GLP-1 in PCOS: the fertility evidence gaps

Polycystic ovary syndrome (PCOS) affects an estimated 8 to 13 percent of women of reproductive age worldwide^[5], making it the most common endocrine disorder among that group. Its hallmark features insulin resistance, androgen excess, and disrupted ovulation overlap neatly with the metabolic pathways that semaglutide targets. That overlap has made GLP-1 receptor agonists an appealing candidate for PCOS management. But a published reply to a letter-to-the-editor in Clinical Nutrition ESPEN, indexed in August 2026 under PMID 42674355, is part of a continuing scientific dialogue that reveals how contested the evidence actually is.

The editorial chain concerns a pilot prospective study (PMID 41421448) that enrolled 20 women with PCOS, obesity, and prediabetes. Participants received combined subcutaneous semaglutide (0.25 to 0.5 mg weekly) and metformin (1,000 mg daily) for six months. The authors reported significant weight reduction and improvements in fertility markers. A comment on that paper raised methodological objections, and the reply to that comment defends the original authors' conclusions. The exchange is a useful window into what the field agrees on, and what it does not.

What the stronger trials show

Moving beyond the pilot study, a prospective randomized controlled trial published in Reproductive Biology and Endocrinology in 2025 compared metformin monotherapy against a combination of semaglutide and metformin in 100 overweight or obese women with PCOS^[1]. After 16 weeks of treatment, the combination group lost a mean of 6.09 kg compared with 2.25 kg in the metformin-only group. Menstrual cycle recovery was documented in 72.5 percent of the combination group versus 42.3 percent with metformin alone. During a 24-week pregnancy tracking period that followed treatment, 35 percent of women in the combination group achieved a natural pregnancy compared with 15 percent in the metformin group^[1]. These are notable numbers from a single-centre, open-label trial of 80 completers, and the authors themselves flag small sample size and short duration as central limitations.

A 2023 meta-analysis in BMC Endocrine Disorders synthesized 11 randomized controlled trials encompassing 840 women with PCOS to assess GLP-1 receptor agonists across the class^[2]. The pooled finding for spontaneous pregnancy rate was encouraging: a relative risk of 1.72 (95% CI 1.22 to 2.43). However, the same analysis found no statistically significant difference for total pregnancy rate or IVF pregnancy rate. The authors also noted that trial duration mattered substantially: women in studies lasting 24 weeks or more showed meaningfully greater menstrual regularity improvements than those in 12-week protocols^[2]. Most of the included trials lasted only 12 weeks, which means the majority of available data may be systematically too short to capture the outcomes that matter most.

The methodological debate the dialogue reveals

A systematic review that integrated IVF data, ovarian physiology, and molecular mechanisms of GLP-1 receptor agonists captured the broader evidence picture^[3]. Across included studies, evidence quality was found to be predominantly low, with marked heterogeneity in study populations, GLP-1 agent types, dosing regimens, and follow-up periods. A consistent pattern across multiple reviews is that existing trials were largely short (12 weeks or less), small, and open-label. Crucially, none of the included RCTs in the major published reviews had specifically evaluated semaglutide at obesity doses; the agents that now dominate clinical practice were essentially absent from the formal trial evidence base^[3].

These concerns sit at the core of the editorial exchange. Pilot and small RCTs enrolling fewer than 100 participants, running for 12 to 16 weeks, using open-label designs, and operating from a single centre cannot reliably answer whether observed fertility improvements stem from the drug itself or from the weight loss it produces. That distinction matters clinically. If the benefit is entirely weight-mediated, then any intervention that achieves comparable weight loss in PCOS should produce comparable reproductive improvements. If the drug has direct effects on ovarian physiology through GLP-1 receptor signalling in granulosa cells, the treatment has implications beyond weight management.

The direct-versus-indirect effect question

A 2026 analysis in Frontiers in Nutrition applied a Bayesian framework to assess the evidence for GLP-1 receptor agonists in PCOS across metabolic and reproductive domains^[4]. The authors distinguished between two mechanistic pathways: the indirect route through weight reduction and improved insulin sensitivity, where the evidence is relatively consistent, and the proposed direct pharmacological effects on ovarian function, where evidence is substantially weaker. As the analysis puts it, outcome definitions, denominators, and follow-up windows varied substantially across studies, and reproducible reproductive endpoints remain sparse. Most trials use menstrual bleeding as a proxy for restored ovulatory function rather than confirming ovulation with hormonal monitoring or ultrasound^[4].

There is also a safety dimension that the broader fertility debate cannot sidestep. Semaglutide carries a precautionary contraindication in pregnancy based on animal data, and the label recommends discontinuing the drug before attempting conception. The Frontiers in Nutrition analysis noted that the absence of a stable teratogenic pattern in available data should not be interpreted as proof of safety^[4]. Limited pregnancy-exposure records, heterogeneous timing of drug cessation relative to conception, and underpowered follow-up prevent definitive conclusions about periconception safety. For a drug being discussed in the context of fertility treatment, that gap is not trivial. See the site's dedicated discussion of semaglutide and menstrual changes for a fuller account of the reproductive side-effect picture.

What the scientific community is calling for

Multiple systematic reviews and the Frontiers in Nutrition framework converge on the same set of research priorities^[3][4]. Trials need standardised reproductive endpoints, including confirmed ovulation rather than menstrual bleeding alone. Studies must run long enough, at least 12 months, to capture effects that short protocols miss. Mediation analyses are needed to separate weight-loss effects from direct pharmacological effects on fertility. And pregnancy-exposure registries need systematic data on drug washout timing and foetal outcomes.

Until those trials exist, the editorial dialogue around the pilot study in Clinical Nutrition ESPEN reflects a genuine scientific impasse. The signal from individual studies is promising enough to generate clinical interest and ongoing trials. The evidence base as a whole is not yet strong enough to support definitive conclusions about whether semaglutide improves fertility in PCOS independently of weight loss, or to settle the question of optimal duration and dosing for reproductive outcomes. Readers interested in the broader picture of how semaglutide works can find that context in the site's mechanism explainer. For a condition as prevalent as PCOS, the open fertility questions remain significant.

This article is for informational purposes only and does not constitute medical advice. Consult a healthcare provider before making any treatment decisions.