In the hushed, warm environment of a neonatal intensive care unit, a fundamental question divides clinicians: when should a preterm infant receive its first feeding? The answer varies dramatically from one hospital to the next, and even among attending neonatologists within the same unit. For extremely low birth weight babies—those weighing less than 1,000 grams—the decision can mean the difference between rapid growth and a devastating intestinal catastrophe. Yet the evidence underpinning these choices remains surprisingly thin.

The Feeding Debate That Divides Neonatologists

The core tension is between providing early nutrition to support brain development and avoiding the rare but lethal complication known as necrotizing enterocolitis (NEC). NEC is an inflammatory necrosis of the bowel that primarily strikes preterm infants in the first weeks of life, carrying a mortality rate of 20 to 30 percent. For decades, clinicians feared that early feeding might trigger NEC by overwhelming an immature gut. Others argued that delaying feeds starves the gut mucosa, increasing vulnerability to infection and inflammation.

Observational studies have fueled both sides. Some large cohort analyses suggest that early trophic feeds—small volumes of milk, often as little as 10–20 mL per kilogram per day—reduce the risk of NEC compared to prolonged bowel rest. Other analyses find no association or even a slight increase in NEC with earlier initiation. The inconsistency reflects confounding by illness severity: sicker infants tend to be fed later, so any protective effect of early feeding may be masked.

Randomized trials have been smaller and underpowered. A 2021 Cochrane review, led by Dr. William McGuire at the University of York, pooled data from 15 trials involving more than 3,000 infants. The review found no statistically significant difference in rates of NEC or all-cause mortality between early and delayed feeding groups. The only clear benefit was a modest reduction in the time to reach full enteral feeds—roughly one to two days earlier. This modest gain has not been enough to shift practice uniformly.

The debate persists because the stakes are high and the signal from the data is weak. As one neonatologist told me, “We are making decisions based on belief systems, not evidence.” The result is a patchwork of protocols that can confuse parents and frustrate nurses who must implement them.

Why Early Feeding Sounded Like a Good Idea

The biological rationale for early feeding rests on the concept of gut mucosal protection. In utero, the fetus swallows amniotic fluid, which bathes the intestinal lining with growth factors and immunoglobulins. After birth, enteral feeds provide similar trophic stimuli. Without them, the gut epithelium can atrophy, tight junctions loosen, and bacterial translocation increase—a setup for NEC.

Animal models reinforced this hypothesis. In piglet studies, those fed colostrum or milk formulas within hours of birth had healthier intestinal villi and lower inflammatory markers compared to those starved for 24 to 48 hours. The data were compelling enough that by the early 2000s, many NICUs shifted from a default waiting period of 48 to 72 hours to initiating trophic feeds within 6 to 24 hours of birth.

Yet the translation from bench to bedside has been rocky. Preterm infants are not piglets. Their gut barrier function, motility, and immune responses are uniquely immature. What protects a full-term piglet may overwhelm a 26-week human infant whose intestines lack coordinated peristalsis and whose microbiome is being shaped by the NICU environment. The timing of NEC’s peak—usually around 10 to 14 days of life—suggests that the risk may be more related to the cumulative load of milk exposure than to the first feed itself.

The historical practice of delayed feeding was not arbitrary; it grew out of clinical observation in the 1960s and 1970s, when early feeds were associated with high rates of NEC. But those early feeds were often large volumes of formula, not the tiny trophic amounts used today. The field has been trying to find a middle path ever since.

The Cochrane Review That Shook the Field

Dr. William McGuire’s 2021 Cochrane review, updated from a 2017 version, remains the most comprehensive synthesis of the evidence. It included 15 randomized controlled trials enrolling infants born at less than 37 weeks or weighing less than 2,500 grams. The intervention was early initiation of enteral feeds (typically within 24 to 48 hours) versus delayed initiation (after 72 hours or more).

The primary outcome—incidence of NEC Bell stage II or higher—showed no statistically significant difference: risk ratio 0.93, with a 95% confidence interval spanning from 0.72 to 1.20. Mortality before hospital discharge was similarly non-significant. The only outcome that reached statistical significance was the time to full enteral feeds, which was reduced by a mean of 1.2 days in the early feeding group.

The review’s authors were cautious. They noted that most trials were small, with wide confidence intervals, and that the definition of “early” varied from 6 to 48 hours. Subgroup analyses by gestational age and birth weight were underpowered. “The available evidence does not provide a clear answer for clinicians,” McGuire wrote in a commentary, “but it does suggest that the risk of harm from early feeding is not large.”

