Higher pre-pregnancy body mass index was associated with adverse pregnancy and perinatal outcomes in women with polycystic ovary syndrome after a freeze-all strategy: A historical cohort study.

Kun Qian and Lei Jin conceived the study and participated in its design; Xinyao Hu and Enqi Yan wrote the paper; Xinyao Hu and Wenju Peng analyzed the data; Yueping Zhou collected the data. All co‐workers have seen and agreed with the contents of the manuscript. The authors have no conflicts of interest.

The study received approval from the Institutional Review Board of Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology (TJ‐IRB20230394) on March 17, 2023.

This work was supported by a grant from the National Key Research and Development Project (grant number 2018YFA0108401).

A total of 1667 patients with PCOS after the freeze‐all strategy were included in the analysis according to the inclusion and exclusion criteria. These patients were categorized into four groups based on their pre‐pregnancy BMI. There were 137 (8.2%) patients in the underweight group, 1130 (67.8%) patients in the normal weight group, 339 (20.3%) patients in the overweight group, and 61 (3.6%) patients in the obese group. The baseline demographics of patients with PCOS in different BMI groups are shown in Table  1 . Among these groups, duration and type of infertility, antral follicle count, anti‐Müllerian hormone level, dose and duration of gonadotropin, and number of thawed, surviving, and transferred embryos were found to be statistically significant. It was observed that patients in the underweight group had a higher rate of primary infertility. Obese women had longer infertility duration and received larger doses of gonadotropin but had a lower level of anti‐Müllerian hormone than patients in the other BMI groups. The overweight and obese groups had higher antral follicle count and longer duration of gonadotropin. Regarding FET cycle characteristics, no significant differences were found in the proportion of survived embryos/thawed embryos, stage and quality of the transferred embryos, endometrial preparation protocol, or thickness of the endometrium among different BMI groups. Characteristics of women with PCOS according to BMI. Note : Continuous data are reported as medians (first quartile, third quartile) and analyzed by Mann–Whitney U tests. Categorical data are reported as no./total no. (%) and analyzed by chi‐squared test. Bold indicates significant values. Abbreviations: 2PN, two pronuclei; AFC, antral follicle count; AMH, anti‐Müllerian hormone; BMI, body mass index; FSH, follicle stimulating hormone; Gn, gonadotropin; GnRH, gonadotropin‐releasing hormone; ICSI, intracytoplasmic sperm injection; IVF, in vitro fertilization; MII, metaphase II; PCOS, polycystic ovarian syndrome. The pregnancy and perinatal outcomes of patients with PCOS in different BMI groups are described in Table  2 . Significant differences did not exist in the biochemical pregnancy rate, miscarriage rate, or ectopic pregnancy rate among these groups. Moreover, a higher CPR (70.9%) and LBR (57.8%) were observed in the normal weight group compared with the other groups. Pregnancy and perinatal outcomes of women with PCOS by maternal pre‐pregnancy BMI. Note : Continuous data are reported as medians (first quartile, third quartile) and analyzed by Mann–Whitney U tests. Categorical data are reported as no./total no. (%) and analyzed by chi‐squared test. Bold indicates significant values. Abbreviations: BMI, body mass index; GDM, gestational diabetes mellitus; HDP, hypertensive disorders of pregnancy; LBW, low birthweight; LGA, large for gestational age; PCOS, polycystic ovarian syndrome; PTB, preterm birth; SGA, small for gestational age. A total of 634 women with PCOS who had singleton deliveries were included in the perinatal outcome analysis. Among these women, there were 52, 440, 113, and 28 singleton deliveries in the underweight group (8.2%), normal weight group (69.4%), overweight group (17.8%), and obese group (4.4%), respectively. There were no significant differences in the rates of PTB, placenta previa, fetal