Associations Between Preconception Weight Loss Drug Use,Gestational Weight Gain,and Development of Gestational Diabetes

Abstract

Background:

Weight loss drugs (WLD) including glucagon-like peptide-1 agonists (GLP-1 agonists) are increasingly used for obesity to promote weight loss and insulin sensitivity. Currently, recommendations are to halt all WLD in pregnancy given undetermined safety profiles or previously identified teratogenic properties. Rebound weight gain following cessation of certain WLD has been identified in prior studies. Though weight gain is anticipated in pregnancy, it is important to limit excessive weight gain. We hypothesize that prepregnancy usage of WLD influences gestational weight gain and risk of gestational diabetes.

Methods:

Epic COSMOS was used in this retrospective observational cohort study. Queries were run on pregnant women between February 20, 2022, and February 19, 2025. Gestational weight gain was compared between those on WLD preconception and those not on such drugs preconception. Development of gestational diabetes was also compared between the two groups.

Results:

Patients who utilized WLD prior to pregnancy weighed more at their initial prenatal encounter compared with those who did not. Moreover, pregestational WLD users had a higher end-of-pregnancy weight and a greater difference in body mass index between beginning and end of pregnancy, compared with nonusers. Moreover, significantly more WLD users developed gestational diabetes in pregnancy than nonusers.

Conclusions:

Preconception use of WLD was associated with higher gestational weight gain. With increasing prevalence in both obesity and WLD use among reproductive-aged women, further work is needed to understand the impact of cessation of WLD on gestational weight gain and on gestational diabetes risk.

Introduction

Gestational weight gain has been shown to have a significant impact on pregnancy outcomes, including gestational diabetes, preeclampsia, and delivery complications including shoulder dystocia.^1–3 Limiting weight gain across pregnancy has been an important tool to decrease these complication rates; however, it may be difficult among women who struggle with obesity prior to conception. The prevalence of clinical obesity at the time of first prenatal visit has increased over the last two decades. ³

New weight loss drugs (WLD), in particular glucagon-like peptide-1 agonists (GLP-1 agonists), have been shown to be effective in weight and diabetes management. Interest in GLP-1 agonists has soared in recent years. Rebound weight gain after discontinuation, however, is a particularly difficult problem to address.^4,5 Limited research is available on the impact of these medications in pregnancy; therefore, discontinuation is recommended immediately upon conception. ⁶ It is unclear what impact discontinuation of these medications has on weight gain in pregnancy as well as the incidence of diabetes among these patients.

The purpose of this study is to evaluate the association between the use of weight loss medications prior to pregnancy and (1) weight gain in pregnancy and (2) incidence of gestational diabetes in pregnancy.

Materials and Methods

This study utilized the Epic COSMOS database, a large nationwide de-identified health database. Four separate queries were run to capture the differences in the change in weight throughout pregnancy as well as the incidence of gestational diabetes in women who had been on WLD within 8 months of establishing prenatal care compared with those who had not.

The four queries included data on all women who had an initial prenatal encounter and had taken WLD prior to pregnancy, women who had an initial prenatal care who had not taken WLD prior to pregnancy, and the last two queries were for the same groups of women (with and without a history of WLD) who had a documented delivery. All women seen for these visits between February 20, 2022, and February 19, 2025, were included.

For this study, patients were included in the WLD group if they had been on any of the following WLD—GLP-1 agonists (including semaglutide, tirzepatide, liraglutide, dulaglutide), metformin, phentermine resin, phentermine-topiramate, bupropion-naltrexone, or bupropion HCL.

An additional set of four queries was run on a subset of the above population who specifically used one of the GLP-1 agonists as listed above prior to conception. Those queries were run to capture the differences in the change in weight throughout pregnancy in women who had been on GLP-1 agonists within 8 months of establishing prenatal care compared with those who had not.

The encounter type “initial prenatal encounter” was used to filter for patients’ weight and body mass index (BMI) at the onset of pregnancy. The 10th revision of the International Classification of Diseases (ICD-10) code Z37* was used to filter for women at the time of their delivery. Patients with pregestational diabetes were identified using ICD-10 codes 024.0 and 024.1. Patients with gestational diabetes were identified by the ICD-10 code 024.41 Gestational Diabetes Mellitus in pregnancy.

COSMOS uses unique conventions of data retrieval and linking events to values. To ensure weights were documented from the visit, and height was not being routinely recorded at every visit, weight was linked to the encounter type or the date the visit was billed for the corresponding diagnosis. There is an exception for BMI. Due to weight being routinely recorded at each visit, height is not as reliably recorded. COSMOS generated a specialized option for “BMI Calculated,” which uses the patient’s current weight and the last recorded height to calculate BMI with the most robust pool of data.

Due to the nature of the COSMOS SlicerDicer tool, which provides aggregated, de-identified summary data, only descriptive statistics were used in this analysis; no inferential statistical comparisons were performed.

Results

All four queries

All four queries generated their own unique sample of the population. Women who presented at their initial prenatal encounter and did not have a documented history of WLD use resulted in 1,457,068 women, with an average weight of 172 pounds (with 95% data completeness), and an average calculated BMI of 31.6 (87% data completeness). There were 3,041,499 women who had a delivery without a history of WLD, who had an average weight of 185 pounds (87% data completeness), and who had an average calculated BMI of 32.6 (96% data completeness) at the time of delivery.

Women who presented at their initial prenatal encounter with a history of WLD included 42,250 cases with an average weight of 211 pounds (94% data completeness) and 35.9 calculated BMI (93% data completeness). Women who had a delivery with a history of using WLD included 132,903 with an average weight of 228 pounds (85% data completeness) and 39.5 calculated BMI (85% data completeness).

