Antibiotic use predicts microbiota dysbiosis in preterm infants

By Dr. Liji Thomas, MDReviewed by Lauren HardakerJun 11 2025

Researchers reveal how antibiotic treatments in premature newborns can trigger long-lasting changes in the gut microbiome, increasing the risk of dangerous infections and longer hospital stays.

Study: Associations between antibiotic exposure intensity, intestinal microbiome perturbations, and outcomes in premature neonates with bacteremia. Image credit: Iryna Inshyna/Shutterstock.com

Infections in newborns are often accompanied by disturbances of the gut microbiome. A study published in the Journal of Perinatology analyzed the associations of antibiotic exposure with bacteremia in newborns and the role of antibiotic-induced gut dysbiosis.

Introduction

Babies born too early or with too low birth weight are likely to develop infections like necrotizing enterocolitis (NEC), sepsis, and meningitis. As such, many infants receive preventive or therapeutic antibiotics from birth for a few days or even months.

Long treatment periods occur when babies have recurrent infections requiring prolonged or repeated antibiotic treatment during the newborn period. Despite the lifesaving effect of antibiotics in severe bacterial infections, they also severely disrupt the normal development of the gut microbiome.

A study found that when mothers were exposed to antibiotics during childbirth, they were relatively unaffected compared to their infants, suggesting the undeveloped gut microbiome makes infants more vulnerable. 

As such, higher antibiotic exposure is associated with a greater risk of infant gut dysbiosis and an increased prevalence of specific pathogens in the gut. This, in turn, is linked to NEC, spontaneous intestinal perforation (SIP), and late-onset neonatal sepsis. The current study focuses on bacteremia, NEC/SIP, and recurrent infections.

Higher numbers of Enterococcus species in the gut microbiome may lead to inflammation, the effect of which varies with antibiotic choice. Antibiotic exposure intensity scores like the Antibiotic Spectrum Index (ASI) may better assess antibiotic protocols among infants in the neonatal intensive care unit (NICU) than other measures like the number of treatment days.  

The current study aimed to identify a potential cause-effect relationship between antibiotic exposure, gut dysbiosis, and newborn illness. The researchers matched cases and controls by gestational and postnatal age. Whilst this helps to partially account for confounding factors such as feeding protocols and developmental stage, they could not independently control for diet or nutrition due to the frequent changes during illness

About the study

The study included preterm infants weighing 2000 g or less at birth, none older than two months at baseline. The analysis included 398 stool samples from 40 bacteremic newborns and 39 controls.

Stool samples were taken over one year, about five per infant, matching newborns with bacteremia to uninfected controls. Whole metagenome sequencing was performed on stools, and the intensity of antibiotic exposure was compared using the ASI.

Study findings

About 70% of infants whose blood cultures were positive for bacteria were infected. The common pathogens were coagulase-negative Staphylococcus (CONS), Klebsiella, Staphylococcus aureus, Escherichia coli, and Enterococcus species.

Bacteremic babies had longer stays in the NICU than the controls. In six of 15 cases of NEC, the diagnosis occurred within 48 hours before the onset of bacteremia.

Ampicillin and gentamicin were the most used antibiotics to prevent early-onset sepsis (EOS). Vancomycin and cefepime were used almost twice as often among cases as the controls.

Among bacteremic infants, the number of different species (α diversity) in the gut microbiota community was reduced at four weeks of life. This was more pronounced among those with NEC.

This difference persisted as late as 20 weeks after birth in babies with recurrent respiratory and urinary infections. Babies who recovered without subsequent infections developed a microbiome that reverted to healthy diversity.  

The ASI was higher among bacteremic infants throughout the postnatal period. These babies showed a greater number of Enterococcus species (>25% in pooled stool samples) but a lower abundance of anaerobes (<25%) compared to controls at 13-16 weeks. Reduced anaerobe abundance is a marker of poor gut microbiome maturation, as this group includes multiple commensals.

Among bacteremic babies, those with NEC showed higher levels of Enterococcus species by 12 weeks of age compared to those without NEC. Enterococcus levels in bacteremic infants increased and stayed elevated between weeks eight and 16. In uninfected babies, these levels rose only around four weeks. Babies with higher Enterococcus abundance were also more likely to experience repeated infections.

This suggests that bacteremia was associated with delayed gut maturation and predicted higher odds of repeated infections, though causation cannot be confirmed. This is a significant finding that could shape management in this high-risk group.

Conclusions

Premature infants with bacteremia found building a normal gut microbiome challenging despite having a comparable composition between birth and four weeks. This effect was more notable if the baby also had NEC or SIP.

Bacteraemic infants had a lower α diversity in their gut microbiota, which was even more pronounced in those with NEC or SIP. They had more Enterococcus species abundance but fewer anaerobes, signaling dysbiosis.

Such infants were more likely to have repeated infections associated with poor gut microbiome composition. Similar findings in graft-versus-host disease following stem cell transplants for bone marrow disorders led to a shift in antibiotic use patterns.

“As signatures of intestinal microbiome dysbiosis in neonates emerge, a similar proactive approach could be taken in the NICU where prescribers more strongly consider the microbiome impact of different antibiotic choices.”

Using ASI ensured objective quantification of antibiotic exposure and better comparison of microbiome responses between cases and controls. This demonstrated that differences in microbiome diversity remained over time, even after accounting for recent antibiotic use, suggesting that cumulative antibiotic exposure has a lasting impact.

This observational study was from a single center. It could not assign effects to specific antibiotics or rule out the influence of other factors such as diet, inflammation, or clinical instability. As a result, the findings may not be generalizable to all NICU populations.

Future studies should evaluate the potential accumulating and long-term effects of repeated antibiotic exposure, especially with broad-spectrum antibiotics. They should also examine other factors like inflammation, infection, and diet for their impact on infant dysbiosis risk.

Download your PDF copy now!