Live birth rates in natural compared to artificial frozen blastocyst transfer cycles.

This research did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.

A total of 905 FET cycles were analysed, which included 164 NC-FET and 741 AC-FET. The baseline characteristics of the patients are summarised in Table 1 . The mean age of the women was 35.1 years (SD: 4.2), while the mean age of their partners was 37.6 years (SD: 5.6). There were no significant differences in age between the NC-FET and AC-FET groups ( P = 0.174 and P = 0.304, respectively). However, BMI and AMH levels differed significantly between the groups, being higher in the AC-FET group compared to the NC-FET group ( P = 0.006 for BMI and P = 0.034 for AMH, respectively). The distribution of infertility diagnoses was comparable between the two groups, with ovulation disorders, male factor infertility, and idiopathic infertility being the most prevalent causes. Basic characteristics of patients according to FET protocol. Data are presented as mean ± SD or as n (%). AC, artificial cycle; AMH, anti-Müllerian hormone; BMI, body mass index; FET, frozen embryo transfer; NC, natural cycle; SD, standard deviation; NS, non-significant. Data on gravidity and parity were recorded for only 426 patients, but there were no statistically significant differences between the two study cohorts. Of these 426 women, 34% had never been pregnant, and 50% were nulliparous. Table 2 outlines the clinical characteristics at the time of FET. Endometrial thickness was significantly lower in the NC-FET group compared to the AC-FET group (7.9 vs 8.7 mm, P = 0.002). The number of blastocysts transferred per cycle was comparable between groups, with the majority of cycles (68%) involving a single blastocyst transfer. Clinical characteristics at the time of FET. Comparison between patient groups by either unpaired t -test or chi-squared test. AC, artificial cycle; FET, frozen embryo transfer; NC, natural cycle; SD, standard deviation. Table 3 presents the clinical outcomes based on the FET protocol. The overall positive β-hCG rate was 53%, with no significant difference between the NC-FET (56%) and AC-FET (52%) groups. However, among patients who had a positive β-hCG, the incidence of spontaneous abortion was significantly lower in the NC-FET group (9%) compared to the AC-FET group (20%) ( P = 0.013). The live birth rates per ET were significantly higher in the NC-FET group (43%) than in the AC-FET group (30%) ( P = 0.001). Likewise, the clinical pregnancy rates per ET were also higher in the NC-FET group (48 vs 41%), although this difference did not reach statistical significance ( P = 0.088). Clinical outcomes according to FET protocol. Data are presented as mean ± SE or as n (%). Comparison between FET protocols In multiple pairwise comparisons, the P -values were adjusted using the Benjamini–Hochberg method AC, artificial cycle; β-hCG, beta human chorionic gonadotropin; ET, embryo transfer; FET, frozen embro transfer; NC, natural cycle; SE, standard error A total of 295 live births resulted in the delivery of 320 neonates, with multiple gestations occurring in 8% of cases. The distribution of singleton and twin births varied between the study groups, with 85% of NC-FET births being singletons compared to 94% in the AC-FET group ( P = 0.015). Singleton birth weights were lower in the NC-FET group compared to those in the AC-FET group ( P = 0.016). Similarly, twin birth weights were lower in the NC-FET group for both first-born ( P = 0.008) and second-born infants ( P = 0.032). The influence of patient characteristics on the likelihood of live birth following FET is presented in Table 4 . In the univariate regression analysis, it was found that NC-FET had a significantly higher likelihood of achieving live birth compared to AC-FET, with an unadjusted OR of 1.76 ( P = 0.001). Conversely, factors such as advanced maternal age (unadjusted OR: 0.92, P < 0.001), advanced paternal age (unadjusted OR: 0.96, P = 0.002), and higher BMI (unadjusted OR: 0.93, P = 0.008) were associated with decreased odds of achieving live birth. The unadjusted and adjusted odds ratios for the influence of patient characteristics on live birth after FET. P -value for univariate logistic regression. P -value for backward stepwise multiple logistic regression after adjustment for all variables listed in the table. AMH, anti-Müllerian hormone; BMI, body mass index; CI, confidence interval; FET, frozen embryo transfer; OR, odds ratio. After adjustment for all patient characteristics presented in Table 4 , the FET protocol was no longer significantly related to live birth outcomes. The analysis revealed that independent predictors of live birth included maternal age, BMI, infertility diagnosis, and the number of blastocysts transferred. Specifically, advanced maternal age (adjusted OR: 0.94, P = 0.034) and higher BMI (adjusted OR: 0.91, P = 0.012) were linked to lower odds of live birth. In addition, male factor infertility (adjusted OR: 0.50, P = 0.038) and cases of idiopathic infertility or uterine abnormalities (adjusted OR: 0.38, P = 0.005) were negatively associated with live birth in comparison to ovulation disorder. Notably, transferring two blastocysts significantly increased the likelihood of achieving a live birth, with an adjusted OR of 2.50 ( P = 0.001).

