Probiotics, prebiotics infant formula use in preterm or low birth weight infants: a systematic review

Randomized controlled trials in preterm or low birth weight infants fed formula. Participants included preterm infants (<37 weeks gestation) or low birth weight infants (<2.5 kg at birth), hospitalized in NICUs and receiving formula feeds. Eight trials were included (four probiotic, four prebiotic). Trial designs were randomized; blinding described in several trials, but allocation concealment and other methods were inconsistently reported. Primary outcomes focused on short-term growth; secondary outcomes included NEC, sepsis, and mortality; intervention durations ranged from 14 to 30 days.

Probiotic regimens: Costalos 2003: Saccharomyces boulardii 1 x 10^9 CFU every 12 hours for 30 days; Indrio 2008: Lactobacillus reuteri ATCC 55730 1 x 10^8 CFU daily for 30 days; Reuman 1986: Lactobacillus (unspecified) 9 x 10^6 CFU per mL formula for 30 days; Stratiki 2007: Lactobacillus acidophilus plus Bifidobacterium lactis 2 x 10^7 CFU per g milk powder for 30 days. Prebiotic regimens: Boehm 2002: GOS/FOS blend (GOS 90%, FOS 10%) 1 g/dL for 28 days; Indrio 2009: scGOS/lcFOS in 9:1 ratio at 0.8 g/dL for 15 days; Kapiki 2007: GOS/FOS mixture 0.4 g/100 mL for 14 days; Mihatsch 2006: prebiotic oligosaccharides (unspecified) 1 g/dL for 15 days.

Probiotic or prebiotic supplementation to formula-fed preterm infants did not show consistent improvements in growth or major clinical outcomes. Probiotics did not significantly affect weight gain; stool frequency increased in one trial. Prebiotics did not improve overall weight gain; some subgroups showed changes in linear growth or stool characteristics, and prebiotics increased bifidobacteria counts and stool frequency. Overall, there is not enough evidence to support routine supplementation of preterm formula with probiotics or prebiotics; longer, well-designed RCTs are needed.

Small sample sizes across trials; short intervention durations; heterogeneity in probiotic/prebiotic strains, doses, and outcomes; inconsistent risk-of-bias reporting (randomization, allocation concealment, blinding); limited reporting of clinically important outcomes (NEC, sepsis, mortality) and lack of data specifically for low birth weight infants; insufficient data for robust meta-analytic conclusions.