Reproductive prognosis factors in women with infertility and polycystic ovary syndrome undergoing in vitro fertilization techniques.

Polycystic ovary syndrome (PCOS) is a significant cause of oligomenorrhea and hyperandrogenism in women and can be readily diagnosed when classic signs and symptoms, such as hirsutism and polycystic ovarian morphology, are present ( Joham et al ., 2022 ). PCOS is the most prevalent endocrinopathy, affecting approximately 5 to 18% of women of reproductive age ( Joham et al ., 2022 ). Its prevalence is often linked to various metabolic disorders, including infertility, obesity, diabetes mellitus, and an increased risk of long-term pregnancy complications and cardiovascular diseases ( Solomon et al ., 2002 ; Moran et al ., 2010 ; Wild et al ., 2011 ; Diamanti-Kandarakis & Dunaif, 2012 ; Christ et al ., 2019 ). There is considerable heterogeneity in the phenotypic manifestations of this syndrome, and the consequences can vary significantly throughout a woman’s life ( Clark et al ., 2014 ; van Keizerswaard et al ., 2022 ). In PCOS, obesity is a common characteristic, estimated to affect approximately half of the women with this condition, with the increase in fat mass often coinciding with the onset of PCOS. However, obesity is not part of the diagnostic criteria, as about 20% of women with PCOS are not obese. Hormonal studies in women with PCOS have revealed elevated levels of testosterone and estrogen, demonstrating an increased ratio of luteinizing hormone (LH) to follicle-stimulating hormone (FSH), as well as higher concentrations of estrone compared to estradiol, and androstenedione levels that are at the upper limits of normal or elevated ( Casanova et al ., 2019 ). Currently, the most widely accepted diagnostic criteria for PCOS are the Rotterdam criteria, which require the presence of at least two of the following three parameters: menstrual cycle irregularities, clinical or biochemical hyperandrogenism, and polycystic ovarian morphology ( Christ & Cedars, 2023 ). This diagnostic method is subject to controversy in the medical community because some classical characteristics are not always present ( Clark et al ., 2014 ). In 2012, the National Institute of Health recommended the use of this classification, identifying specific subtypes for PCOS ( NIH, 2012 ). Several laboratories, clinical, and physical parameters can be evaluated in PCOS, including total and free testosterone hormone levels, dehydroepiandrosterone sulfate, androstenedione, 17-hydroxyprogesterone ( Longcope, 1986 ; Burger, 2002 ) and anti-Müllerian hormone (AMH), which are not included in the Rotterdam criteria. Due to the heterogeneity associated with ultrasound evaluation in PCOS, there has been growing interest in using AMH as a possible surrogate marker for ovarian morphology assessment ( Dumont et al ., 2015 ), because it is typically elevated in women with PCOS compared to normo-ovulatory women ( Cook et al ., 2002 ). According to international guidelines for the assessment and treatment of PCOS, anovulatory infertility can be effectively managed with first-line therapies, including ovulation induction using either monotherapy or a combination of medications. For those who do not respond to first-line treatments, second-line therapy may involve ovarian stimulation in conjunction with assisted reproductive techniques, such as in vitro fertilization (IVF) ( Teede et al ., 2018 ). Despite the most of women with PCOS responding favorably to ovarian stimulation for IVF, with a higher than average number of growing follicles, their reproductive outcomes tend to be unfavorable compared to women with infertility due to other causes ( Rajani et al ., 2012 ; Piomboni et al ., 2014 ; Yilmaz et al ., 2016 ; Artimani et al ., 2018 ). Factors related to oocyte quality ( Qiao & Feng, 2011 ) and endometrial competence ( Piltonen, 2016 ) may negatively impact reproductive outcomes, suggesting that this condition could deteriorate the follicular microenvironment and compromise the physiological functions of oocytes ( Palomba et al ., 2017 ). AMH is a polypeptide secreted by the granulosa cells of preantral and small antral ovarian follicles ( van Rooij et al ., 2002 ), and it is widely recognized as a sensitive marker of ovarian reserve ( La Marca et al ., 2010 ; Tal et al ., 2014 ). This hormone plays a crucial role in inhibiting the activity of primary follicles and regulating their growth and development ( Chang et al ., 2019 ). AMH levels are significantly elevated in women with PCOS, and numerous studies have sought to establish a correlation between AMH levels and the morphological characteristics of polycystic ovaries ( Swanson et al ., 1981 ; Reed & Carr, 2000 ; Slayden et al ., 2001 ; DeUgarte et al ., 2006 ; Escobar-Morreale et al ., 2012 ; Legro et al ., 2013 ; Prior et al ., 2015 ; Bull et al ., 2019 ; Carmina et al ., 2019 ; Starace et al ., 2020 ). Additionally, AMH has been considered a predictor of ovarian response to controlled ovarian stimulation ( Delamuta et al ., 2024 ). It is known that both AMH levels and the number of ovarian follicles decline as a woman age, occurring in both women with and without PCOS ( Erdem et al ., 2002 ; Alsamarai et al ., 2009 ; de Kat et al ., 2016 ). This observation has led some researchers to suggest the