This retrospective, single-centre study included 905 patients who underwent vitrified-thawed blastocyst transfer at BetaPlus Center for Reproductive Medicine between January 2016 and April 2023. Our centre does not offer a donation programme; thus, all blastocysts were derived from patients’ own oocytes. The study received approval from the Institutional Review Board (Number 16/2023), and all participating couples provided written informed consent for the use of anonymised medical data for research purposes. The flow chart of the study is presented in Fig. 1 . Flow chart of the study. Patients included in this study had supernumerary blastocysts from previous ovarian stimulation cycles for IVF at our centre. The number of blastocysts transferred was determined by the number of previously vitrified and thawed blastocysts per carrier, along with patient preference, as our centre, at that time, did not enforce a mandatory single blastocyst transfer policy. A normal uterine cavity, confirmed via transvaginal ultrasound within 1–3 months before initiating a thawing cycle, was required for inclusion. If endometrial polyps or other uterine abnormalities were suspected, operative hysteroscopy was performed before proceeding with thawing. Patients were also required to have thyroid-stimulating hormone (TSH: <2.5 mIU/L) and prolactin levels within the reference limits of the laboratory providing the results. Patients were excluded from the study if they met any of the following criteria: FET cancellation before thawing for any reason. Blastocysts derived from previously cryopreserved oocytes. Embryos initially cryopreserved at the cleavage stage (day 2–4) and subsequently cultured to blastocyst. Use of different endometrial preparation agents not specified in this study. Presence of Müllerian anomalies of the uterus. Diagnosis of extrauterine or heterotopic pregnancy. Administration of human chorionic gonadotropin (hCG) injection. FET cancellation before thawing for any reason. Blastocysts derived from previously cryopreserved oocytes. Embryos initially cryopreserved at the cleavage stage (day 2–4) and subsequently cultured to blastocyst. Use of different endometrial preparation agents not specified in this study. Presence of Müllerian anomalies of the uterus. Diagnosis of extrauterine or heterotopic pregnancy. Administration of human chorionic gonadotropin (hCG) injection. Blastocysts were vitrified on day 5 or 6 according to predefined criteria: a blastocoel grade of 2–4, an inner cell mass (ICM) of at least grade B, and a trophectoderm of at least grade B as outlined in the Istanbul Consensus and update ( Alpha Scientists in Reproductive Medicine and ESHRE Special Interest Group Embryology 2011 , Coticchio et al. 2025 ). The scoring system is based on a combination of numbers and letters. The number (ranging from 1 to 6) indicates the width of the blastocoel. The first letter represents the quality of the ICM, graded from A to D, while the second letter reflects the quality of the trophectoderm, also graded from A to D ( Gardner & Schoolcraft 1999 ). Figure 2 presents blastocysts resulting in live births, shown with their morphological grades. Blastocyst images with grades. Vitrification was performed using Kitazato vitrification solutions and an open system carrier (Cryotop, Kitazato, Japan). Blastocysts were incubated in an equilibration solution for 15 min, followed by placement in a vitrification solution and multiple washes over 60 s before being loaded onto the carrier. The carrier was then plunged into liquid nitrogen and capped. Blastocysts were graded before ET, which was conducted using standard transfer catheters (Embryo Transfer Catheter Set, Cook). Two protocols were used: natural cycle FET (NC-FET) and artificial cycle FET (AC-FET). NC-FET: no medication was administered. The transfer day was determined using pre-planned luteinising hormone (LH) ovulation tests. Figure 3 shows an example of test strip images provided by patients. If ovulation was not confirmed, the cycle was cancelled and an AC-FET protocol was scheduled in the following cycle. AC-FET: patients received oestradiol (oral or transdermal), with some also receiving triptorelin acetate, followed by vaginal progesterone capsules to prepare the endometrium. The goal was an endometrial thickness ≥8 mm, assessed via transvaginal ultrasound on cycle days 9–11. If satisfactory, progesterone was initiated, and FET was performed 5 days later. Anovulatory patients were exclusively included in this group. NC-FET: no medication was administered. The transfer day was determined using pre-planned luteinising hormone (LH) ovulation tests. Figure 3 shows an example of test strip images provided by patients. If ovulation was not confirmed, the cycle was cancelled and an AC-FET protocol was scheduled in the following cycle. AC-FET: patients received oestradiol (oral or transdermal), with some also receiving triptorelin acetate, followed by vaginal progesterone capsules to prepare the endometrium. The goal was an endometrial thickness ≥8 mm, assessed via transvaginal ultrasound on cycle days 9–11. If satisfactory, progesterone was initiated, and FET was performed 5 days later. Anovulatory patients were exclusively included in this group. If a cycle was cancelled due to inadequate endometrial thickness, the next cycle included additional oestradiol supplementation in the form of patches. A small subset of patients who failed to achieve satisfactory endometrial thickness in multiple attempts was excluded from the analysis. Example of a test strips image provided by patients. The primary outcome was the live birth rate (live birth of an infant ≥24 weeks’ gestation) per cycle protocol (NC-FET or AC-FET). Secondary outcomes included spontaneous abortion rate (defined as loss occurring after confirmation of an intrauterine gestational sac) and biochemical pregnancy rate (at least 9 days following FET, serum β-hCG ≥5 mIU/mL). Data were collected using medical records from software routinely used at our clinic (Meditex, Germany). The data were gathered by the authors using Meditex’s Report Editor, specifically designed for this study. After collection, the data were cleaned to ensure uniformity in reporting and interpretation before being transferred and analysed using SPSS version 25 (SPSS Inc., USA). A P -value of less than 0.05 was considered statistically significant. Numerical data were presented as means with standard deviations (SDs) or standard errors (SEs), while categorical data were expressed as frequencies. To compare the two study groups, unpaired t -tests or chi-squared tests were utilised. In addition, multiple pairwise comparisons were conducted using the Benjamini–Hochberg method to control for the false discovery rate. Univariate and multiple logistic regression analyses were employed to identify the potential impacts of patient characteristics on live birth following FET. The final model of significant predictors was constructed using the backward stepwise selection method with a P -value threshold of 0.1. Results were presented as odds ratios (ORs) with 95% confidence intervals (CIs).

