Rethinking Early-Life Exposures: A Q&A with Sagori Mukhopadhyay, MD, MMSc

From the moment of birth — particularly for preterm infants — clinical decisions can shape a lifetime of health. Sagori Mukhopadhyay, MD, MMSc, a Children’s Hospital of Philadelphia (CHOP) neonatologist who practices at Pennsylvania Hospital, is at the forefront of research examining how early-life exposures, including infection and antibiotic use, influence long-term outcomes. 

An associate professor of Pediatrics, Dr. Mukhopadhyay leads research at the intersection of neonatal infectious diseases, microbiome science and clinical epidemiology. Her work spans longitudinal birth cohorts, translational science and pragmatic clinical interventions, with the goal of informing more precise and effective care in the NICU. 

In the Q&A below, Dr. Mukhopadhyay discusses how early-life factors shape lifelong health; the evolving science of the neonatal microbiome; and how risk-based approaches to antibiotic stewardship are reshaping care for the most vulnerable patients. 

What first drew you to study early-life exposures and long-term outcomes in neonatology?

In most areas of medicine, the goal is to restore patients to their previous state of health. In neonatology, that baseline doesn’t exist. We are seeing patients at the very beginning of life — particularly preterm infants, where there is no clear definition of “normal.” 

That shifts the way you think about clinical care. Many of the decisions we make early on can’t be fully understood without looking at downstream outcomes. Over time, it became increasingly clear that even routine interventions could have far-reaching effects, and that those effects needed to be studied systematically.  

How has our understanding of early antibiotic exposure and the developing microbiome evolved?

We now have strong epidemiologic data linking early antibiotic exposure to a wide range of later health outcomes, including immune disorders, obesity and metabolic disease. At the same time, mechanistic studies — largely in model systems — have demonstrated that these effects can be mediated through alterations in the microbiome. 

However, there remains a critical gap. While we have human data showing associations and experimental data explaining mechanisms, connecting the two directly in human populations is more challenging. The microbiome is not a single organism — it is a complex ecosystem. Antibiotics don’t simply eliminate bacteria; they reshape that ecosystem in ways that are often difficult to predict.  

Your recent work on delivery-based risk stratification for early-onset sepsis has reduced empiric antibiotic use in very low birth weight infants. What is the key clinical takeaway?

One of the most important insights is that there are subsets of very preterm infants in whom early-onset sepsis is extremely unlikely based on delivery characteristics alone. By identifying those infants, we can safely withhold empiric antibiotics rather than treating broadly. 

This represents a shift in mindset. Historically, the approach has been to treat first and reassess later. What this work shows is that, with a mechanism-based strategy, we can identify infants who are unlikely to benefit from antibiotics in the first place — even among very high-risk populations.  

What practical strategies can NICUs use to improve antibiotic stewardship while maintaining safety?

The most effective strategies are often local. Before implementing changes, it’s essential to understand where and why antibiotics are being used within a given unit. 

Even among centers with highly developed stewardship programs, there remain areas for improvement. For example, moderate preterm infants — who have a relatively low risk of infection — often receive antibiotics in the absence of tailored risk-based strategies. Identifying these gaps through data tracking is a critical first step toward more targeted stewardship.  

Why is simplicity so important when implementing stewardship approaches in the NICU?

Clinical decision-making in the NICU is time-sensitive and often reflexive. Tools that rely on readily available data — such as routine obstetric information — can be integrated seamlessly into existing workflows. 

Requiring new diagnostics or complex tools can slow adoption. In contrast, strategies that are simple, reproducible and embedded in current practice environments are far more likely to be implemented consistently and safely.  

How do you translate large-scale cohort data into changes at the bedside?

Many research questions start with observations in clinical care — patterns that raise concern or suggest opportunities for improvement. These observations can then be tested in larger datasets, including multi-center cohorts, to determine whether they represent broader trends. 

Importantly, because these questions originate at the bedside, there is a clear pathway back to clinical implementation. The goal is not only to generate knowledge, but to directly inform and improve practice.  

What opportunities do longitudinal birth cohorts and biospecimen repositories create?

These resources allow for prospective, targeted data collection that can be leveraged as new technologies emerge — particularly in areas like microbiome sequencing. 

They also ensure that advanced analytical tools are applied to the populations where they are most needed, including preterm infants. Without these cohorts, many clinically relevant questions in neonatology would remain unaddressed.  

How do you see microbiome science influencing neonatal care in the next decade?

Microbiome-directed therapies have tremendous potential, but they require more precise application. Historically, interventions such as probiotics have been applied broadly, without accounting for differences in baseline microbiome composition. 

As technologies improve, there is an opportunity to move toward more targeted approaches — phenotyping infants based on their microbiome and tailoring interventions accordingly. This could enable more precise strategies to support healthy colonization and reduce downstream risk. 

What role does collaboration play in advancing perinatal infectious disease research?

By definition, perinatal infectious diseases span multiple disciplines, including neonatology, obstetrics, infectious disease, microbiology and data science. 

Addressing these questions requires integrated expertise. Multidisciplinary collaboration is not simply advantageous — it is essential to understanding complex disease processes and translating discoveries into improved outcomes.  

What are the most important unanswered questions in your field?

A central challenge is understanding how individual infants respond to antibiotics at the level of the microbiome. While we know antibiotics alter microbial communities, we cannot yet predict those effects reliably for a given patient. 

Future work will focus on identifying phenotypes of infants and microbiome patterns that can guide more personalized therapies — both to treat infection and to mitigate unintended consequences. 

What emerging trends should clinicians be watching?

Advances in molecular diagnostics are rapidly transforming infectious disease management. New technologies are enabling faster, more precise identification of pathogens — sometimes independent of traditional culture methods. 

These tools have the potential to support more targeted antibiotic use, but their successful implementation will depend on integration into clinical workflows and provider familiarity with their interpretation.  

What motivates your work as a neonatologist and physician-scientist?

The resilience of newborns is extraordinary. Even after severe illness, many infants recover and go on to thrive. 

That potential — and the opportunity to shape outcomes at the very beginning of life — is deeply motivating. Antibiotics are powerful, lifesaving tools, but the goal is to use them as precisely as possible, minimizing harm while maximizing benefit for each child.  

Looking ahead, Dr. Mukhopadhyay’s work will continue to advance a more nuanced understanding of neonatal infectious disease — one that integrates epidemiology, microbiome science and clinical care. By refining how and when interventions are used, her research aims to improve not only immediate outcomes in the NICU, but long-term health trajectories for children.