Researchers analyzing gut viruses and bacteria in 129 premature babies found that distinct viral patterns appear before necrotizing enterocolitis develops, with early-onset disease predicted 75% accurately by phage-bacteria interactions and late-onset disease predicted 83% accurately by antibiotic-resistant bacteria. According to Gram Research analysis, these findings suggest the disease has different microbial causes depending on when it occurs, potentially enabling earlier detection and targeted prevention strategies in the future.
Researchers studied the gut bacteria and viruses of 129 premature babies to understand what causes necrotizing enterocolitis, a dangerous intestinal disease. By analyzing stool samples from 43 babies who developed the disease and 86 who stayed healthy, scientists discovered that different types of viruses and bacteria appear before the illness starts. The study found that early cases of the disease could be predicted by looking at how viruses and bacteria interact, while later cases were better predicted by antibiotic-resistant bacteria. These findings could help doctors identify at-risk babies earlier and develop new treatments targeting the gut microbiome.
Key Statistics
A 2026 study of 129 premature babies across three US hospitals found that early-onset necrotizing enterocolitis could be predicted with 75% accuracy by examining interactions between gut viruses and bacteria before disease symptoms appeared.
According to research published in Gut in 2026, late-onset necrotizing enterocolitis was predicted with 83% accuracy using antibiotic-resistant bacteria profiles, suggesting different microbial mechanisms drive early versus late disease in premature infants.
A 2026 analysis of 1,825 stool samples from premature babies revealed that dormant viruses in the gut carried genes that may help harmful bacteria survive and resist antibiotics, potentially explaining why some babies develop serious intestinal disease.
The Quick Take
- What they studied: Whether patterns of viruses and bacteria in premature babies’ guts can predict which babies will develop necrotizing enterocolitis, a serious intestinal infection.
- Who participated: 129 premature babies from three US hospitals: 43 who developed the disease and 86 healthy babies of similar age and birth weight, tracked over time with stool samples.
- Key finding: Gram Research analysis shows that early-onset disease could be predicted with 75% accuracy by examining how viruses and bacteria interact, while late-onset disease was predicted with 83% accuracy using antibiotic-resistant bacteria patterns.
- What it means for you: If these findings hold up in larger studies, doctors may eventually be able to identify which premature babies are at highest risk for this dangerous disease before symptoms appear, allowing for earlier intervention. However, this research is still in early stages and not yet ready for clinical use.
The Research Details
Researchers collected stool samples from 129 premature babies at three hospitals and analyzed them using advanced genetic sequencing technology. They looked at two main things: the bacteria in the gut and the viruses that infect those bacteria, called bacteriophages or phages. The team collected samples over time from each baby, creating a detailed picture of how the gut microbiome changed day by day. They used machine learning, a type of artificial intelligence, to see if patterns in the viruses and bacteria could predict which babies would later develop necrotizing enterocolitis. The researchers also looked at what medications and foods the babies received, since these factors can change the gut microbiome.
Most previous research focused only on bacteria and ignored viruses, which play an important role in controlling bacterial populations. By including viruses in their analysis, the researchers could see a more complete picture of what happens in the gut before disease develops. The study also recognized that early-onset and late-onset disease might have different causes, so they analyzed these separately. This approach is more realistic because it matches how the disease actually occurs in real babies.
This study has several strengths: it included a large number of samples (1,825 stool samples total), used advanced genetic technology to identify viruses and bacteria accurately, and followed babies over time rather than taking just one snapshot. The researchers also validated their findings using machine learning, which tests whether patterns actually predict disease. However, the study was conducted at only three hospitals, so results may not apply to all premature babies everywhere. The findings need to be confirmed in larger, independent studies before doctors can use them clinically.
What the Results Show
The research revealed that premature babies who later developed necrotizing enterocolitis had different patterns of viruses in their guts compared to healthy babies, even before symptoms appeared. For babies who developed the disease early (within 40 days of birth), the best predictor was the interaction between specific viruses and bacteria, achieving 75% accuracy in identifying which babies would get sick. The study found that these viruses were present but not actively reproducing, suggesting they were dormant or integrated into bacterial cells. The viruses appeared to carry genes that helped harmful bacteria survive and resist antibiotics. For babies who developed the disease later (after 40 days), antibiotic-resistant bacteria were the strongest predictor, with 83% accuracy. This suggests that different mechanisms drive early-onset and late-onset disease.