Before performing multiple regression analysis, our data suggested that NC-FET was superior to AC-FET in achieving live births from frozen-thawed blastocysts. In addition, we observed a higher incidence of chemical pregnancies and spontaneous abortions in the AC-FET group. However, after adjusting for patient characteristics, the observed difference between NC-FET and AC-FET was no longer statistically significant. Similarly to our study, recent research by Wang et al. (2023) reported a higher birth rate with NC-FET and an increased rate of spontaneous abortions in the AC-FET group. Their study analysed data from 7,824 ovulatory women undergoing their first FET cycle with a single blastocyst. Although our sample size is smaller and includes real-world patients with varying ovulatory status, BMI, and AMH levels, our findings align with their conclusions. In our univariate regression analysis, NC-FET appeared to be associated with a higher birth rate compared to AC-FET. However, after adjusting for confounding factors in the multiple regression model, this association was no longer significant. Real-world clinical practice involves a diverse patient population with varying ages, BMIs, infertility diagnoses, and ET protocols, which interact in complex ways that may not be fully captured by statistical models. Since both our raw data and previous studies suggest a potential advantage of NC-FET, these findings should be considered when making clinical decisions. NC-FET may be particularly beneficial for young ovulatory women with a healthy BMI and cryopreserved blastocysts. This hypothesis is supported by our results and some previous studies ( Huang et al. 2023 , Wang et al. 2023 ), but not all ( Groenewoud et al. 2016 , Alur-Gupta et al. 2018 , Carosso et al. 2023 , Ho et al. 2024 ). However, the study by Ho et al. included only ovulatory patients and a mix of day 3 and day 5 ETs, which may make our two studies incomparable. Our study closely resembles the design of Alur-Gupta et al. which analysed 923 AC-FET blastocyst transfers and 103 NC-FET blastocyst transfers, but found no differences in positive HCG, live birth rates, biochemical pregnancies, or spontaneous abortions. However, Alur-Gupta et al. in contrast to our study, observed a statistical difference in the number of embryos transferred in the AC-FET group, with 37.5% of patients having a transfer of two embryos. Since double ET leads to a higher live birth rate per woman (odds ratio 1.94) ( Pandian et al. 2005 ), this might account for the similar success rates in their patients. In addition, while Alur-Gupta et al. analysed 103 NC-FET cycles, we analysed 175, so our larger sample size may have contributed to the difference in live birth, potentially improving the reliability of our findings. Huang et al. ( Huang et al. 2023 ) also found no significant difference in live birth rates between NC-FET and AC-FET. However, their data showed a trend towards higher birth rates in NC-FET (37.6 vs 30.1%, P = 0.119). Their lack of statistical significance may be attributed to a smaller sample size ( n = 384) and stringent inclusion criteria (requiring at least three frozen embryos and regular menstrual cycles), which resulted in both groups consisting of patients with favourable prognoses, making it difficult to detect differences. Another study by Mendes Godinho et al. (2024) reported a lower live birth rate and higher miscarriage rate in AC-FET compared to NC-FET, findings that are consistent with our results. Although only 18.2% of our patients underwent NC-FET, the groups were well matched in terms of age, partner’s age, and the number of embryos transferred. The primary differences were in BMI and AMH, which was expected considering that PCOS/anovulatory patients – who were almost exclusively included in the AC-FET group – tend to have higher BMI and AMH. While we could have designed the study to compare patients with similar BMI and AMH, we focused on real-world population comparisons. We also observed a significantly thinner endometrium in the NC-FET group (7.9 mm on average). This may be due to the potentially detrimental effects of supraphysiological oestradiol levels in AC-FET cycles, or our clinic’s inclusion criteria requiring an endometrial thickness of