For some neonatologists, that was enough to continue early feeding. For others, the lack of benefit argued for a more conservative approach. The review did not settle the debate; it exposed the uncertainty.

How NICUs Actually Decide: A Survey Snapshot

A survey published in Pediatrics in 2020 captured the breadth of variation across US neonatal units. Researchers contacted 248 level III and IV NICUs and asked about their feeding protocols for extremely low birth weight infants. The reported timing for first enteral feed ranged from 6 hours to 72 hours after birth. About 40 percent of units started within 24 hours, another 40 percent between 24 and 48 hours, and the remaining 20 percent waited beyond 48 hours.

The survey also found that unit culture—the collective experience and preferences of senior physicians—was a stronger predictor of practice than published evidence. Units with a history of conservative feeding were unlikely to change even after the Cochrane review. Conversely, units that had adopted early feeding tended to cite observational data and a single influential trial from 2008 that showed faster growth without an increase in NEC.

The variation was most pronounced for infants born at 23 to 25 weeks’ gestation, where the risk of NEC is highest and the benefit of nutrition most critical. Some units started trophic feeds at 12 hours for all stable infants; others waited until 72 hours and required a Doppler ultrasound showing adequate mesenteric blood flow. “We have no standardized approach,” one survey respondent wrote. “Each attending does what they think is best.”

This variability is not unique to the United States. A 2022 survey of European NICUs found similar disparities, with French units tending toward earlier feeds and German units more likely to delay. The lack of consensus has real consequences for research, as multi-center trials struggle to recruit sites with compatible protocols.

Another dimension of variation concerns the type of milk used for the first feed. Some units exclusively use expressed breast milk, others use donor milk, and a minority still use preterm formula. Each choice carries its own risk-benefit profile. Breast milk contains immunoglobulins and prebiotics that may protect against NEC, but its availability varies. Donor milk is pasteurized, which reduces some protective factors. Formula is nutritionally consistent but may increase the risk of NEC. The interplay between feeding timing and milk type is poorly studied, adding another layer of uncertainty.

For example, a 2018 study from the University of Iowa found that among infants fed exclusively with breast milk, early initiation was associated with lower NEC rates, but among those fed formula, early initiation showed a trend toward higher NEC. This suggests that the optimal timing may depend on what is being fed—a nuance that most protocols ignore.

Parental Anxiety Meets Clinical Uncertainty

For parents of preterm infants, the feeding question adds another layer of stress to an already overwhelming experience. Many expect that their baby will be fed immediately, as a newborn would be in a healthy term birth. When they see a tiny infant with a nasogastric tube but no milk flowing, they often interpret it as neglect or incompetence.

“I kept asking the nurse why they weren’t feeding him,” one mother told me. “She said they were waiting for his gut to ‘wake up.’ I had no idea what that meant.” Parents are rarely included in the decision-making process; the medical team decides based on protocol and clinical judgment. When a unit’s policy changes without explanation—or differs from what a parent read online—trust can erode quickly.

Shared decision-making is uncommon in this context, partly because the evidence is too ambiguous to present a clear risk-benefit tradeoff. How do you explain equipoise to a sleep-deprived parent? Some units have developed written materials that acknowledge the uncertainty, but these are the exception. The emotional toll is compounded when a baby develops feeding intolerance or NEC, and parents wonder whether a different timing might have prevented it.

The guilt is not theirs to carry, but they carry it anyway. As one NICU social worker observed, “Feeding is the most primal thing parents do. When they can’t do it, and when the team seems unsure, it shakes their confidence in the whole system.”

To address this, a handful of NICUs have started including parents in feeding rounds, explaining the rationale for timing decisions in plain language. At the University of California, San Francisco, a pilot program provides each family with a written feeding plan that includes a range of possible timelines and the reasons for any delays. Early feedback suggests that parents feel more informed and less anxious, though the approach has not been formally evaluated.

Interestingly, parental attitudes can also influence clinician behavior. In a 2021 qualitative study, neonatologists reported feeling pressure from parents to start feeds earlier, especially when parents had read about the benefits of early feeding online. Some clinicians admitted to deviating from their preferred protocol to accommodate parental requests, even when they believed it was not in the infant’s best interest. This dynamic underscores the need for better communication tools that convey uncertainty without undermining trust.

What the Ongoing Trials Might Resolve

Two large randomized trials are currently underway, each hoping to provide definitive answers. The NEO-EXPRESS trial, led by researchers at the University of Oxford and funded by the UK National Institute for Health and Care Research, aims to enroll 2,400 infants born before 30 weeks’ gestation. Participants are randomly assigned to immediate trophic feeds (within 6 hours of birth) or delayed trophic feeds (after 72 hours). The primary outcome is survival without NEC at 36 weeks’ postmenstrual age.

