WHAT’S NEW ….. CLINICALLY
A new relevance of AMH levels in IVF?
Whether anti-Müllerian hormone (AMH) levels matter during IVF in patients with polycystic ovary syndrome (PCOS) has, to the best of our knowledge, never been seriously investigated—until now. A recent Chinese study published in Reproductive Biology and Endocrinology aimed to do exactly that.
Though retrospective in nature and thus subject to the usual limitations of this study design, the investigation included 4,719 women diagnosed with PCOS, aged between 20 and 40. Unfortunately, the paper does not specify the criteria used for diagnosing PCOS—an important omission. It also fails to clarify which PCOS phenotypes were included and in what proportions—details that are highly relevant and would have added valuable context.
Nonetheless, the data are intriguing. The researchers divided the study population into three groups based on the 25th and 75th percentiles of AMH levels:
Low: <4.98 ng/mL, n=1,198
Medium: 4.98–10.65 ng/mL, n=2,346
High: >10.65 ng/mL, n=1,175
The study assessed live birth rates, clinical pregnancy rates, cumulative pregnancy rates, and miscarriage rates.
Live birth rates for fresh embryo transfers were 39.8%, 35.9%, and 30.4% for the low, medium, and high AMH groups, respectively (P=0.017). Miscarriage rates were 11.3%, 17.1%, and 21.8% (P=0.018), while clinical and cumulative pregnancy rates did not differ significantly across groups. After adjustment, the differences remained statistically significant for live birth rates between the high and low AMH groups, and for miscarriage rates between the low and both the medium and high AMH groups. In other words, in PCOS patients, high AMH levels were associated with lower live birth rates and higher miscarriage rates (1).
What makes this paper especially interesting is its resonance with findings from nearly a decade ago. In 2016, the Center for Human Reproduction (CHR) published a study in the Journal of Translational Medicine showing a similar trend in infertile women—though not necessarily PCOS patients. That study observed a U-shaped relationship: at younger ages, higher AMH was associated with improved live birth rates; however, as age advanced, miscarriage rates increased with rising AMH, leading to reduced live birth rates despite still-decent pregnancy rates.
Like the current Chinese study, CHR’s investigators concluded that excessively high AMH levels—through still unclear mechanisms—reduce live birth rates, largely by increasing miscarriage rates (2). And this is precisely what the new study suggests as well.
The key question that remains is: How does very high AMH contribute to miscarriage?
References
1. Zhao et al., Reprod Biol Endocrinol 2025;23:15
2. Gleicher et al., J Transl Med 2016;14:172
Gestational carrier (GC) pregnancies are riskier than widely appreciated
A recent study by California investigators, analyzing a national dataset of 14,312,619 deliveries, provided compelling insights into gestational carrier (GC) cycle outcomes. Published within the past year, the study offers one of the most credible datasets on this topic to date. Somewhat surprisingly—given the supposedly rigorous selection of GCs—it revealed that GC pregnancies are associated with a notable increase in adverse pregnancy outcomes (1).
An earlier investigation—a systematic review and meta-analysis on the same subject—had already indicated a significant rise in GC pregnancies, reporting a 55% increase between 2017 and 2020 (2). There is every reason to believe that this upward trend has accelerated further in the post-COVID era.
In the more recent study, GC pregnancies were more likely to involve multiple gestations (14.7% vs. 1.8%, P<0.001), placental abruption (3.5% vs. 1.1%), and low-lying placenta (1.6% vs. 0.2%). Among singleton pregnancies, GC cases showed higher rates of late preterm birth (10.8% vs. 6.4%), peri-viable deliveries (1.1% vs. 0.4%), and postpartum hemorrhage (12.2% vs. 4.1%). Interestingly, the likelihood of Cesarean delivery was higher in non-GC pregnancies (31.6% vs. 23.6%), and the strength of most other associations diminished in multiple (primarily twin) pregnancies.
These last two findings are noteworthy in their own right. They highlight two important issues: the persistently inflated national Cesarean section rate and the long-held but questionable belief that twin pregnancies pose significantly greater maternal and neonatal risks. Both topics have been extensively discussed in the literature by investigators from the Center for Human Reproduction (CHR) in New York City, so we will not revisit them in detail here.
What conclusions can be drawn from these two datasets?
(i) GC pregnancies should be classified as higher-risk and managed accordingly with appropriate medical attention.
(ii) It may be time to reevaluate the GC selection process. At CHR, especially since the COVID-19 pandemic, we have observed a noticeable decline in the quality of GC candidates approved by many agencies.
References
1. Masjedi et al., J Assist Reprod Genet 2025;42:201-211
2. Matsuzaki et al., JAMA Network open 2024;7(7):e2422634
More on Endometrial Scratching
The field of IVF certainly does not lack controversial treatments, but few have generated as much contradictory outcome data as Endometrial Scratching (ES). When such discrepancies arise, they can almost always be traced back to two core issues: differences in patient populations and/or variations in how the treatment was applied. Unfortunately, peer review in our field is still too often inattentive to these fundamental factors, even though they are essential to properly interpreting any clinical study.
A particularly striking (or should we say “terrible”) example is a recent paper by U.S. and Brazilian investigators published in JARG (1). This paper exemplifies many of the problems currently plaguing the peer review process in reproductive medicine—and, in doing so, contributes to the confusion surrounding numerous clinical treatments.
The study in question was a retrospective case-control analysis, which inherently introduces bias, as there must have been underlying reasons why some patients received ES while others did not. This bias is indirectly acknowledged in the paper by the surprising finding that patients who underwent ES had higher rates of endometritis. One could reasonably argue that if those patients had not been selectively assigned to ES, the treatment might have shown a benefit over non-scratching.
Nevertheless, the authors concluded that endometrial scratching prior to frozen euploid embryo transfer did not improve pregnancy outcomes and was associated with an increased risk of second-trimester pregnancy loss. One could also interpret the findings to suggest that ES may contribute to chronic endometritis, which in turn could increase miscarriage risk. However, such interpretations are speculative and, ultimately, invalid in this case—because the authors failed to appropriately adjust their already marginal statistical findings.
We highlight this paper to underscore a point that has been made many times before: publishing studies like this—even when authored by prominent names—does not help to resolve controversial issues. On the contrary, such papers tend to deepen confusion and delay progress by feeding ongoing debates rather than clarifying them. The result is that patients may continue to miss out on the most appropriate treatments.
Reference
Lindekugel et al., J Assist Reprod Genet 2025. doi: 10.1007/s10815-024-03360-x. Online ahead of print.