Systematic review quantifies the incretin-contraceptive interaction

A systematic review published as an abstract in Obstetrics & Gynecology reports that GLP-1 and dual GLP-1/GIP medications leave estrogen levels from combined oral contraceptives (COCs) essentially untouched, but can lower and delay absorption of the progestin component — with the biggest effects seen with tirzepatide and short-acting agents ^[1]. The authors pair that analysis with what they describe as the first documented case of oral contraceptive failure in a person taking semaglutide ^[1].

What the review found

Following PRISMA methods, the researchers searched PubMed, EMBASE, Web of Science and ClinicalTrials.gov, and also included FDA reviews and package inserts ^[1]. Of 372 records screened, 10 studies covering 305 participants met criteria: seven randomized crossover trials and three FDA reviews ^[1]. Most enrolled healthy premenopausal participants; four evaluated postmenopausal women ^[1].

Ethinyl estradiol exposure was unchanged across studies ^[1]. The progestin picture was different and varied by drug. Semaglutide and dulaglutide showed minimal change in levonorgestrel (LNG) area under the curve — a measure of total drug exposure — ranging from −8% to +5% ^[1]. Tirzepatide, lixisenatide, and exenatide given immediately before the pill reduced LNG maximum concentration (Cmax) by 45–66% and LNG AUC by 18–23% ^[1].

Across the pooled data, the time to peak concentration was consistently pushed later, by a mean difference of 1.16 hours (95% CI, 0.10–2.22) ^[1]. Short-acting agents produced the largest pharmacokinetic changes when taken before the contraceptive ^[1]. The authors also note a theoretical model of delayed gastric emptying that predicts these timing delays may persist longer with tirzepatide, even though the stomach-slowing effect of these drugs tends to fade over time ^[1].

The case report

The same report describes a 39-year-old patient with obesity who conceived after four months of weekly semaglutide while using a monophasic ethinyl estradiol/levonorgestrel pill ^[1]. The authors call it the first documented COC failure on semaglutide ^[1]. A single case cannot establish that the medication caused the failure, and the review itself found semaglutide had minimal effect on levonorgestrel exposure ^[1].

Why it matters for patients

Millions of people taking GLP-1 drugs are of reproductive age, and questions about whether these medicines interfere with the pill come up constantly. This review offers the most specific numbers so far, and its bottom line is nuanced rather than alarming: the authors conclude that GLP-1 and dual GLP-1/GIP agonists "generally preserve estrogen exposure but can delay and reduce progestin bioavailability," and that these changes are "unlikely to compromise COC efficacy in typical use" ^[1].

They do flag areas where the margin is thinner. Counseling, they write, should cover timing strategies, backup contraception during high-risk periods, and caution with progestin-only and emergency contraceptive methods ^[1]. That last point follows logically from the data: if the drop is concentrated in the progestin, then methods that rely on progestin alone have less cushion than a combined pill that still delivers full estrogen exposure ^[1].

The differences between molecules also matter. Under this analysis, semaglutide (Ozempic, Wegovy, Rybelsus) and dulaglutide behaved differently from tirzepatide (Mounjaro, Zepbound) and the short-acting agents lixisenatide and exenatide ^[1]. That means general statements about "GLP-1 drugs and birth control" may not apply equally to every product.

Several limitations are worth keeping in view. This is a conference abstract presented at the American College of Obstetricians and Gynecologists' Annual Clinical and Scientific Meeting, not a full peer-reviewed paper, and the underlying pool is small — 305 participants across 10 studies, most of them healthy volunteers rather than people using contraception in daily life ^[1]. The abstract does not report real-world pregnancy rates among people on these medications, so how often these pharmacokinetic shifts translate into actual contraceptive failure is not yet known from this work ^[1].

What happens next

The abstract appears in the May 2026 supplement of Obstetrics & Gynecology (147(5S):6S) ^[1]. The sources do not indicate whether a full manuscript, updated labeling, or new professional guidance will follow.