Loading ...

Oral corticosteroids are a clinical staple during pregnancy — prescribed for asthma flares, autoimmune conditions, preterm labor risk, and fetal lung maturation, among other indications. And for just as long, many clinicians have worried about what they do to blood sugar.

That concern has real biological grounding. Corticosteroids drive insulin resistance, and that reality has shaped counseling and monitoring protocols for years. The assumption that steroid exposure during pregnancy meaningfully raises gestational diabetes risk has become almost reflexive.

A large cohort study published in JAMA Internal Medicine is now challenging that assumption, and the findings may prompt primary care providers to reconsider both what they tell patients and how aggressively they monitor.

Study Findings Challenge a Longstanding Assumption

The cohort study drew on nationwide Korean health data from 2010 through 2021, including 1,325,940 pregnancies from a pool of more than 3.8 million. Among eligible pregnancies, 6% involved oral corticosteroid exposure between weeks 1 and 27 of gestation — a rate that underscores how routine this clinical scenario is.

On the surface, the data seemed to confirm the prevailing assumption. Gestational diabetes occurred in 9.5% of corticosteroid-exposed pregnancies vs 7.36% in unexposed pregnancies — a gap that looks meaningful in the raw numbers. 

After propensity score-based weighting to adjust for comorbidities and other confounders, however, the risk ratio landed at just 1.01 (95% CI, 0.99-1.03), statistically indistinguishable from no effect. Subgroup analyses found no meaningful differences by maternal age, steroid indication, dosage, timing, or duration of exposure. The one exception: Women exposed between 4- and 6-weeks’ gestation showed a modest increase (weighted RR, 1.10; 95% CI, 1.03-1.17), a finding the authors noted without a clear explanation.

Not every clinician is ready to treat the question as settled. Florence Comite, MD, an endocrinologist and precision medicine specialist and founder of Comite Center for Precision Medicine & Healthy Longevity, said she remains cautious about corticosteroid prescribing in pregnancy regardless of what population-level data show. 

"I try to avoid prescribing corticosteroids like prednisone if I can. They're not great for anyone, especially pregnant patients — they drive up sugar and insulin and decrease bone turnover, leading to significant bone loss," Comite said.

Jeff Chapa, MD, national medical director of maternal-fetal medicine at Obtelecare, took a more measured view.

"Emerging evidence showing no significant link between oral corticosteroids and gestational diabetes is reassuring, though it doesn't change our approach to monitoring. What it does affect is counseling," Chapa said.

Hyperglycemia vs Gestational Diabetes: A Distinction That Matters

The study's headline finding rests on a distinction clinicians encounter in practice constantly: Steroid-induced hyperglycemia and true gestational diabetes are not the same thing.

Corticosteroids predictably cause transient glucose elevations, typically peaking within 24 to 48 hours of administration and resolving as the drug clears. That window can overlap with routine gestational diabetes screening, and when it does, an elevated result may not reflect a genuine diagnosis. Sorting out whether a value reflects baseline metabolic risk, steroid exposure, or true new-onset gestational diabetes requires clinical context the number alone cannot provide.

Monitoring and Counseling in Practice

"I always prescribe a continuous glucose monitor and also suggest another wearable: a sleep tracker so the patient can take an active role in managing her sugars," said Comite, who uses technology early as part of her monitoring approach.

David Ghozland, MD, an ob/gyn affiliated with Cedars-Sinai Medical Center in Los Angeles, flags repeat exposures as a separate risk category. 

"I'm also extremely cautious about repeat courses of steroids within the same week. That is not a number I dismiss lightly, especially in a patient who already has baseline insulin resistance. Anyone needing a second round of corticosteroids gets an automatic endocrinology consult," Ghozland said.

Andrews noted that early-pregnancy oral corticosteroid use opens an option many clinicians miss. 

"If a patient is taking corticosteroids early in pregnancy, then there is an option for an early glucose screening test, such as the 1-hour glucose tolerance test. This can be done in the first trimester," she said.

On counseling, getting ahead of fear matters as much as the clinical decision itself.

"The first thing I do is shut down the panic. Most patients hear 'diabetes risk' and immediately think we're giving them a lifelong disease. I sit down and say their blood sugar is going to spike for about 48 hours and then it's done," Ghozland said.

Emma Rodriguez, MD, a maternal-fetal medicine specialist at CHRISTUS Children's Center for Maternal and Fetal Care in San Antonio, Texas, shared the same theme. 

"I emphasize that most glucose changes are temporary and can be managed through monitoring, nutrition, and, in some cases, medication," Rodriguez said.

Chapa put the broader calculus plainly. 

"Untreated or poorly controlled maternal disease often poses a greater risk to the baby than appropriately used medication. Managing the mother's condition effectively is a key way to protect the pregnancy," he said.

"We're choosing a temporary glucose bump over a baby who may need a ventilator because their lungs aren't ready," Ghozland said.

Where Caution Still Applies

The null finding does not apply equally to every patient. High baseline insulin resistance changes the picture.

"Patients with obesity and PCOS are the ones I watch like a hawk. Their baseline insulin resistance is already high, so adding steroids on top of that is asking for trouble with glucose control," Ghozland said.

Rodriguez draws a clear clinical line around glycemic instability. "In patients who have poorly controlled pregestational or gestational diabetes, I consider admission for glycemic control of the anticipated steroid-induced hyperglycemia as well as for increased fetal surveillance," she said.

Andrews flagged prior difficult-to-control gestational diabetes as a separate concern. "I would be especially cautious in a patient who has a history of GDM that was difficult to control or a current diagnosis of poorly controlled GDM," Andrews said.

Reassuring Data, Individualized Care

A study of more than 1.3 million pregnancies that found no association between oral corticosteroids and gestational diabetes is not a finding to set aside lightly, and for many providers it is already shifting how they frame the conversation with patients.

Population-level reassurance doesn't erase individual risk, though. Monitoring remains standard, counseling still matters, and patients with preexisting metabolic vulnerabilities warrant closer attention.

Comite, who argues that nearly everyone carries some degree of metabolic vulnerability, offered a broad reminder: "Every pregnant woman needs to be cautious."

No disclosures were reported.