necessity of age-adjusted diagnostic criteria. While specific age-related thresholds have yet to be widely accepted by the scientific community, recognizing these changes over time is important, particularly when assessing older women with PCOS ( Christ & Cedars, 2023 ). Although exploring this relationship holds promise, considerable uncertainty remains regarding the methodologies used in studies and the ranges of AMH levels that have been investigated. Due to these limitations, AMH is not included in current guidelines as a substitute for imaging evaluations of polycystic ovaries or as a standalone test for diagnosing PCOS ( Teede et al ., 2018 ). Originally, it was believed that the elevated levels of AMH in individuals with PCOS resulted from the presence of polycystic ovaries, which contain approximately twice as many preantral and antral follicles. However, recent insights indicate that increased AMH levels are primarily due to enhanced production by individual follicles ( Das et al ., 2008 ) and may also correlate with the severity of PCOS ( La Marca et al. , 2004a ; Nardo et al. , 2009 ; Piouka et al. , 2009 ; Lin et al. , 2011 ; Delamuta et al ., 2024 ). Multiple recent studies have demonstrated that AMH possesses significant predictive validity in diagnosing PCOS, particularly concerning the severity of ovulatory dysfunction ( La Marca et al ., 2010 ). Therefore, AMH could serve as an important indicator in assessing the reproductive prognosis of women with PCOS who are undergoing assisted reproductive technologies (ART) ( Vale-Fernandes et al ., 2023 ). While still a matter of debate, the latest evidence suggests that AMH levels may also be associated with reproductive outcomes in ART, including pregnancy rates and live birth rates (LBR), independent of age ( Eldar-Geva et al ., 2005 ; Nelson et al ., 2007 ; Elgindy et al ., 2008 ; Gleicher et al ., 2010 ; Blazar et al ., 2011 ; La Marca et al ., 2011 ; Brodin et al ., 2013 ), as well as with a higher yield of mature oocytes and embryos ( Fallat et al ., 1997 ; Hazout et al ., 2004 ; La Marca et al ., 2004b ; Eldar-Gevat et al ., 2005; Lekamge et al ., 2007 ). A recent study also suggested that AMH levels could guide the selection of ovarian stimulation protocols, with the aim of improving treatment outcomes for women of advanced reproductive age ( Meczekalski et al ., 2016a ; 2016b ). However, conflicting data exist regarding the effectiveness of AMH as a prognostic marker in IVF techniques, leading to its status as a tool not universally accepted ( Kaya et al ., 2010 ; Tal et al ., 2014 ; Liu et al ., 2022 ). Anovulatory infertility is often the first sign of PCOS ( Cunha & Póvoa, 2021 ). When first-line treatments prove ineffective, these women may pursue second-line infertility therapies, such as in IVF or intracytoplasmic sperm injection (ICSI) ( Costello et al ., 2019 ). In recent years, some reproductive prognostic factors have been identified as predictors of success after ART ( Gao et al ., 2021 ; Wang et al ., 2023 ). These factors include age, body mass index (BMI), baseline concentrations of AMH and follicle-stimulating hormone (FSH) at the start of the menstrual cycle, and follicle count ( Tsakos et al ., 2014 ). This study was designed to clarify the prognostic value of AMH levels, age, BMI, and the LH:FSH ratio in women with PCOS undergoing IVF/ICSI techniques.

The sample of women with PCOS in this study has a mean age of 34.78 years (±3.55), a mean BMI of 24.76 kg/m 2 (±5.12), and a mean AMH level of 3.38 ng/mL (±2.77). At the beginning of the menstrual cycle, the women in the sample had a serum LH level of 7.20mIU/mL (.81), FSH level of 7.67mIU/mL (±5.51), basal estradiol level of 49.33mIU/mL (±41.78), and the LH:FSH ratio was 1.02 (±0.66). According to the technique, 52.2% underwent ICSI and 47.8% underwent IVF. Regarding smoking, 21.2% of women reported this habit, while 40.3% of male partners too. Only 9.9% of women had other causes of infertility besides PCOS, and 66.4% had concomitant male infertility. Additional results can be found in Table 1 . The IVF/ICSI cycles (n=494) were categorized into groups based on AMH levels (n=346) ( Table 2 ). Women with PCOS and low AMH levels tended to be older ( p <0.001). When comparing women with high and low AMH levels, those in the high AMH group exhibited significantly lower FSH levels (6.50 vs . 10.37/mL; p <0.001), a higher LH:FSH ratio (1.37 vs . 0.68; p <0.001), and higher LH levels (8.35 vs . 6.92mIU/mL), although the difference in LH levels was not statistically significant. No significant differences in BMI were observed between the groups (25.33 vs . 24.25 vs . 25.68 kg/m 2 ). However, when examining the subgroup of women with AMH levels above the 95th percentile (n=17), it was found that they tended to have a higher weight (73.59 kg; p =0.020) and a higher BMI (28.09 kg/m 2 ; p =0.009) compared to those below the 95th percentile. Additionally, this group of women with very high AMH levels exhibited a continued trend of an elevated LH:FSH ratio ( p <0.001). The IVF cycle characteristics and outcomes are presented in Table 3 . The gonadotropin dose used in stimulation was lower in the high AMH group compared to the low AMH group (2082.06 IU vs . 2658.91 IU; p <0.001). The number of follicle count (5.46 vs . 13.29 vs . 