Differences in weight and BMI by history of WLD use

Women without a history of WLD use prior to pregnancy gained about 13 pounds on average over the course of their pregnancies. The average BMI among this group increased from 31.6 to 32.6, only a one-point increase throughout the course of their pregnancies (Fig. 1).

FIG. 1.

Weight gain in pregnancy compare by weight loss drug (WLD) use.

Conversely, women with a history of WLD use prior to pregnancy gained about 17 pounds on average over the course of their pregnancies. These patients began with an average BMI of 35.9 and delivered with an average BMI of 39.5, a four-point increase throughout the course of their pregnancies (Fig. 2).

FIG. 2.

Body mass index (BMI) across pregnancy by WLD use.

Differences in weight and BMI by history of GLP-1 agonist use

Given the current popularity of GLP-1 agonists, the authors further examined associations between preconception use of this drug in particular and gestational weight gain. Patients who had taken GLP-1 agonists prior to their pregnancy weighed more at their initial prenatal encounter relative to women who were not taking GLP-1 agonists prior to their initial encounter (232 pounds and 173 pounds, respectively) (Fig. 3). Women having taken GLP-1 agonists in preconception presented with a higher BMI as well (BMI 39.2 vs. 31.6) (Fig. 4).

FIG. 3.

Weight gain across pregnancy by glucagon-like peptide-1 (GLP-1) agonist use.

FIG. 4.

BMI across pregnancy by GLP-1 agonist use.

Additionally, women who did not take GLP-1 agonists in preconception had their BMI only change marginally (BMI difference +1.8 points) compared with women who had historically taken GLP-1 agonists (BMI difference +4.2 points) (Fig. 4). Patients on GLP-1 agonists prior to their pregnancy gained twice as much weight as patients who did not take GLP-1 agonists prior to pregnancy (weight gain 26 pounds and 13 pounds, respectively) (Fig. 3).

Gestational diabetes and preexisting diabetes mellitus type I and II

Of the 132,903 women who had a history of using WLD, 20% of them had a history of preexisting diabetes and about 45% developed gestational diabetes. This is substantially greater than the women who did not have a history of WLD use. In this latter group, only 2% had a diagnosis of pregestational diabetes, and only 11% developed gestational diabetes.

Discussion

In this study, the use of WLD prior to pregnancy was associated with an increase in average weight gain throughout the pregnancy as well as an increase in incidence of both pregestational and gestational diabetes. Use of GLP-1 agonists specifically in the preconception period was associated with a twofold increase in gestational weight gain in comparison to nonuse. This is congruent with the concept that metabolic disorders existing prior to pregnancy will persist during pregnancy. An increased average weight gain among these patients is also an important risk factor to consider in this patient population.

Women without a history of WLD gained on average 13 pounds during pregnancy, and an approximately one-unit increase in BMI (31.6–32.6). Those with a history of WLD gained about 17 pounds, and their average BMI increased by 3.6 units (35.9–39.5). While the BMI increase appears larger than expected for a 17-pound weight gain, this discrepancy reflects that weight and BMI were aggregated separately across encounters rather than recalculated from identical weight–height pairs. Additionally, as aggregate values do not require 100% data completeness, these models were run with 85%−95% data completeness; therefore, missing data used to calculate the average weight and BMI changes may have impacted these results slightly. However, the key pattern remains: Women with WLD use preconception started pregnancy at a higher weight and gained slightly more over the course of the pregnancy.

The increased risk of gestational diabetes observed among the WLD use group was notable. The American College of Obstetricians and Gynecologists’ guidelines do not currently recommend early screening for pregestational and gestational diabetes due to WLD use alone; however, early screening is recommended for those with prepregnancy BMI >40 kg/m². This in turn reflects a large portion of the study population using preconception WLD. ⁷

The strength of this study includes the use of a large national database with data that are generalizable. The COSMOS database allows granular data collection on a macroscopic scale, facilitating analysis of large populations. This is particularly useful in the instance of trending patient weight and BMI. This database also provides significant limitations, however, as data accuracy is limited to the accuracy of the documenting providers. Additionally, due to limitations within the database, multiple variables could not be compared simultaneously between patients, limiting the authors’ ability to control for confounds. While the inability to perform inferential analyses is a methodological limitation, the findings are derived from a large, diverse, population-level dataset, enhancing the generalizability and interpretive value of the descriptive trends observed.

Future studies should examine these relationships in a prospective direction to better account for potential confounding factors and obtain more complete individual-level data. Additionally, such studies should consider examination of each individual WLD and its relationships with gestational weight gain.

Conclusions

With increasing prevalence in both obesity and WLD use among reproductive-aged women, further work is needed to understand the impact of cessation of WLD on gestational weight gain and on gestational diabetes risk. Identifying an appropriate time frame in which to cease use prior to pregnancy, not only for fetal safety but to avoid excessive weight gain in pregnancy, is critical. In addition, attention to preconception WLD use, regardless of BMI at the first prenatal visit, should be considered when deciding whether to perform early glucose screening.

Improving counseling for such patients would allow reproductive-aged individuals who pursue healthy weight loss in the preconception period to continue such benefits in pregnancy.

Authors’ Contributions

K.M.: Research question formulation, research design, manuscript creation and editing. E.C.: Research design, Statistical analysis, manuscript creation and editing. S.L.: Research design, manuscript creation and editing. V.H.: Manuscript creation and editing. T.J.: Manuscript creation and editing. B.P.: Manuscript creation and editing.