malformation, LBW, SGA, VLBW, or very SGA among the four groups. However, singletons born to obese and overweight women were prone to be delivered by cesarean section and had significantly higher birthweight as well as Z‐score, the proportions of macrosomia, LGA, and very LGA than those singletons born to women in the other groups. Furthermore, the obese group was associated with a significantly higher rate of obstetric complications, including HDP (14.3%) and GDM (14.3%), and an increased percentage of very PTB (10.7%) when compared with the other groups. Multivariate logistic regression analyses were conducted to clarify the influence of pre‐pregnancy BMI on pregnancy and perinatal outcomes in patients with PCOS. The results of this analysis are presented in Table  3 . The results demonstrated that the overweight group had a significantly lower CPR (aOR = 0.571; 95% CI 0.437–0.746), overall LBR (aOR = 0.606; 95% CI 0.469–0.783), and LBR (aOR = 0.614; 95% CI 0.466–0.809) after blastocyst transfer compared with the normal weight group. In addition, the obese group had a higher risk of biochemical pregnancy (aOR = 2.320; 95% CI 1.052–5.118) and a lower risk of multiple pregnancy rate (aOR = 0.324; 95% CI 0.106–0.992) than the normal weight group. However, pregnancy outcomes in underweight patients were comparable when the normal weight group served as the reference group. Adjusted odds ratios of pregnancy and perinatal outcomes in women with PCOS. Note : Bold indicates significant values. Abbreviations: aOR, adjusted odds ratio; CI, confidence interval; GDM, gestational diabetes mellitus; HDP, hypertensive disorders of pregnancy; LBW, low birthweight; LGA, large for gestational age; PCOS, polycystic ovarian syndrome; PTB, preterm birth; SGA, small for gestational age. Adjusted for age, number of embryos transferred, stage of embryo development, endometrial preparation protocol, fertilization method, cause of infertility, endometrial thickness, and number of oocytes retrieved. Adjusted for age, number and stage of embryos transferred, cause of infertility, and endometrial preparation protocol. Adjusted for age, number and stage of embryos transferred, and cause of infertility. Adjusted for age, number of embryos transferred, endometrial preparation protocol, fertilization method, cause of infertility, endometrial thickness, and number of oocytes retrieved. Adjusted for age, cause of infertility, stage of embryo development, number of embryos transferred, and duration of infertility. For perinatal outcomes, underweight women with PCOS had a significantly higher risk of fetal malformation (aOR = 11.222; 95% CI 1.264–99.588) than normal weight women. However, overweight women with PCOS had significantly higher risks of cesarean section (aOR = 2.536; 95% CI 1.179–5.457), GDM (aOR = 1.974; 95% CI 1.004–3.879), macrosomia (aOR = 3.188; 95% CI 1.622–6.267), LGA (aOR = 2.909; 95% CI 1.803–4.694) and very LGA (aOR = 4.061; 95% CI 2.141–7.704) compared with normal weight women. Moreover, the risks of very PTB, macrosomia, LGA, very LGA, HDP, and GDM were higher in the obese group than in the normal weight group, but this lacked significance. As the number of patients with PCOS in the obese group was limited within this cohort, individuals classified as overweight and obese (BMI ≥25 kg/m 2 ) were combined as one group when conducting multiple regression analyses (Table  S1 ). The results were consistent with the above results. Furthermore, we recategorized the patients with PCOS based on their BMI following the BMI classification for Asian populations to confirm our results. The results of pregnancy and perinatal outcomes are shown in Table S2 . In brief, the normal weight group exhibited a higher CPR (71.4%) and LBR (58.6%) compared with the other BMI categories. The results of the logistic regression analysis can be found in Table  S3 . The findings using the BMI classification for Asian populations aligned with the findings using the WHO classification of BMI. Restricted cubic splines regression analyses were conducted to explore the relationship between BMI and reproductive outcomes in