggestive, or limited and non-conclusive. The review included 2,742 primary studies representing more than 160 million participants.
Of the included reviews, 8% were rated high quality, 11% moderate, 36% low, and 45% critically low. Substantial heterogeneity was observed across pooled estimates, and publication bias was evaluated in only half of the included reviews.
The researchers noted that most included systematic reviews were of low or critically low methodological quality, which limited the certainty of conclusions. Heterogeneity across studies was considerable and overlap of primary data was frequent. Because the analysis relied on existing publications, causation could not be established. Limited regional representation—especially from low- and middle-income countries—also reduced generalizability.
The review found that while many neonatal deaths remain associated with preventable factors, the evidence base requires substantial strengthening. The researchers called for better-designed studies, standardized outcome definitions, and improved data representation from underreported regions. They emphasized that higher-quality evidence could better guide public health strategies and clinical practice aimed at reducing neonatal deaths globally.
The authors reported no conflicts of interest.
Source: eClinicalMedicine