at least 8 mm for scheduling FET. However, this finding may not have clinical significance, as previous studies have reported that normal endometrial thickness in spontaneous cycles ranges from 7 to 14 mm ( Bakos et al. 1993 ), with no significant impact on pregnancy rates. Similarly, studies on oocyte donation cycles and FET cycles with endometrial thicknesses as low as 6 mm have found no significant differences in pregnancy or live birth rates ( Dain et al. 2013 , Arce et al. 2016 ). A thinner endometrium has been associated with lower birth weights in IVF/FET ( Zhang et al. 2023 ). Our data appear to support this, as birth weights in the NC-FET group were lower (3,095.3 g), which may be linked to the thinner endometrial lining. However, this difference was not clinically significant. AC-FET has been increasingly scrutinised due to reports of higher perinatal risks ( Maheshwari et al. 2016 ). These risks are consistently confirmed in the literature, but it is important to recognise that some patients cannot undergo NC-FET, making these risks unavoidable. In our cohort, birth weights were significantly higher in the AC-FET group, a finding consistent with previous studies ( Rosalik et al. 2021 ). However, this did not affect the mode of delivery or Apgar scores. Similarly, Wang et al. (2023) reported higher birth weights in AC-FET cycles in a cohort of over 7,000 patients, further supporting this association. Our NC-FET group consisted solely of true spontaneous cycles, without ovulation induction or hCG triggering. Although the initial analysis suggested a higher live birth rate in the NC-FET group despite the same number of blastocysts transferred, this difference was lost after adjusting for patient characteristics. In addition, we observed a significantly higher rate of biochemical pregnancies and spontaneous abortions in the AC-FET group ( P < 0.001). The pathophysiology of biochemical pregnancies remains unclear, but is associated with chromosomal abnormalities, parental age, lifestyle factors, and immunological mechanisms. Some studies suggest that elevated oestrogen levels in AC-FET cycles may impair vascular invasion, potentially contributing to poor implantation and increased biochemical pregnancy rates ( Winter 2002 , Wei et al. 2024 ). The most likely explanation for the higher pregnancy loss rate in our study, which was not significant in multivariate analysis, is the inclusion of both ovulatory and anovulatory patients, as anovulatory cycles (often linked to PCOS) are associated with higher pregnancy loss rates ( Matorras et al. 2023 ). Patients in the NC-FET group used home-based LH test strips for ovulation detection, which was convenient for both patients and our practice, as many patients travel long distances for appointments. Recent research ( Holder et al. 2023 ) concluded that different NC-FET approaches yield comparable outcomes while offering flexibility for both patients and physicians. Home-based monitoring reduces medical visits, lowers costs, and saves time, making it a practical alternative to hospital-based monitoring. Although clinicians must educate patients on potential challenges, such as cycle cancellations due to absent LH surges, inadequate endometrial lining, or scheduling conflicts, our data suggest that home-based monitoring did not negatively impact outcomes. Currently, there is no clear consensus on whether NC-FET or AC-FET is superior, and the optimal endometrial preparation protocol remains debated ( Ho et al. 2024 , Ghobara et al. 2017 ). Our study contributes valuable insights into this discussion, particularly in real-world clinical settings. Based on our results and other findings in the literature, it seems that NC-FET should be offered to all patients except anovulatory ones. As a medical community, we should find ways to reduce the risks potentially associated with AC-FET when they are unavoidable, such as in anovulatory patients. The strength of our study lies in its relatively large sample size, despite its retrospective design. A limitation is the inclusion of anovulatory patients in the analysis; however, their inclusion reflects a true real-life population for AC-FET. Further prospective studies with larger sample sizes are needed to build upon these findings.