women with PCOS (Figure  1 ). There was a linear relationship between pre‐pregnancy BMI and CPR ( p for nonlinear = 0.721), LBR after blastocyst transfer ( p for nonlinear = 0.597), and rates of PTB ( p for nonlinear = 0.742), HDP ( p for nonlinear = 0.111), macrosomia ( p for nonlinear = 0.846), GDM ( p for nonlinear = 0.453), LGA ( p for nonlinear = 0.187), and very LGA ( p for nonlinear = 0.882). CPR and LBR after blastocyst transfer decreased with an increase in the BMI of patients with PCOS (Figure  1A,B ). In addition, the risks of PTB, GDM, macrosomia, LGA, and very LGA increased with increasing maternal BMI (Figure  1C,E–H ). The risks of HDP decreased as pre‐pregnancy BMI increased, and subsequently increased after a normal BMI was reached (Figure  1D ). The relationship between pre‐pregnancy body mass index (BMI) and different outcomes including clinical pregnancy rate (CPR) (A), live birth rate (LBR) after blastocyst transfer (B), preterm birth (PTB) (C), hypertensive disorders of pregnancy (HDP) (D), gestational diabetes mellitus (GDM) (E), macrosomia (F), large for gestational age (LGA) (G), and very LGA (H) in women with polycystic ovary syndrome (PCOS) by restricted cubic spline regression. The model was adjusted for maternal age at frozen‐thawed embryo transfer (FET), cause of infertility, fertilization method, number of oocytes retrieved, number of embryos transferred, stage of embryo development, endometrial preparation regimens, and the thickness of endometrium in CPR (A). For LBR after blastocyst transfer, the model was adjusted for maternal age at FET, cause of infertility, fertilization method, number of oocytes retrieved, number of embryos transferred, endometrial preparation regimen, and endometrial thickness (B). Analyses of PTB (C), HDP (D), GDM (E), macrosomia (F), LGA (G), and very LGA (H) were adjusted for the following confounders: maternal age at FET, infertility cause, duration of infertility, number of embryos transferred, stage of embryo development, fertilization method, and endometrial preparation regimen. The solid line and blue shading represent the estimated risk (odds ratio; OR) and its 95% confidence intervals (CI). Additionally, we compared the pregnancy and perinatal outcomes of women with PCOS and women without PCOS in each BMI group to investigate whether a diagnosis of PCOS could result in adverse reproductive outcomes (Table  4 ). The result showed that compared with women without PCOS, women with PCOS had a higher clinical pregnancy rate (aOR = 1.416; 95% CI 1.259–1.694) and higher live birth rate (aOR = 1.190; 95% CI 1.035–1.368) but a higher risk of SGA (aOR = 1.596; 95% CI 1.016–2.508) in the normal weight group. Women with PCOS in the overweight group exhibited higher risks of very PTB (aOR = 4.195; 95% CI 1.345–13.090) and GDM (aOR = 2.186; 95% CI 1.012–4.722) compared with women without PCOS. Moreover, there was no significant difference between women with PCOS and women without PCOS in the underweight and obese groups. The effect of BMI on pregnancy and perinatal outcomes in the population of women without PCOS was also investigated (Table  S4 ). The results of multiple logistic regression of women without PCOS in different BMI groups are shown in Table  S5 . Reproductive outcomes of women with PCOS comparing with women without PCOS. Note : Categorical data are reported as no./total no. (%). Bold indicates significant values. Abbreviations: aOR, adjusted odds ratio; CI, confidence interval; GDM, gestational diabetes mellitus; HDP, hypertensive disorders of pregnancy; LBW, low birthweight; LGA, large for gestational age; PCOS, polycystic ovarian syndrome; PTB, preterm birth; SGA, small for gestational age; VLBW, very low birthweight. Adjusted for age, number of embryos transferred, stage of embryo development, endometrial preparation protocol, fertilization method, cause of infertility, endometrial thickness, and number of oocytes retrieved. Adjusted for age, number of embryos transferred, endometrial preparation protocol, fertilization method, cause of infertility, endometrial thickness, and number of oocytes retrieved.