21.58; p <0.001 and p <0.001), the number of COC (6.35 vs . 11.22 vs . 13.91; p <0.001; p =0.015), and embryos with 2PN (3.12 vs . 5.04 vs . 6.33; p <0.001; p =0.033) were higher in the high AMH group compared to the low and moderate AMH groups. Women with low AMH levels obtained fewer embryos compared to the high AMH group (1.73 vs . 2.73; p <0.001), with no statistically significant differences between the moderate and high AMH groups. Across AMH percentiles, there is a statistically significant trend of increasing number of freeze all IVF cycles between groups (23.9% vs . 47.7% vs . 61.6%; p <0.001). It is also noted that the clinical pregnancy rate and cumulative pregnancy rate increase across AMH percentiles; however, these differences between groups are not statistically significant. Regarding biochemical pregnancy, this factor tends to be higher with higher AMH levels, with statistical significance between groups (35.2% vs . 52.3% vs . 53.4%; p =0.004). As for LBR, there are significant differences between AMH groups, with an increase between the low and moderate AMH group (21.7% vs . 33.8%), but a slight decrease between the moderate and high AMH group (33.8% vs . 32.4%; p =0.015). The CLBR increased across AMH percentiles (22.4% vs . 39.3% vs . 41.4%; p =0.021), with statistically significant differences between groups. There were no statistically significant results for the number of stimulation days, nor for fertilization rates, embryo cleavage rates, blastocyst rates, and endometrial thickness on the day of final monitoring and on the day of oocyte pick-up. The remaining results are shown in Tables 2 and 3 . Conducting a logistic regression analysis for age as a potential confounding factor in the AMH groups, it is observed that, considering the same reproductive prognostic variables, with the sample divided into two groups of women below and above 35 years old, respectively, differences exist ( Tables 4 and 5 ). In the group of women below 35 years old, the LH:FSH ratio was significantly higher between the high AMH group and the moderate AMH group (1.55 vs . 1.12; p <0.001). The dose of gonadotropins used in stimulation became statistically significant, with higher dose in the low AMH group ( p <0.001). And the number of embryos obtained continued to trend upwards across groups but ceased to be statistically significant in this age group. The number of cryopreserved embryos was significantly higher in the high AMH group compared to the moderate AMH group and no longer showed statistically significant differences between the low and high AMH groups. Women below 35 years old in the high AMH group had higher progesterone levels at trigger (1.38 vs . 0.94 ng/mL; p =0.044) and a higher number of freeze-all IVF cycles (72.5% vs . 44.8%; p =0.002) compared to the moderate AMH group. The LBR and CLBR increased with AMH levels in younger women; however, without statistical significance between groups. In the analysis of women over 35 years old, statistically significant differences were observed among all groups of AMH levels. Within this age range, women with elevated AMH levels achieved a higher number of embryos ( p <0.001 and p =0.047), a greater number of cryopreserved embryos ( p <0.001 and p =0.031), and a higher biochemical pregnancy rate compared to the other groups. In this age group, the LBR and CLBR decreased from the moderate AMH group to the high AMH group, with LBR dropping from 27.1% to 17.1% ( p =0.039) and CLBR decreasing from 29.4% to 17.1% ( p =0.038). Some results from the 17 women who had AMH levels above the 95th percentile changed after detailed analysis by age. Besides observing a greater number of embryos obtained in younger women (4.00 vs . 2.00), this number was also higher compared to women with AMH levels below the 95th percentile ( p =0.004) below 35 years old. This result is not observed in the group of women over 35 years old. Also, the clinical pregnancy rate, cumulative clinical pregnancy rate, LBR, and CLBR were markedly lower in women above 35 years old compared to younger ones. Specifically, in the group of women with high (above the 75th percentile) and very high (above the 95th percentile) AMH levels, the LBR was 45.1% and 40.0% below 35 years old and 17.1% and 14.3% above 35 years old, respectively. The statistical analysis of the remaining variables did not change after logistic regression. The remaining results are in Tables 4 e 5 . The IVF/ICSI cycles (n=494) were divided into groups according to BMI classes (n=493). It was found that the duration of infertility tended to be higher in groups of women with higher BMI, with statistical significance observed when comparing normal weight ( p <0.001) and overweight classes ( p =0.018) with obese women ( Table 6 ). The IVF cycle characteristics and outcomes are presented in Table 6 . Women with higher BMI tended to require higher doses of gonadotropins compared to underweight and normal weight women ( p =0.025), overweight women ( p =0.004), and obese women ( p <0.001), as well as between normal weight and obese women ( p =0.041). Across BMI groups, there was a trend towards a greater number of stimulation days, with statistical significance between normal weight and overweight classes ( p =0.003) and normal weight and obese classes ( p <0.001). There was a