Probiotics and Risk of Death, Necrotizing Enterocolitis, and Sepsis in Very Preterm Infants

Abstract

Importance Probiotics are reported to benefit preterm infants, but population-based clinical evidence is limited. In 2020, the Swedish Neonatal Society issued a national recommendation to supplement very preterm infants (28 weeks 0 days’ to 31 weeks 6 days’ gestation) with probiotics. Objective To investigate whether probiotic supplementation was associated with risk of death, necrotizing enterocolitis (NEC), and/or culture-proven sepsis in infants born very preterm. Design, Setting, and Participants This population-based cohort study examined data from the Swedish Neonatal Quality Register from January 1, 2017, through December 31, 2024. Live-born, very preterm infants, excluding those who died within the first 3 days of life or due to congenital or chromosomal abnormalities, were included. Exposure Daily probiotic supplementation of 1 billion colony-forming units of freeze-dried Bifidobacterium infantis , Bifidobacterium lactis , and Streptococcus thermophilus initiated following birth and discontinued at postmenstrual week 34. Main Outcomes and Measures Composites of death and/or NEC (Bell stage 2-3), death and/or surgical NEC (Bell stage 3), and death and/or culture-proven sepsis were assessed. Modified Poisson regression, propensity score–matched, and inverse probability of treatment weighting analyses were used for replication analyses. A multivariate bayesian modeling analysis accounting for the association between NEC and mortality was performed to estimate the association between probiotic exposure and mortality. Results Among 4695 infants (mean [SD] gestational age, 30.2 [1.1] weeks; 2550 male [54.3%]), probiotic exposure increased from 0% before the national recommendation in 2020 to 16.9% in 2020 and a mean (SD) of 64.8% (4.7%) in 2021 to 2024. Probiotic supplementation vs no supplementation was associated with lower risk of death and/or NEC (rate, 0.8% vs 3.4%; adjusted relative risk [ARR], 0.28 [95% CI, 0.14-0.56]). When assessed separately, risks of death (rate, 0.3% vs 1.4%; ARR, 0.16 [95% CI, 0.06-0.45]) and NEC (rate, 0.6% vs 2.5%; ARR, 0.25 [95% CI, 0.10-0.61]) were both lower in the probiotic group. Multivariate bayesian modeling suggested that a substantial proportion of the observed mortality reduction was mediated via reduced NEC incidence (42.3% [credible interval, 16.1%-100%]). Probiotic supplementation was not associated with death and/or culture-proven sepsis (rate, 3.4% vs 4.4%; ARR, 0.86 [95% CI, 0.58-1.26]). Conclusions and Relevance This cohort study of very preterm infants found that implementation of a multistrain probiotic was associated with a significant reduction in death and/or NEC, which may help inform shared decision-making between clinicians and families. These results should not be generalized to extremely preterm infants (<28 weeks’ gestation).

The authors' abstract, as published at the source. JAMA Network Open, 2026 · DOI ↗