This historical cohort study focused on women with PCOS who had all of their embryos frozen undergoing their first FET cycles. The findings showed that elevated BMI had a negative impact on the rates of clinical pregnancy and live birth and adversely affected maternal and neonatal outcomes, increasing the risks of cesarean section, GDM, LGA, very LGA, and macrosomia when compared with normal BMI. Women with PCOS are predisposed to insulin resistance and metabolic dysfunction, which have negative effects on pregnancy outcomes and may be exacerbated by a higher BMI. 25 The mechanism by which obesity affects reproductive outcomes has been debated. Jungheim et al. 26 conducted a meta‐analysis including individuals using donor eggs and found that there was no association between obesity (BMI ≥30 kg/m 2 ) and the clinical pregnancy rate after IVF, as well as the rates of implantation, miscarriage, and live birth. This study indicated that the overriding factor was oocyte quality rather than endometrial receptivity, explaining the poor pregnancy outcomes of obese women using autologous eggs. However, the patients enrolled in this meta‐analysis were not limited to women with PCOS. Qiu et al. 10 conducted a similar study to ours, and they reported a decrease in the number of top‐quality embryos as BMI increased. They failed to differentiate whether the poor IVF outcomes were attributable to compromised egg quality or impaired endometrial receptivity. However, because of the retrospective nature of this study, it was unable for us to assess the quality of all obtained embryos. The quality of the embryos transferred in FET cycles was evaluated and the proportion of transfers with at least one good‐quality embryo was comparable across subgroups. Therefore, similar to egg quality, compromised endometrial receptivity may also play an important role in causing poorer IVF outcomes in overweight or obese women with PCOS. In fact, the molecules expressed by endometrial cells participate in the insulin signaling pathway, with glucose serving as the main source of energy for the normal function of the endometrium, such as decidualization. 27 , 28 The metabolic disorders arising from maternal obesity could have detrimental effects on the endometrial environment. Bellver et al. 29 analyzed the sequencing data of gene expression during the implantation window in obese women with PCOS and found that it differed from that of healthy controls. Moreover, the nuclear factor‐κB signaling pathway, which is involved in the regulation of inflammation, the cell cycle, and apoptosis, is altered under the conditions of PCOS and obesity. 30 Investigators have shown that the secretion of adiponectin, primarily secreted by adipose tissue and known for its anti‐inflammatory and insulin‐sensitizing properties, is decreased in obese women when they are exposed to hyperandrogenic and hyperinsulinemic conditions. 31 Additionally, tumor necrosis factor‐α, a major effector of the canonical nuclear factor‐κB pathway, is increased in obese women with PCOS compared with normal weight women with PCOS or non‐PCOS obese patients. 30 These changes in molecules and signaling pathways indicated that endometrial receptivity may be compromised in obese women with PCOS. With obesity becoming a public issue, the prevalence of obese women being treated at fertility clinics is growing. 32 Although only a few studies have investigated the impact of pre‐pregnancy BMI on the reproductive outcome of women with PCOS, they have drawn contradictory conclusions. Hynes et al. 33 used data from the 2012–2015 SART CORS database and observed that the proportion of LGA infants and prematurity decreased with increasing BMI in women with PCOS undergoing fresh embryo transfer cycles. Guo et al. 34 found that maternal overweight is an adverse indicator for delivering LGA newborns whereas maternal underweight is a protective factor against LGA in women undergoing FET cycles. Nevertheless, in 2019, the same research team concluded that there was no apparent association between neonatal outcomes and pre‐pregnancy BMI based on different data. 11 The present study found that overweight women with PCOS had higher incidences of LGA, very LGA, and macrosomia, and were prone to deliver by cesarean section. Previous studies, however, failed to determine the effect of pre‐pregnancy BMI on the obstetrics complications of women with PCOS. We analyzed the incidence of HDP and GDM and found an increasing tendency of these outcomes with elevated BMI, although it was not statistically significant. This could be attributed to the small sample size of the obese women with PCOS included in this study. Furthermore, there was no association between maternal underweight and the risk of adverse perinatal outcomes in women with PCOS. The adverse outcomes associated with an elevated BMI observed in the present study may be attributed to early fetal and placental dysfunction resulting from increased insulin resistance and accompanying hyperinsulinemia, oxidative stress, and an inflammatory uterine microenvironment. 35 To investigate the effect of PCOS on reproductive outcomes, the pregnancy and perinatal outcomes were compared between the women with PCOS and without PCOS in each BMI subgroup. The result showed that compared with the non‐PCOS population, a diagnosis of PCOS resulted in a higher clinical pregnancy rate and live birth rate among normal weight women, which was consistent with a previous study. 36 Another study compared the neonatal outcomes between PCOS and non‐PCOS in women with normal BMI and found that PCOS was associated with higher risks of PTB and very PTB. 5 They speculated that this may be attributed to the underlying endocrine and metabolic effect related to PCOS. Further research is needed to reveal the underlying mechanism. This research still had some limitations. First, this study was retrospective, and as a result, we failed to record weight gain during pregnancy. Therefore, it was not possible to determine the effect of body weight changes on perinatal outcomes in women with PCOS. Second, our sample size was relatively small when concerning the analysis of perinatal outcomes. However, the strength of this study was that we investigated the effect of elevated BMI on the pregnancy and birth outcomes of women with PCOS, as both continuous and categorical variables, which increased the reliability of our results. In addition, further analysis using the BMI classification for Asian populations was consistent with our results, which confirmed our conclusions. The findings of this study highlight the clinical significance of pre‐pregnancy weight management in women with or without PCOS.

