Prescribing Trends in Glucagon-like Peptide-1 Medications Among Pregnant and Postpartum Persons

Abstract

With rising obesity rates and increasing glucagon-like peptide-1 receptor agonist (GLP-1RA) use, understanding perinatal prescribing patterns is important. We conducted a retrospective cohort study to examine semaglutide and tirzepatide prescribing among pregnant patients in the U.S. from 2019–2024. We analyzed prescriptions during the year before and after delivery, grouping deliveries into 6-month periods and applying segmented linear regression with data-driven change-point detection to identify prescribing trend shifts. GLP-1RA prevalence increased from 0.2 to 6.4 per 1,000 pre-delivery and from 0.3 to 14.6 per 1,000 post-delivery, with significant prescribing change-points indicating accelerated prescribing in June 2022 for the pre-delivery period and March 2021 for the post-delivery period. These findings suggest rapid adoption of GLP-1RAs in the perinatal period and underscore the need for evidence-based safety data for these medications.

PRECIS

Glucagon-like peptide-1 (GLP-1) receptor agonist prescribing to pregnant and postpartum people increased significantly from 2019 to 2024, highlighting a need for perinatal safety data.

INTRODUCTION

Prescription rates for glucagon-like peptide-1 receptor agonists (GLP-1RA) have increased in the general U.S. population.¹ While over 40% of pregnant women globally are overweight or obese,² current guidelines recommend against GLP-1RAs during pregnancy and breastfeeding due to limited safety data.^3,4 Although some data thus far have been provisionally reassuring,^5,6 population-based U.S. prescribing trends remain uncharacterized. We thus evaluated trends in GLP-1RA prescribing among pregnant people in a national electronic health records cohort.

METHODS

We used the TriNetX database to assess GLP-1RA prescribing trends among U.S. individuals aged 25–50 with documented deliveries between 1/1/2019 and 12/31/2024. Detailed methods are in Appendix 1 (available online at http://links.lww.com/xxx). This work was exempt from human participant review by the Carilion Clinic human research protection office. For each delivery, we identified ≥1 tirzepatide or semaglutide prescription during the year before delivery (predelivery) and the year after delivery (post-delivery). To assess temporal trends, deliveries were grouped into 6-month calendar periods by delivery date, yielding 12 time periods for analysis. For each delivery, we defined two independent 12-month exposure windows: the 365 days before delivery (pre-delivery) and the 365 days after delivery (post-delivery). We calculated the proportion of individuals with ≥1 GLP-1RA prescription within each window for each time period; individuals could contribute to one or both windows independently. Common comorbidities were identified using ICD-10 codes. Segmented linear regression was used to detect change points in prescribing trends. Monotonic trends were also assessed using a Cochran-Armitage test (see Appendix 1, available online at http://links.lww.com/xxx). Analyses were performed in SAS 9.4, with two-sided p<0.05 considered statistically significant.

RESULTS

Across all years (2019–2024), deliveries per 6-month period ranged from 59,660 to 69,705. In the most recent year of data (2024), the mean age was 33.3 (SD=5.22) years, with the sample identified as 65.5% White, 18% Black, and 9.9% Hispanic/Latine. As shown in Figure 1, the prevalence of GLP-1RA prescribing before delivery rose from 0.2 (January 2019) to 6.4 per 1,000 (June 2024). Pre-delivery prescribing was flat before June 2022 and accelerated thereafter (increasing 2.9 per 1,000/year; 95% CI=2.5–3.4). Post-delivery prescribing rose from 0.3 (January 2019) to 14.6 per 1,000 by June 2024; rates were flat before March 2021 and accelerated after (increasing 4.4 per 1,000 /year; 95% CI=3.8–5.0). Monotonic trend significance was confirmed using Cochran-Armitage tests (p < 0.0001 for both pre- and post-delivery windows).

Figure 1:

Trends in predelivery (A) and postdelivery (B) prescribing of GLP-1 receptor agonists (GLP-1RAs), including semaglutide and tirzepatide, among U.S. individuals with a delivery claim between 2019 and 2024. The y-axis reflects the number of individuals per 1,000 deliveries with ≥1 GLP-1RA prescription during the year before delivery (A) or the year after delivery (B), grouped by 6-month calendar periods. Predelivery prescribing rose from 0.2 per 1,000 (January 2019) to 6.4 per 1,000 (June 2024), with a data-driven inflection point in June 2022 (τ=7.80; 95% CI, 6.97–8.64). Postdelivery prescribing rose from 0.3 per 1,000 (January 2019) to 14.6 per 1,000 (June 2024), with a change point in March 2021 (τ=5.41; 95% CI, 3.99–6.82). Colored lines represent predicted trends from segmented (piecewise) linear regression models, and dots represent observed prescribing rates. Model R²>0.97 for both periods.

Among those who delivered in 2024 and had received a GLP1-RA in the year preceding and following delivery, 57.3% (95% CI: 55.1%−59.5%), were obese in the year before delivery. In addition, 38.9% (95% CI: 36.8%−41%) had anxiety disorders, 29.3% (95% CI: 27.3%−31.3%) had mood disorders, 21.4% (95% CI: 19.6%−23.2%) had gestational diabetes, and 16.4% (95% CI: 14.8%−18%) had type 2 diabetes.

DISCUSSION

We observed a rise in GLP-1RA prescribing to pregnant and postpartum people in the U.S. from 2019 to 2024, with change points aligning with broader adoption of semaglutide and tirzepatide. For context, there were 3.6 million U.S. births in 2024;⁷ if 6.4 per 1,000 of pregnant people received a GLP-1RA in the year before delivery, this equates to over 20,000 exposures. Despite guidelines advising against use in pregnancy and lactation due to limited safety data,⁴ our data shows that prescribing is increasingly common before and after delivery, primarily among patients with obesity.

GLP-1RAs may provide some metabolic benefits relevant to pregnancy, including reduced risk of gestational diabetes, hypertensive disorders, and postpartum cardiometabolic disease.⁶ Preclinical data and case reports have not shown teratogenicity, but robust human safety data remain lacking.⁴ Given the high prevalence of obesity in pregnancy and rising perinatal GLP-1RA use,^1–3 clarifying safety and efficacy has become an urgent clinical and research priority. As a limitation, adjusted or subgroup analyses could not be performed because TriNetX provides aggregated, period-level data rather than individual-level covariates required for multivariable modeling. Additional study is needed to evaluate demographic and clinician-level factors, including insurance coverage and prescriber specialty.

As GLP-1RA use increases, these findings suggest that obstetric clinicians may increasingly encounter patients who are using or have recently used these medications during the perinatal period. This may prompt consideration of factors not traditionally emphasized in routine prenatal and postpartum care, such as nutritional adequacy in the context of recent weight loss, or a history of eating disorders or substance use. While behavioral interventions for postpartum weight loss have shown limited effectiveness,⁸ GLP-1RAs could represent a future option for select patients.⁹ Further research is needed to clarify potential frameworks for thoughtful, evidence-based GLP-1RA prescribing in pregnancy and lactation.