Advancements in in vitro fertilisation (IVF) have significantly improved success rates, with the increasing use of frozen embryo transfer (FET) playing a key role ( Roque et al. 2019 ). FET not only increases cumulative live birth rates by allowing the transfer of embryos in a more physiologically optimal uterine environment but also reduces the incidence and severity of ovarian hyperstimulation syndrome (OHSS) and the need for repeated ovarian stimulation cycles ( Drakopoulos et al. 2019 ). In addition, FET has been associated with improved obstetric outcomes, including lower rates of multiple gestations, preterm birth, low birth weight, and small-for-gestational-age infants compared to fresh embryo transfer (ET) ( Maheshwari et al. 2016 ). However, emerging evidence suggests that FET may be linked to an increased risk of hypertensive disorders of pregnancy ( Moreno-Sepulveda et al. 2021 ), as well as larger-for-gestational-age and high birth weight infants ( Maheshwari et al. 2018 , Ginström Ernstad et al. 2019 ), highlighting the need for further research into the mechanisms underlying these associations. FET now constitutes over 30% of all assisted reproductive technology (ART) procedures in Europe ( Gliozheni et al. 2023 ). Endometrial preparation for FET can be achieved through several protocols, including natural cycles (NC-FET), modified natural cycles, and artificially programmed cycles (AC-FET), with or without the use of a gonadotropin-releasing hormone (GnRH) agonist. Despite the widespread use of these approaches, no consensus has been established regarding the optimal method. A Cochrane review concluded that available data are insufficient to recommend one approach over another ( Ghobara et al. 2017 ), and a recent randomised trial by Ho et al. found no significant differences in live birth rates between different FET preparation protocols ( Ho et al. 2024 ). Concerns have been raised about the absence of the corpus luteum and its hormonal contributions in AC-FET cycles, which may impact implantation and foetal development, potentially increasing the risk of hypertensive disorders ( Maheshwari et al. 2018 , Pereira et al. 2021 , Zaat et al. 2023 ). These findings suggest that NC-FET might offer physiological advantages, although further research is required to determine whether AC-FET indeed contributes to pregnancy complications. Despite these potential concerns, AC-FET remains the preferred approach in clinical practice due to its logistical advantages. A recent survey of ART clinics in the United States ( Lee et al. 2023 ) found that although most clinics offer NC-FET, it is used in only a minority of FET cycles. Reported barriers include the unpredictability of ovulation, the need for frequent monitoring, and increased scheduling difficulties. AC-FET, on the other hand, provides greater flexibility in transfer timing, allowing for adjustments of up to several days. This enables patients and clinics to schedule ETs at more convenient times, reducing logistical challenges. In addition, AC-FET minimises the risk of unexpected follicular development, leading to fewer cancelled cycles. Given the increasing reliance on FET in ART, determining the optimal endometrial preparation method remains a crucial clinical question. The present study aimed to compare live birth rates following NC-FET and AC-FET in a real-world, non-selected patient population at a single private fertility centre. Understanding the impact of these protocols on pregnancy outcomes may help refine clinical decision-making and optimise FET strategies to improve both success rates and maternal–foetal health.

The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the work reported.

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

DR was responsible for study design, data analysis and interpretation, and writing the manuscript. MB contributed to study design, critical discussion, and writing the manuscript. KPP participated in data acquisition, analysis, and drafting of the manuscript. SV participated in data acquisition, analysis, and drafting of the manuscript. All authors approved the final version for publication and is accountable for the accuracy and integrity of the work.

Couples gave written informed consent for their anonymised medical data to be used for research purposes.

The study was approved by the Institutional Review Board (‘Etičko povjerenstvo’).