This historical cohort study focused on women with PCOS who had all of their embryos frozen and subsequently underwent their first FET cycles. Our findings indicated that an increased BMI resulted in adverse live birth outcomes as well as higher risks of cesarean section, LGA, very LGA, and macrosomia in comparison with a normal BMI. Furthermore, the analysis of reproductive outcomes of PCOS and non‐PCOS women showed that a diagnosis of PCOS produced a higher clinical pregnancy rate and live birth rate in normal weight women but higher risks of perinatal complications in normal weight and obese women. Hence, our research findings indicate that weight management is a crucial factor in ensuring the health of the infants of these women.

The prevalence of polycystic ovary syndrome (PCOS), an endocrine disorder associated with female infertility, varies between 5% and 20% among women of reproductive age globally. 1 In addition to anovulation, PCOS is characterized by hyperandrogenism, polycystic ovaries, irregular menstrual cycles, and a higher risk of insulin resistance. 2 , 3 Frozen–thawed embryo transfer (FET) may be a safer option for women with PCOS than fresh embryo transfer, as it can mitigate the risks of ovarian hyperstimulation syndrome (OHSS) and offer patients the opportunity for preimplantation genetic testing. 4 However, recent studies also demonstrated that women with PCOS undergoing FET cycles were at an increased risk of adverse infant outcomes such as preterm birth. 5 It is well known that the pregnancy and perinatal outcomes of patients with PCOS can be influenced by numerous factors such as body mass index (BMI), oligo‐ovulation or anovulation, and metabolic abnormalities. Obese women were found to have a higher rate of miscarriage, particularly during the first trimester, and a lower clinical pregnancy rate (CPR) compared with women of normal weight. 6 , 7 The detrimental effects of obesity on female fertility include impaired endometrial receptivity and reduced oocyte quality. 8 , 9 However, for women with PCOS, the effect of pre‐pregnancy BMI on pregnancy and perinatal outcomes remains unclear. Qiu et al. 10 found that the live birth rate (LBR) decreased but the rate of miscarriage increased in obese women with PCOS, whereas another study 11 suggested that BMI had no significant effect on the pregnancy and perinatal outcomes of women with PCOS after FET cycles. Furthermore, whether a diagnosis of PCOS could result in adverse pregnancy and perinatal outcomes in women with different BMIs remains unknown. The application of the “freeze‐all” policy has become widespread among women with PCOS because of its lower risk of OHSS and higher LBR. The association between pre‐pregnancy BMI and pregnancy and perinatal outcomes in women with PCOS who undergo a freeze‐all strategy is still unclear. Therefore, a historical cohort study was conducted to investigate whether pre‐pregnancy BMI had any impact on pregnancy and perinatal outcomes for women with PCOS after a freeze‐all policy and whether a diagnosis of PCOS could result in adverse reproductive outcomes in women with different BMIs.