significant increase in the number of COC ( p =0.047), but a decrease in fertilization rate ( p =0.020) in overweight women compared to underweight women. The remaining prognostic factors were not associated with significant differences between groups. Regarding LH levels, FSH levels, LH:FSH ratio, basal estradiol, and progesterone at trigger, no statistically significant differences were found between BMI classes. There is a decrease in the clinical pregnancy rate, cumulative pregnancy rate, LBR, and CLBR in the underweight and obese groups compared to the other BMI classes, with higher values observed in normal weight and overweight women. However, these findings are not statistically significant. The remaining results are in Tables 6 and 7 Splitting the sample into two groups below and above 35 years old (n=238 and n=256, respectively), it is observed that there are more women with primary infertility causes in the group below 35 years old (86.6% vs . 71.0%; p <0.001), the duration of this infertility is shorter (48.05 months vs . 54.04 months; p =0.041), and the LH:FSH ratio is lower in women above 35 years old (0.90 vs . 1.15; p <0.001) ( Table 8 ). The IVF cycle characteristics and outcomes are presented in Table 9 . Regarding intra-cycle variables, younger women underwent more ICSI (59.2% vs . 45.7%; p =0.003) than FIV. The dose of gonadotropins used in stimulation was higher in older women (2472.63 IU vs . 2148.43 IU; p <0.001), while the number of COC (9.49 vs . 11.93; p <0.001), follicle count (11.40 vs . 16.30; p <0.001), number of embryos with 2PN (4.59 vs . 5.57; p =0.007), blastocyst rate (5.34% vs . 8.70%; p =0.032), number of freeze-all cycles (36.9% vs . 48.3%; p =0.010), biochemical pregnancy rate (41.8% vs . 58.0%; p <0.001), clinical pregnancy rate (29.9% vs . 47.4%; p <0.001), cumulative pregnancy rate (35.5% vs . 52.9%; p <0.001), LBR (23.7% vs . 42.5%; p <0.001), CLBR (28.4% vs . 47.1%; p <0.001), and endometrial thickness on oocyte pick-up day (10.06 mm vs . 11.02 mm; p =0.006) were lower in this group compared to women below 35 years old. The remaining results are in Tables 8 and 9 . Women with an LH:FSH ratio greater than 1 are significantly younger compared to women with a ratio lower than 1 (35.36 vs . 33.84 years; p <0.001). There is also a shorter duration of infertility in women with an LH:FSH ratio greater than 1 (54.2 vs . 46.0 months; p <0.008) ( Table 10 ). The IVF cycle characteristics and outcomes are presented in Table 11 . Regarding intra-cycle variables, in women with an LH:FSH ratio greater than 1, there is a reduction in the dose of gonadotropins used in stimulation (1921.25 IU vs . 2527.59 IU; p <0.001), a higher number of follicle count (18.59 vs . 11.36; p <0.001), a higher number of COC (13.06 vs . 9.45; p <0.001), a higher number of embryos with 2PN (6.11 vs . 4.52; p <0.001), a higher number of embryos obtained (2.87 vs . 2.34; p =0.006), a higher number of embryos cryopreserved (1.82 vs . 1.37; p =0.046), increased LBR (38.3% vs . 29.2%; p =0.007) and CLBR (45.2% vs . 32.5%; p =0.010), a higher number of freeze-all cycles (50.0% vs . 38.8%; p =0.017), and cases of biochemical pregnancy (59.6% vs . 43.6%; p =0.004). The remaining results are in Tables 10 and 11 . Statistically significant positive correlations were found between AMH and LH serum levels at the beginning of the menstrual cycle ( p <0.001), the LH:FSH ratio ( p <0.001), the number of COC ( p <0.001) and follicle count ( p <0.001), LBR ( p =0.035), and CLBR ( p =0.008). Negative correlations were observed between AMH serum levels and age ( p <0.001), FSH levels ( p <0.001), and the dose of gonadotropins used in stimulation ( p <0.001). No statistically significant differences were found between AMH and BMI, basal estradiol, clinical pregnancy rate, and cumulative clinical pregnancy rate. Regarding age, positive correlations were found with FSH levels at the beginning of the menstrual cycle ( p =0.006) and the dose of gonadotropins used in stimulation ( p <0.001), and negative correlations were found with LH levels at the beginning of the menstrual cycle ( p <0.001), the LH:FSH ratio ( p <0.001), the number of COC ( p <0.001) and follicle count ( p <0.001), clinical pregnancy rate ( p <0.001), cumulative clinical pregnancy rate ( p <0.001), LBR ( p <0.001), and CLBR ( p <0.001). Regarding BMI, positive correlations were found between the dose of gonadotropins used in stimulation ( p =0.003) and the number of follicle count ( p =0.010), and negative correlations were found between LH and FSH levels at the beginning of the menstrual cycle ( p <0.001 and p <0.001, respectively) and basal estradiol ( p =0.003). The remaining variables showed no statistically significant correlations with BMI. Positive correlations were observed between the LH:FSH ratio and LH levels at the beginning of the menstrual cycle ( p <0.001), the number of COC ( p <0.001) and follicle count ( p <0.001), clinical pregnancy rate ( p =0.030), cumulative clinical pregnancy rate ( p =0.014), LBR ( p =0.012), and CLBR ( p =0.003). Negative correlations were found between the LH:FSH ratio and the dose of gonadotropins used in stimulation ( p <0.001) and FSH levels at the beginning of the menstrual cycle ( p <0.001). The remaining correlation coefficients are presented in Table 12 .

