Evidence increasingly challenges the idea that probiotics work mainly in moderately preterm babies, with multistrain products showing potential to reduce necrotising enterocolitis and possibly sepsis in the most immature infants.
Claims that probiotics work better in moderately preterm babies than in the most immature infants are increasingly hard to sustain, according to a recent review of the evidence on necrotising enterocolitis, or NEC, in extremely preterm babies. The problem, the review notes, is not that the youngest infants have been excluded from research entirely, but that many trials have failed to report results separately for babies by gestational age or birth weight, leaving systematic reviews with only a limited evidence base for the highest-risk group. The Cochrane review published in 2023 found more than 10,000 infants across all included trials, but could analyse only about 1,800 extremely low birthweight or extremely preterm babies, and even that subgroup included a large British study in which Bifidobacterium breve did not show benefit.
That uncertainty has prompted closer attention to the smaller studies that do focus on the most premature infants. An Australian trial of a multistrain probiotic that included Bifidobacterium infantis enrolled about 460 babies born before 28 weeks, and while the subgroup estimate in that age band was imprecise, the broader trial showed a statistically significant reduction in NEC with no clear sign that birth weight altered the treatment effect. A separate randomised trial in extremely low-birthweight infants found that multistrain probiotics changed the gut microbiome in a favourable direction, increasing Bifidobacterium and Lactobacillus, reducing Candida, and helping babies reach full feeds sooner. More recent systematic review evidence has also pointed in the same direction, suggesting possible reductions in NEC and mortality in extremely preterm infants, although the quality of evidence was rated low to very low.
Observational data are filling some of the remaining gaps. A Norwegian population-based study of extremely preterm infants born between 2014 and 2021 followed practice after the country recommended probiotics for this group. During that period, units used different products, including Infloran, Labinic droplets and later Proprems, all of which contained B infantis. The study reported lower mortality and less surgical NEC among babies given probiotics, while late-onset sepsis was similar overall despite a handful of cases of probiotic sepsis. The findings matter because the probiotic group was, on average, smaller and less mature at birth and had worse illness severity scores, meaning the adjusted analysis was essential.
These results fit with a broader pattern seen in meta-analyses of neonatal sepsis. A 2025 systematic review concluded that multistrain probiotic mixtures reduce late-onset sepsis, and a separate review by Nick Embleton, Chris van den Akker and Belal Alsheikh concluded that probiotic supplementation does not increase sepsis overall and is more likely to lower it, even though rare cases of probiotic sepsis do occur. That distinction is important because concerns about contamination or bloodstream infection have long shaped neonatal practice, yet the balance of evidence still suggests that the infection risk from probiotics has not outweighed the benefit in aggregate.
The strongest challenge to routine use has come not from the efficacy data themselves but from regulatory backlash. After US Food and Drug Administration warning actions, one analysis using the Pediatrix data warehouse found that probiotic use fell sharply in neonatal units that had previously used them most often. In those units, both medical and surgical NEC rose after probiotics were stopped, while late-onset sepsis did not materially change. Taken together, the trial and observational data provide little support for drawing a line between moderately preterm and extremely preterm infants when deciding whether probiotics may help prevent NEC. On the contrary, the evidence increasingly points to multistrain preparations, particularly those containing B infantis, as the most plausible option for reducing NEC and perhaps offering a modest reduction in sepsis as well.
Disclaimer: This content is for informational purposes only and is not intended to be a substitute for professional medical judgment, advice, diagnosis, or treatment.