The researchers discovered that antibiotic use, feeding methods, and other medications significantly shaped which viruses and bacteria colonized the babies’ guts. Babies who received antibiotics showed different patterns of viral and bacterial diversity compared to those who did not. The study also found that viruses in early-onset cases had low gene expression, meaning they were not actively making proteins, which is consistent with a dormant state. These dormant viruses may have been protecting harmful bacteria from the immune system or helping them survive in the gut environment.
Previous studies of necrotizing enterocolitis focused almost exclusively on bacteria, missing an important part of the gut ecosystem. This research is among the first to systematically examine viruses in premature babies with this disease. Earlier work suggested that bacterial imbalance causes the disease, but this study shows that the relationship between viruses and bacteria is equally important. The finding that early and late disease have different causes challenges the idea that necrotizing enterocolitis is a single disease with one underlying mechanism.
The study was conducted at only three hospitals in the United States, so results may not apply to premature babies in other countries or healthcare settings. The researchers identified patterns that predict disease but did not prove that these viruses and bacteria actually cause the disease. The study is also relatively small for machine learning research, so the accuracy rates may be higher than what would occur in a larger, independent group of babies. Finally, the research is very new and has not yet been tested in clinical practice, so doctors cannot use these findings to make treatment decisions yet.
The Bottom Line
Based on this research, there are no immediate changes parents or doctors should make. The findings are promising but still experimental. Future research should test whether identifying at-risk babies using these viral and bacterial patterns actually helps prevent the disease through early intervention. Doctors should continue following current guidelines for preventing necrotizing enterocolitis, including careful feeding practices and judicious use of antibiotics in premature babies.
This research is most relevant to neonatologists (doctors who care for premature babies), researchers studying the microbiome, and parents of extremely premature infants. Parents of premature babies should be aware that this research may eventually lead to better ways to predict and prevent this serious disease. Healthcare systems and hospitals may want to follow this research as it develops, since it could eventually inform new screening and prevention strategies.
This research is in early stages. It typically takes 5 to 10 years for laboratory discoveries to become clinical tools that doctors can use in hospitals. Before these findings can be applied in practice, larger studies must confirm the results, and researchers must develop a practical test that hospitals can perform. Parents should not expect changes in premature baby care based on this research for several years.
Frequently Asked Questions
Can doctors predict which premature babies will get necrotizing enterocolitis?
New research suggests yes, but not yet in clinical practice. Scientists can now identify viral and bacterial patterns that predict the disease with 75 to 83% accuracy, but these tests are still experimental and require further validation before hospitals can use them routinely.
What role do viruses play in premature baby gut infections?
Viruses that infect bacteria appear to protect harmful microbes and help them resist antibiotics. A 2026 study found these dormant viruses carry genes that enhance pathogenic bacteria survival, making them important players in disease development alongside bacteria.
Is necrotizing enterocolitis caused by one type of bacteria or many?
Research shows multiple bacteria and viruses work together to cause the disease, and the specific combination differs depending on when the disease develops. Early cases involve different microbial interactions than late cases, suggesting distinct disease mechanisms.
How soon will these findings help treat premature babies?
These discoveries are still in research stages. Larger studies must confirm results and develop practical clinical tests before doctors can use this information. This typically takes 5 to 10 years from laboratory discovery to hospital implementation.
Should parents of premature babies change anything based on this research?
No immediate changes are needed. Parents should continue following their neonatologist’s guidance on feeding and medications. Future research may lead to new prevention strategies, but current medical care remains the standard until these findings are validated clinically.
Want to Apply This Research?
- For parents of premature babies, track daily stool characteristics (color, consistency, frequency) and any feeding changes, as these relate to gut microbiome changes. Note antibiotic use, formula type, and any signs of feeding intolerance.
- If this research eventually leads to clinical testing, parents could work with their neonatal team to understand their baby’s microbiome risk profile and make informed decisions about feeding methods and antibiotic use in consultation with doctors.
- Long-term tracking would involve monitoring the baby’s digestive health during the first 3 to 6 months of life, the critical window when necrotizing enterocolitis typically develops. Parents should maintain detailed records of feeding, medications, and any gastrointestinal symptoms to share with their medical team.
This research is preliminary and not yet ready for clinical application. The findings describe patterns associated with necrotizing enterocolitis but do not establish definitive cause-and-effect relationships. Parents of premature babies should not make medical decisions based on this research alone. All questions about premature infant care, feeding, and disease prevention should be directed to the baby’s neonatologist or pediatrician. This article is for educational purposes and should not replace professional medical advice.
This research translation is published by Gram Research, the science division of Gram, an AI-powered nutrition tracking app.