This study is an original article based on a retrospective observational analysis involving a sample of women with PCOS (n=393) who underwent IVF/ICSI using autologous oocytes at the Centro Materno-Infantil do Norte (CMIN), Porto, Portugal. The participants were monitored from the initial study phase through to the completion of reproductive techniques, pregnancy, and childbirth. Data were collected from each oocyte pick-up (n=494) between January 2017 and December 2022, with follow-up on pregnancies and births conducted from the date of the initial embryo transfer until December 2023. The study utilized the current Rotterdam diagnostic criteria for participant inclusion ( Christ & Cedars, 2023 ). Cases presumed to be PCOS but lacking the requisite signs and symptoms outlined in the Rotterdam criteria, as well as those exhibiting clinical or biochemical hyperandrogenism due to other causes, were excluded. Additionally, cases involving endometriosis, thyroid disease, and other endocrine disorders were also excluded. Electronic clinical records were reviewed to extract the necessary data for the research. The study protocol received approval from the Ethics Committee of the Unidade Local de Saúde de Santo António (ULSSA) on February 1, 2024, under the Reference Number: 2023.262(224-DEFI/214-CE). The sample was divided according to serum levels of AMH: below the 25th percentile (AMH 4.67 ng/mL; n=86). The terms “low”, “moderate”, “high”, and “very high” correspond to the results relative to the groups below the 25th percentile, between the 25th and 75th percentiles, above the 75th percentile, and above the 95th percentile, respectively, and are mentioned to simplify the analysis and discussion of results, not necessarily indicating reference values or biological normality. A more detailed analysis was performed in the subgroup of women with AMH levels above the 95th percentile (AMH >8.67ng/mL; n=17). The sample was divided categorizing the sample into two groups according to the age of the female subject: below 35 years (n=238) and above 35 years (n=256). The sample was divided categorizing the sample into BMI classes: underweight (BMI <18.5 kg/m 2 ; n=23), normal weight (BMI between 18.5 and 24.9 kg/m 2 ; n=285), overweight (BMI between 25.0 and 29.9 kg/m 2 ; n=106), and obesity (BMI ≥30.0 kg/m 2 ; n=79). The sample was divided categorizing the sample into two categories of the LH:FSH ratio, below (n=305) and above 1 (n=178). The women underwent an antagonist protocol of gonadotropin-releasing hormone (GnRH) with controlled and individualized ovarian stimulation, based on standard tests of ovarian reserve. On the third day after oocyte pick up, one or two embryos were selected for transfer, while the remaining embryos were cryopreserved or allowed to develop until the blastocyst stage for fresh embryo transfer or to be frozen for subsequent frozen embryo transfer (FET). The decision to freeze all embryos was made based on local criteria, which included factors such as the occurrence or risk of ovarian hyperstimulation syndrome, inadequate endometrial environment, premature elevation of progesterone, COVID-19 infection, hydrosalpinx, or personal circumstances favoring FET rather than fresh embryo transfer. The primary outcomes included LBR and the cumulative live birth rate (CLBR). Live birth was defined as the birth of one or more live infants. LBR was defined as the number of live births divided by the number of women in a group. CLBR was defined as the rate of live births that occurred during the fresh IVF cycle and subsequent FET cycle(s) after a single oocyte pick-up until achieving a live birth. Clinical pregnancy was defined as the visualization of one or more gestational sacs on ultrasound. Pregnancies without visualization of a gestational sac and/or without visualization of a gestational sac with a heartbeat were considered miscarriages and did not contribute to clinical pregnancy rate. The cumulative pregnancy rate was defined as the proportion of women who became pregnant from a single oocyte pick-up. The secondary outcomes included the rate of oocyte immaturity, defined as the proportion of immature oocytes per number of cumulus-oocyte complexes (COC); fertilization rate, defined as the proportion of embryos with two pronucleated oocytes (2PN) per number of COC; cleavage rate, defined as the proportion of the number of 2PN embryos by the number of 2PN; blastocyst rate, defined as the proportion of the number of blastocysts by the number of 2PN; the percentage of freeze-all cycles and the number of embryos obtained and cryopreserved. The primary and secondary statistical analysis involved comparing group characteristics, FIV/ICSI cycles, and obstetric outcomes. Categorical variables were presented as number of cases and percentages, and continuous variables as mean, standard deviation, minimum, and maximum. For categorical variables in group comparisons, the Chi-square test was used, and for continuous variables, the One-Way ANOVA test was used. The degree of association of AMH levels, age, BMI, and LH:FSH ratio with reproductive prognostic factors was calculated using the Spearman correlation coefficient. Logistic regression was performed to verify age as a confounder in the analysis by percentiles of AMH groups. Data from this study were analyzed using IBM SPSS Statistics 29.0.0.0 Software. A statistical significance level of 0.05 (5%) was considered.