The authors declare that there are no competing interests.

A historical cohort study was carried out at the Reproductive Medical Center of Tongji Hospital. We enrolled women applying the freeze‐all strategy for their first in vitro fertilization (IVF) or intracytoplasmic sperm injection cycles. The diagnosis of PCOS was based on the 2003 Rotterdam revised criteria. 12 In our reproductive medicine center, the indications for using the freeze‐all policy include an elevated level of serum progesterone (>1.5 ng/mL), a high risk of OHSS (manifesting as the retrieval of more than 20 oocytes or a high level of estradiol [>7000 pg/mL]) and other conditions when the endometrial environment is unsuitable for embryo transfer. 13 The data analyzed in this study were collected from January 2016 to December 2020. The present study focused on patients' first FET cycles. The exclusion criteria were as follows: (a) patients undergoing preimplantation genetic testing; (b) patients with a history of chronic hypertension or diabetes mellitus before the current pregnancy; and (c) patients receiving oocyte donation or sperm donation. Additionally, patients who were lost to follow up or had missing core data, such as data on gestational weeks and neonatal birthweight, were not included in the data analysis. Patients who met the inclusion criteria were grouped into four categories based on their pre‐pregnancy BMI: underweight (BMI <18.5 kg/m 2 ), normal weight (BMI of 18.5 to <25 kg/m 2 ), overweight (BMI of 25 to <30 kg/m 2 ), and obese (BMI ≥30 kg/m 2 ). The classification of BMI groups followed the World Health Organization BMI classification. 14 Previous literature has provided descriptions of the protocols used for ovarian stimulation, oocyte retrieval, and fertilization procedures. 15 , 16 A good‐quality cleavage‐stage embryo was defined as an embryo with seven or eight blastomeres, with <20% fragmentation, and without multinucleation. Blastocyst morphological grading was based on the Gardner scoring system, 17 and good‐quality blastocysts were defined as those with inner cell mass grades of A or B and trophectoderm grades of A or B. There are several endometrial preparation protocols applied for patients in FET cycles, which have been described elsewhere. 5 , 18 Generally, patients who had regular menstrual cycles used natural cycles for endometrial preparation. For patients who had irregular menstrual cycles, artificial or stimulated cycles were mainly used. In stimulated cycles, letrozole was added from Days 3 to 5 of the menstrual cycle, and the monitoring of follicle development was initiated on Day 10. Once the dominant follicle reached 14 mm on Day 10, no additional medication was administered until ovulation was triggered. If the diameter of the dominant follicle was less than 14 mm, a daily dosage of 75–150 IU of human menotropins was added to facilitate follicle development. For luteal phase support, oral dydrogesterone (20 mg, Duphaston; Solvay Pharmaceuticals BV) was used 1 day after ovulation in natural cycles or stimulated cycles. In artificial cycles, oral estradiol (Progynova; Bayer) was administered at a dosage of 2 mg/day from Days 1 to 4, 4 mg/day from Days 5 to 8, and 6 mg/day from Days 9 to 12. Patients' ovulation and endometrial thickness were then evaluated on Day 13. The dose of oral estradiol was adjusted dynamically based on endometrial thickness. Once the endometrial thickness reached 8 mm, the administration of 40 mg progesterone via intramuscular injection was started. For all endometrial preparation protocols, luteal phase support was sustained until 10 weeks of gestation. FET was performed on the 4th day of intramuscular progesterone for cleavage‐stage embryos and on the 6th day for blastocysts. 