Secondary outcomes include growth velocity, time to full feeds, duration of parenteral nutrition, and neurodevelopmental assessment at 2 years corrected age. The trial is expected to report results in 2025. Its size and rigorous design should provide enough statistical power to detect a moderate effect on NEC, if one exists.

A second trial, the Early vs. Delayed Enteral Feeding (EDEN) trial in Australia and New Zealand, has a similar design but focuses on infants weighing less than 1,500 grams. Both trials have faced recruitment challenges, as some clinicians are reluctant to randomize when they have strong personal preferences. “We are asking people to accept uncertainty,” the NEO-EXPRESS principal investigator told me. “That is hard in a field where every decision feels life-or-death.”

Even if the trials show a clear benefit for one strategy, implementation will take years. Changing unit culture is slow, and the results may not generalize to settings with different feeding practices or baseline NEC rates. But at minimum, they should reduce the range of acceptable timing, narrowing the window from 6 to 72 hours down to something more evidence-based.

There is also the question of whether the trials will address the interaction with milk type. Neither NEO-EXPRESS nor EDEN stratifies randomization by breast milk versus formula, so any subgroup findings will be exploratory. If the results are equivocal, the debate may shift to whether the timing question is even answerable without considering what is being fed.

Some researchers argue that the field should move beyond timing alone and focus on optimizing the composition and advancement rate of feeds. For instance, a 2022 meta-analysis suggested that slow advancement of feed volumes (15–20 mL/kg/day) reduces NEC compared to faster advancement, regardless of when feeds are started. This raises the possibility that the timing debate may be less important than how feeds are escalated. Future trials may need to test combinations of timing and advancement strategy to provide actionable guidance.

A Practical Middle Ground for Clinicians

Until the trials report, most experts recommend an individualized approach that balances nutritional needs with risk factors. Infants with signs of hemodynamic instability, such as low blood pressure or high lactate levels, are generally kept nil per os until they stabilize. Those who are stable and have normal abdominal exams can start trophic feeds within 24 to 48 hours, with gradual advancement.

Emerging technologies may help refine this decision. Doppler ultrasound of the superior mesenteric artery can assess blood flow before feeds; low flow has been associated with feeding intolerance and NEC. Biomarkers like intestinal fatty acid binding protein (I-FABP), measured in urine or stool, can detect early intestinal injury. Some units are incorporating these into feeding algorithms, though they are not yet standard.

Standardized feeding protocols, when implemented consistently, have been shown to reduce variation and improve outcomes. A 2019 study found that units using a written protocol had lower rates of NEC and faster time to full feeds, regardless of the specific timing chosen. The key is consistency: nurses and doctors follow the same rules, reducing individual bias and improving adherence.

The middle ground is not a compromise; it is a recognition that the evidence does not support a single right answer. As one neonatologist put it, “We need to be humble. We don’t know yet. But we can be systematic in how we manage that uncertainty.” That humility, combined with ongoing research and a willingness to adapt, may ultimately serve preterm infants better than any rigid protocol.

One approach that has gained traction is the use of “feeding bundles”—checklists that include timing, volume advancement, and monitoring parameters. A 2020 quality improvement project at a large US NICU implemented a bundle that standardized initiation at 24 hours for stable infants, with daily assessments for advancement. Over 18 months, the unit saw a 30% reduction in feeding intolerance and a non-significant trend toward lower NEC. While not a randomized trial, such initiatives demonstrate that even imperfect evidence can be operationalized to reduce harm.

Another area of active investigation is the role of probiotics in modulating gut health during the transition to enteral feeds. Several meta-analyses suggest that probiotics reduce the risk of NEC in preterm infants, but the optimal strain, dose, and timing remain unclear. Some units now combine early trophic feeds with prophylactic probiotics, though this practice is not yet universal. The interaction between feeding timing and probiotic efficacy is unknown, but it represents a promising avenue for future research.

Ultimately, the debate over when to start feedings reflects a deeper tension in neonatology: the desire for evidence-based practice collides with the reality that evidence is often insufficient for the complex, fragile patients in our care. Until large trials provide clearer answers, clinicians must navigate uncertainty with humility, transparency, and a commitment to learning from every infant they treat. For parents, that uncertainty is a source of anxiety; for clinicians, it is a call to do better.

This article is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare provider for decisions regarding infant feeding and medical care.