Premature infants exposed to maternal diabetes during pregnancy may face a higher risk of developing severe retinopathy of prematurity (ROP), according to a retrospective cohort study published in Ophthalmology Science.
Researchers found that the association held even after accounting for gestational age, birthweight and other major clinical factors, pointing to prenatal metabolic exposure as a potential contributor to disease progression.
Let’s start with the basics: Why might maternal diabetes affect ROP risk?
ROP develops because retinal blood vessels are not fully formed at birth in premature infants. As those vessels continue to grow outside the womb, the process can become disorganized, leading to abnormal vessel formation and, in more severe cases, retinal detachment.
The connection to maternal diabetes comes from the fact that ROP shares key features with diabetic retinopathy, particularly the role of ischemia-driven abnormal vessel growth.
Because glucose and other circulating factors can cross the placenta, it suggests that exposure to maternal hyperglycemia may influence how the retinal vasculature develops before birth, potentially setting the stage for more severe disease after delivery.
Why has this question remained unclear?
Previous studies looking at maternal diabetes and ROP have not agreed with each other. Some found a link, while others did not.
A key limitation is that many grouped all types of diabetes together, rather than distinguishing between gestational, type 1 (T1D), and type 2 diabetes (T2D).
That matters because the timing and duration of exposure are different. T1D and T2D are present throughout pregnancy, while gestational diabetes is usually diagnosed later, which could affect how much the developing retina is exposed to elevated glucose levels.
So where does this new research come in?
This study set out to clarify that gap by looking not just at whether maternal diabetes is linked to severe ROP, but whether the risk differs by diabetes subtype.
To do that, researchers analyzed a large NICU database from Vanderbilt University Medical Center, focusing on whether prenatal exposure to maternal diabetes was associated with Stage 3 to 5 ROP, the stages most likely to require treatment.
Who was included in the study?
The study included 3,139 preterm infants cared for between 2004 and 2021 who met standard ROP screening criteria based on gestational age or birthweight.
Out of that group, 311 infants, about 10%, developed Stage 3 to 5 ROP. Another 2,121 infants had no detectable disease and were used as the comparison group.
Maternal diabetes was present in 9% of cases, and the researchers further broke that down into gestational diabetes, T1D and T2D to see whether the risk differed across groups.
What did the researchers find?
After adjusting for known risk factors like gestational age, birthweight, sex, comorbidities and birth location, maternal diabetes was associated with a higher likelihood of severe ROP.
One detail that stands out is that infants exposed to maternal diabetes tended to have higher birthweights and gestational ages, which would normally point to a lower risk of ROP. Despite that, the increased odds of severe disease remained.
That disconnect suggests that relying only on birthweight and gestational age could overlook a subset of infants who appear lower risk but may still be vulnerable.
Did risk differ by diabetes subtype?
Yes—and this is one of the most important parts of the study.
The increased risk was driven by pregestational diabetes. Both T1D and T2D were significantly associated with Stage 3 to 5 ROP, with substantially higher odds compared to infants not exposed to maternal diabetes.
Gestational diabetes showed a trend toward increased risk, but the result was not statistically significant.
The authors note that this could be related to timing, since many of the most premature infants are born around the same window when gestational diabetes is typically diagnosed, which may lead to underrecognition.
What else stood out in the data?
As expected, lower gestational age and lower birthweight were still strongly linked to more severe ROP. Comorbidities like necrotizing enterocolitis, bronchopulmonary dysplasia and intraventricular hemorrhage were also more common in infants with advanced disease.
At the same time, infants exposed to maternal diabetes had lower rates of some of these complications and were more likely to be born in the study center rather than transferred in. That makes the diabetes signal harder to dismiss as a byproduct of overall illness severity.
In other words, even when some factors pointed toward a more stable clinical picture, the association with severe ROP still held.
Any limitations to consider when interpreting the findings?
This was an observational study, so it cannot prove that maternal diabetes directly causes severe ROP.
There were also gaps in maternal data, particularly for infants born outside the study center, including missing information on hemoglobin A1c levels. That means the study could not assess how the degree of glycemic control might influence risk.
In addition, the analysis relied on electronic health record data, which can introduce inconsistencies, although the researchers performed manual chart reviews to confirm key outcomes.
What might this mean for ROP screening decisions?
Current ROP screening guidelines focus heavily on gestational age and birthweight, with some flexibility for clinicians to screen additional high-risk infants.
This study suggests that maternal diabetes exposure could be one of those factors worth considering, especially because it may identify infants who do not fit the typical high-risk profile but still have an elevated chance of disease progression.
Anything else?
One of the more practical takeaways is that maternal diabetes does not follow the usual pattern clinicians expect when assessing ROP risk. The infants exposed to it may look more mature on paper, but still carry added risk once other variables are taken into account.
The authors also highlight the need for future research that includes more detailed measures of maternal glucose control to better understand what aspect of diabetes exposure is driving the association.
Take home.
Maternal diabetes, particularly T1D and T2D, was associated with higher odds of vision-threatening ROP in premature infants, even after accounting for traditional risk factors.
The findings suggest prenatal metabolic exposure may play a role in disease progression and could eventually help refine how clinicians identify and monitor at-risk infants.