19 In the present study, both pregnancy outcomes and perinatal outcomes were included. The information collected for data analysis was obtained from electronic records followed up by trained staff after each patient's delivery. The primary outcomes of this study were pregnancy outcomes. A positive pregnancy test was confirmed by serum human chorionic gonadotropin levels 2 weeks after embryo transfer. Clinical pregnancy was confirmed by the presence of one or more gestational sacs with a fetal heartbeat on transvaginal ultrasound. Biochemical pregnancy was defined as a positive human chorionic gonadotropin test that did not result in a clinical pregnancy. Miscarriage referred to the spontaneous loss of a pregnancy occurring earlier than 24 weeks of gestation. The definition of live birth was the delivery of at least one live newborn after 24 weeks of gestation. The secondary outcomes examined in this study were perinatal outcomes. To minimize the bias caused by vanishing twin syndrome, only women with singleton pregnancies were included in our analysis of obstetric complications and neonatal outcomes. Abnormal obstetrical complications included hypertension disorders in pregnancy (HDP), placenta previa, and gestational diabetes mellitus (GDM). The diagnosis of HDP followed the consensus of the International Society for the Study of Hypertension in Pregnancy, 20 which encompassed both preeclampsia and gestational hypertension. The diagnosis of GDM was based on previously established consensus criteria. 21 Neonatal outcomes included fetal malformation, gestational age, preterm birth (PTB), delivery mode, and newborn birthweight. In FET cycles, the determination of gestational age involved adding 17 days for cleavage‐stage embryo transfer or 19 days for blastocyst transfer, starting from the date of FET. PTB was defined as delivery occurring before 37 weeks, whereas very PTB was defined as delivery occurring before 32 weeks. The evaluation of singleton birthweight included Z‐score , low birthweight (LBW), very low birthweight (VLBW), small for gestational age (SGA), very SGA, macrosomia, large for gestational age (LGA), and very LGA. Adjustments were made for infant gender and gestational age when calculating the Z‐score using the following formula: Z‐score  = ( x  −  μ )/ σ . In this formula, x represents infant birthweight, μ denotes the average birthweight of the reference population for the corresponding gender, and σ is the standard deviation as indicated in the reference chart. 3 LBW and VLBW were defined as birthweights less than 2500 and less than 1500 g, respectively. Birthweights <10th and <3rd centiles were defined as SGA and very SGA, respectively. An infant with a birthweight exceeding 4000 g was considered macrosomic. The definitions of LGA and very LGA were birthweights above the 90th and 97th centiles, respectively. The reference centiles for birthweight including Z‐score , SGA, very SGA, LGA, and very LGA were based on the Chinese population at different gestational weeks. 22 All data analyses were conducted using SPSS software version 26.0 (IBM, Armonk, NY, USA) and R (version 4.3.0). In the present study, continuous variables are presented as medians (interquartile range), whereas categorical variables are shown in the form of frequencies and proportions. To assess the normality of continuous variables, the Kolmogorov–Smirnov or Shapiro–Wilk test was used. The comparison of continuous variables was carried out using the Kruskal–Wallis test. An analysis of categorical variables was conducted by Pearson's chi‐squared test or Fisher's exact test, as appropriate. Multiple comparisons were made using Bonferroni correction, if necessary. To explore the impact of pre‐pregnancy BMI on pregnancy and perinatal outcomes after a freeze‐all strategy, multivariate logistic regression analyses were performed. The covariates in the multiple logistic regression included maternal age at FET, cause of infertility, fertilization method, number of oocytes retrieved, number of embryos transferred, stage of embryo development, duration of infertility, endometrial preparation regimen, and endometrial thickness. The results are described as crude odds ratios (crude ORs) or adjusted odds ratios (aORs) with 95% confidence intervals (CIs). Because categorizing continuous variables may reduce the statistical power for detecting correlations between the exposure and outcome, 23 restricted cubic splines with three knots were used to analyze the nonlinear relationship between pre‐pregnancy BMI and various outcomes after adjusting for covariates. All statistical tests were two‐tailed. Statistical significance was defined as a p value less than 0.05. Furthermore, to validate the results of this study, we categorized patients with PCOS based on the criteria of the BMI classification for Asian populations 14 , 24 : underweight (BMI <18.5 kg/m 2 ), normal weight (18.5 kg/m 2  ≤ BMI < 23 kg/m 2 ), overweight (23 kg/m 2  ≤ BMI < 27.5 kg/m 2 ), and obese (BMI ≥27.5 kg/m 2 ).

Figure S1. Table S1. Table S2. Table S3. Table S4. Table S5.