The purpose of this research was to explore some basic reproductive characteristics, hormonal levels, and intra-cycle variables of IVF/ICSI associated to reproductive success, particularly LBR and CLBR, and consider their role in investigating reproductive prognosis in a sample of women with PCOS. Through comprehensive statistical analysis, the impact of these factors on reproductive outcomes was analyzed. A weak negative correlation was found between AMH and age, supporting some descriptions in the literature indicating that this hormone tends to decrease with age ( Erdem et al ., 2002 ; Alsamarai et al ., 2009 ; de Kat et al ., 2016 ; Liu et al ., 2022 ), although this correlation was not found in other studies ( Tal et al ., 2014 ; Vale-Fernandes et al ., 2023 ). Regarding LH levels at the beginning of the menstrual cycle, they increase with AMH levels, as in the study by Liu et al. (2022) . As for FSH levels, they were significantly lower in moderate AMH group, when comparing with the low AMH group, as described in other studies on the subject that have published data with samples of women with AMH levels close to those in this study ( Liu et al ., 2022 ; Wang et al ., 2023 ). In this work, no differences were found between AMH groups and BMI, which may be explained by the fact that about 20% of women with PCOS are not obese ( Casanova et al ., 2019 ), supporting results already described ( Sahmay et al ., 2011 ; Liu et al ., 2022 ; Vale-Fernandes et al ., 2023 ). However, in women with very high AMH levels, above the 95th percentile (n=17), AMH was positively related to weight, the LH:FSH ratio, and BMI, consistent with some recent evidence ( Lin et al ., 2011 ). This conclusion may reinforce the well-established association between PCOS and conditions such as overweight, obesity, and increased fat mass, which affect approximately half of the women diagnosed with this endocrine disorder. This clinical relationship can be understood through the mechanisms that lead to dysregulation of the gonadotropin axis, particularly concerning hormones such as LH, androstenedione, estrone, testosterone, insulin, and appetite regulation ( Casanova et al ., 2019 ). The LH:FSH ratio significantly increased with AMH levels, supporting similar results in several studies ( Casanova et al ., 2019 ; Liu et al ., 2022 ). Additionally, the relationship between elevated AMH levels and higher progesterone levels at trigger can be discussed, which are associated with a higher number of IVF cycles with freeze all observed in this study. In fact, premature elevation of progesterone has implications for the decision of whether to perform embryo transfer or to freeze all embryos for subsequent FET, using the freeze-all method, as described in the literature ( Healy et al ., 2016 ; Wang et al ., 2017 ). Premature elevation of progesterone appears to have a negative impact on implantation rates and pregnancy rates due to desynchronization with the endometrium ( Aghahosseini et al ., 2017 ). Age-related variations show that women above 35 with high serum levels of AMH exhibit lower LBR and CLBR compared to younger women in the same percentile range of this hormone. This finding leads to the conclusion that AMH levels may contribute to LBR and CLBR, but age is also an important factor to consider. Specifically, in groups of women with elevated AMH (above the 75th percentile) and very high AMH (above the 95th percentile), the LBR was 45.1% and 40.0% below 35 years and 17.1% and 14.3% above 35 years, respectively. This result revealed that excessively high levels of AMH, as expected in a group of women with PCOS ( Teede et al ., 2018 ; Costello et al ., 2019 ; Christ & Cedars, 2023 ), may contribute to a worse reproductive prognosis, as some studies have indicated ( Lukaszuk et al ., 2014 ; Vale-Fernandes et al ., 2023 ), but age is an important factor to be considered. According to authors Blazar et al., AMH level is related to reproductive outcome in the context of second-line infertility treatments, particularly in terms of pregnancy rate and LBR, regardless of age ( Gleicher et al ., 2010 ; Blazar et al ., 2011 ), strengthening that AMH levels may carry greater significance than age. In women above 35, the biochemical pregnancy rate and cumulative pregnancy rate increased from low to moderate AMH level, but increases were no statistically significant. Also in women above 35, LBR and CLBR increased from low to moderate AMH levels but decreased by 10% ( p =0.039) and 12.3% ( p =0.038), respectively, from moderate to high AMH group, results that are opposite to some research on the subject ( Tal et al ., 2020 ; Liu et al ., 2022 ; Vale-Fernandes et al ., 2023 ), but consistent with other studies ( te Velde & Pearson, 2002 ; Gnoth et al ., 2003 ; van Rooij et al ., 2005 ; Nelson et al ., 2007 ; Gleicher et al ., 2010 ; Blazar et al ., 2011 ; La Marca et al ., 2011 ; Brodin et al ., 2013 ; Acharya et al ., 2022 ; Wang et al ., 2023 ; Delamuta et al ., 2024 ). Acharya et al. conclude that women with AMH levels above 5 ng/mL (similar to the group of women in this study with AMH levels above 4.66 ng/mL), undergoing IVF with autologous fresh embryo transfer, experience a 3% decrease in the likelihood of achieving a live birth for each unit increase in AMH ( Acharya et al ., 2022 ). Discrepancies may be attributed to differences in the populations of various studies and sample sizes, but also to inaccuracies in the definitions of LBR and CLBR. LBR is defined in most studies as the number of live births per woman in a group, however, there are articles that assume the number of births equals the number of live births ( Olivius et al ., 2002 ). Regarding CLBR, which is already a complex concept but suggested as an appropriate way to report the success of an IVF cycle encompassing fresh embryo transfer and subsequent FET ( Germond et al ., 2004 ), the definitions diverge even further. Defining the numerator and denominator of CLBR is a challenge and it is possible that this concept considers varying data across articles ( Maheshwari et al ., 2015 ). Some studies used the attainment of the first live birth as the numerator ( Thurin-Kjellberg et al ., 2009 ; Luke et al ., 2012 ; Stern et al ., 2013 ; Bodri et al ., 2014 ), while others included all live births in a cycle ( Li et al ., 2014 ). As for the denominator, it can be the number of women who underwent ovarian stimulation or all those who underwent oocyte pick-up. In addition to the above, there is no established consensus on the interpretation of results in cases where women undergo a repeated IVF cycle after achieving a live birth, as some authors simply exclude subsequent cycles ( Bodri et al ., 2014 ) and others consider them as new cases ( Gnoth et al ., 2011 ). There are articles that define CLBR as the number of live births that occur in a group of women after embryo transfer and subsequent FET until all cryopreserved embryos are finished ( Daya, 2005 ), as in this study, but others consider the number of live births that occur in a group of women after embryo transfer and subsequent FET until a predetermined number of IVF cycles ( Li et al ., 2014 ). These different methods of calculating CLBR, based on different assumptions, cannot be directly compared and discussed in this and other studies. Although PCOS is associated with oligo-ovulatory cycles, these women typically respond favorably to ovarian stimulation for subsequent IVF procedures, achieving a higher number of follicles compared to the average of women undergoing the same treatments. However, they are associated with an unfavorable reproductive outcome compared to women with infertility due to other causes ( Rajani et al ., 2012 ; Piomboni et al ., 2014 ; Yilmaz et al ., 2016 ; Artimani et al ., 2018 ). It is hypothesized that PCOS may be detrimental to the follicular microenvironment, disadvantaging optimal conditions for normal oocyte physiological functions ( Palomba et al ., 2017 ), their quality ( Qiao & Feng, 2011 ), and endometrial competence ( Piltonen, 2016 ). Regarding endometrial data in this study, it was found that endometrial thickness on the day of the last monitoring and on the day of oocyte pick-up did not show significant differences between groups. It is known that endometrial thickness influences implantation rates, although the exact appropriate thresholds are still not consensus. A study conducted in 2018 concluded that increasing endometrial thickness from 9mm to 16mm has a slight effect on increasing the probability of pregnancy, demonstrating this link ( Bashiri et al ., 2018 ; Liu et al ., 2022 ), although it is not in line with the results of this research. The dose of gonadotropins used in stimulation showed a tendency to decrease with increasing AMH levels, consistent with other studies suggesting that AMH may be useful in predicting ovarian response to gonadotropins, a mechanism possibly explained by decreased follicular sensitivity to FSH, premature follicle growth, and decreased estradiol ( Gnoth et al ., 2008 ). As expected, the number of antral follicles correlates positively (moderate correlation coefficient) with AMH levels, consistent with the results described in the study by Li et al. (2018) . This correlation supports the assertion that AMH production by antral follicles is increased in PCOS, where the number of antral follicles is approximately doubled ( Gnoth et al ., 2008 ; Kaya et al ., 2010 ). The number of embryos obtained decreased with AMH, contrary to described evidence ( Delamuta et al ., 2024 ). However, the oocyte immaturity rate, the fertilization rate and cleavage rate did not show statistically significant differences between groups, results consistent with other studies conducted in women with PCOS ( Vale-Fernandes et al ., 2023 ). In this study, no significant differences were found between BMI classes regarding the number of antral follicles, which is also consistent with a study assessing the impact of BMI on reproductive success in women undergoing IVF techniques ( Chaves et al ., 2023 ). However, in this 2023 study, obese women obtained more mature oocytes compared to women with normal weight, a finding not observed in the present sample, where no significant differences were found between BMI classes regarding the rate of oocyte immaturity. There was a significant increase in fertilization rate in overweight women compared to those in the first BMI class studied (65.1% vs . 46.6%), which is also consistent with the results of other studies ( Shah et al ., 2011 ; Chaves et al ., 2023 ). These data may be explained considering the failure of oocyte cytoplasm in transporting condensed chromosomes or in the incorporation of sperm into oocytes ( Shah et al ., 2011 ; Machtinger et al ., 2012 ). It is noted that the clinical pregnancy rate, cumulative clinical pregnancy rate, LBR and CLBR were lower in obese women compared to other BMI classes; however, this difference did not reach statistical significance as expected and as described in several studies ( Shah et al ., 2011 ; Bellver et al ., 2013 ; Goldman et al ., 2019 ; Moreira et al ., 2025 ). The LBR was lower in women with obesity, although it was not statistically significant, but this trend is also found in a study that suggests that the impact of BMI is more significant in younger ages and that this LBR is more influenced by the prevalence of aneuploidies than by BMI ( Goldman et al ., 2019 ). The decrease in clinical pregnancy rate in that study was observed in cases of higher degrees of obesity ( Shah et al ., 2011 ), while in the present study, an analysis was only conducted with BMI above 30 kg/m 2 , representing a limitation. A study comparing IVF outcomes between women below and above 35 years reported that, for younger women, there was no difference between BMI categories in the number of mature oocytes, number of embryos obtained, fertilization rates, and clinical pregnancy rates ( Shah et al ., 2011 ; Vural et al ., 2016 ). Except for the clinical pregnancy rate, which was significantly lower in women above 35 years old in this sample, all other factors were consistent with the results of this cited study. The authors Vural et al., describe that in the group of women above 35 years old, the clinical pregnancy rate was lower in the obesity group, a result that is in line with this research, where obese women had a clinical pregnancy rate of 31.4% compared to clinical pregnancy rates in the normal weight and overweight classes of 40.3% and 41.4%, respectively ( Vural et al ., 2016 ). However, the number of mature oocytes, embryos obtained, and fertilization rate did not differ between BMI categories. Thus, Vural et al. suggest that different stages of IVF may be affected by the age of the woman undergoing treatment ( Vural et al ., 2016 ). The embryo cleavage rate did not show statistically significant differences between BMI classes, but there is a study reporting that embryos from women with obesity experience a delay in cleavage compared to normal-weight women, suggesting that there is slower embryonic development associated with higher maternal BMI ( Bartolacci et al ., 2019 ). The hypersecretion of LH appears to impact fertility, as well as reproductive outcomes in women with PCOS. A high LH:FSH ratio is associated with poorer outcomes in terms of pregnancy and LBR, and in this study, a reduction in LBR was observed when this ratio was above 1. This finding is consistent with a study that evaluated the effect of this ratio in a sample of women with PCOS and found no statistically significant relationship with the number of embryos obtained and clinical pregnancy rate ( Taylor et al ., 1997 ). Furthermore, another article describes that in women with PCOS undergoing IVF techniques, an LH:FSH ratio greater than 1.5 decreased the clinical pregnancy rate ( Wiser et al ., 2013 ). In the present investigation, differences were found in the fertilization rate, which was lower when the LH:FSH ratio was higher, contrary to the study by Taylor et al. , where no differences were found in dividing samples into ratios greater or less than 1 ( Taylor et al ., 1997 ). Positive correlations were found between the LH:FSH ratio and LBR and CLBR, opposite to what Singh et al., and other authors describe: a high LH:FSH ratio does not impair the outcome of IVF/ICSI cycles in women with PCOS ( Ganor-Paz et al ., 2016 ; Singh et al ., 2021 ). It is thought that the deleterious effects of high LH may be potentially greater in women with PCOS with high LH:FSH ratios. In a 2012 study, the impact of this elevated ratio on reproductive outcomes in IVF treatments was evaluated, associating that a GnRH agonist protocol likely induces a sustained reduction in LH levels, affecting oocyte quality and implantation potential ( Orvieto et al ., 2012 ). This study had some limitations such as its retrospective nature, which, for example, did not allow for the reassessment of anthropometric data during treatment, making it impossible to detect changes in BMI during this period. Additionally, BMI is a nonspecific marker of body composition, failing to distinguish between lean mass and fat mass. The differing definitions of CLBR from article to article made it difficult to compare existing studies with the data obtained in this investigation.

AMH measurement supports clinical decision-making but should not be considered an isolate predictor of reproductive success in IVF techniques. Due to some conflicting data, there remains an urgent need to further investigate the effect of this hormone, exploring both existing and emerging hypotheses, not only as a marker of ovarian reserve. Age has a greater impact on LBR than BMI. Couples facing infertility primarily seek to understand their chances of achieving a live birth before embarking on ART treatments, whether they involve fresh embryo transfers, the use of any surplus frozen embryos from their initial attempt, or subsequent additional treatments. Therefore, in practice, CLBR is the most useful metric as it summarizes the likelihood of a couple achieving a live birth over the entire treatment period. However, although CLBR can be communicated in research articles, due to the use of varied definitions based on different assumptions, it is necessary to standardize its definition and values to accurately reflect the success rate of IVF/ICSI.