Repeated cryopreservation is associated with poor pregnancy outcomes after frozen–thawed embryo transfer in women aged < 35 years: a propensity-score matched cohort study Authors: Zengshu Huang, Chunhui Zhang, Nan Wang, Huiqing He, Song Li, Changzhong Li Published: 2025 ABSTRACT AbstractBackgroundRepeated cryopreservation of embryos in frozen–thawed embryo transfer (FET) has raised concerns regarding its impact on pregnancy outcomes. This study aimed to reveal the association between repeated cryopreservation by vitrification and FET outcomes, and to identify factors influencing these outcomes.MethodsA single-center retrospective cohort study was conducted, analyzing 13,624 FET cycles (from 2018 to 2022) involving embryos with cryopreservation or repeated cryopreservation. Propensity score matching was employed in this study. The primary outcome was clinical pregnancy rate (CPR) while live birth rate (LBR), miscarriage along with neonatal outcomes were also assessed. Multivariate logistic regression and stratified analyses were used to explore the effect of female age and the quality of embryo on CPR after adjusting for confounders.ResultsIn total, 321 FET cycles met the inclusion and exclusion criteria were included after matching and divided into re-cryopreservation group (n = 107) and cryopreservation group (n = 214). The re-cryopreservation group showed lower CPR (27.10% vs. 45.33%; aOR, 0.48; 95% CI 0.28–0.84) and LBR (18.69% vs. 38.32%; aOR, 0.35; 95% CI 0.19–0.65), with no differences in miscarriage or biochemical pregnancy rates. Stratified analyses revealed an adverse effect of women aged < 35 years (aOR, 0.40; 95% CI 0.19–0.83) and single-embryo transfers (aOR, 0.51; 95% CI 0.28–0.94) on CPR while transfer of good quality embryos reduced these effects (aOR, 0.55; 95% CI 0.21–1.44). Gestational age, birthweight and congenital anomalies proportion were not significantly different between the two groups.ConclusionsRepeated cryopreservation was associated with reduced clinical pregnancy and live birth rates, particularly in younger patients. The neonatal outcomes were potentially not affected. Clinicians may consider limiting repeated cryopreservation for women under 35 years and prioritize good-quality embryos for re-vitrification when medically justified. These findings underscored the need for tailored strategies in FET cycles involving re-cryopreserved embryos.Supplementary InformationThe online version contains supplementary material available at 10.1186/s40001-025-03425-2.Keywords: Frozen–thawed embryo transfer, Re-cryopreservation, Clinical pregnancy, Live birth, Perinatal outcomes Introduction Since the first success obtained using frozen human embryos, cryopreservation techniques have been described to assisted reproduction treatment (ART). It has revolutionized the ability to perform dissociation of oocyte retrieval from embryo transfer, accounting for a new era of infertility therapies. The evolution of embryo cryopreservation ensures that frozen–warmed transfer does not worsen perinatal outcomes or increase the events of major congenital anomalies compared with fresh embryo transfer [1–3]. The worldwide application of frozen–thawed embryo transfer (FET) has been not only increasing the cumulative pregnancy rate within a single ovarian stimulation cycle but also minimizing the risks of multiple pregnancy and ovarian hyperstimulation syndrome. It also enables embryo banking for all patient populations seeking future fertility preservation, with particular significance for poor responders. However, surplus surviving embryos for repeated cryopreservation may occasionally occur in FET cycles under specific circumstances: some patients request thawing multiple cleavage-stage embryos for blastocyst culture and transfer; transfer may be canceled at the last minute due to personal or medical reasons; patients may need to await biopsy or rebiopsy results for blastocysts with inconclusive preimplantation genetic testing (PGT) diagnoses. These scenarios create new challenges for the disposition of repeated cryopreserved embryos in storage. As the introduction of more operations during embryo freezing and thawing, concerns about the safety of re-cryopreservation of embryos have been raised. A meta-analysis including 14 clinical studies discussed the adverse influence of repeated cryopreservation on embryo viability, subsequently leading to increased the risk of miscarriages and decreased live births, but showed no impact on neonatal outcomes [4]. Thus, Wang et al. addressed a warning of detrimental role of repeated cryopreservation and warming on embryos and called for long-term follow-up studies with large sample sizes. It is noted that equivalent live birth rates may be expected after the transfer of re-cryopreserved embryos in other studies [5]. The conflicting conclusions from data on the FET outcomes of embryos after being cryopreserved and thawed twice present a decision-making dilemma for clinicians and patients. More importantly, the patients impacted most or least after re-cryopreserved embryo transfer remain poorly understood. Identifying the risk of pregnancy failure in women and/or couples using re-cryopreserved embryos could allow fertility centers to judge who would benefit from repeated embryo cryopreservation. Therefore, a retrospective cohort study was designed to compare the FET outcomes of embryo transfers that underwent two rounds of cryopreservation and thawing with those that underwent one round and explore the effect of female age and embryo quality per transfer on clinical pregnancy. We examined the embryo implantation rate (IR), clinical pregnancy rate (CPR), live birth rate (LBR) and followed the maternal complications and neonatal outcomes, such as birthweight and anomalies. The research could provide new evidence for clinicians to optimize strategies for re-frozen–thawed embryo transfer. Methods Study design and cohort selection This single-center retrospective cohort study was reviewed and approved by the Ethics Committee of Peking University Shenzhen Hospital (reference number 2024-154). All FET cycles between January 1, 2018 and December 31, 2022, were assessed. These cycles were classified into once-cryopreservation embryo transfer cycle group (cryopreservation group) and repeated cryopreservation embryo transfer cycle group (re-cryopreservation group) based on the number of frozen–thawed times before embryo transfer. Prior studies recommended vitrification as a better choice than slow freezing after assessment of the impact of different cryopreservation strategies on pregnancy outcomes [4]. Hence, all the embryos in FET cycles of this study for propensity-score matching were cryopreserved via vitrification. The exclusion criteria were as follows: (1) female age > 45 years, (2) couples with chromosomal abnormalities, (3) PGT cycles, (4) transfer cycles canceled after embryo warmed, (5) patients with in vitro fertilization (IVF) or intra-cytoplasmic sperm injection (ICSI) performed before in other hospitals, (6) severe uterine abnormalities (including double uterus, unicornuate uterus, etc.), (7) triple-or-more-cryopreserved embryo transfer cycles, (8) errors of patient ART number or missing final IVF/ICSI data. Embryo freezing and warming procedures After IVF or ICSI, all embryos were assessed by the routine evaluation system. Cleavage-stage embryos on day 3 after oocyte retrieval (D3) with ≥ 6 blastomeres, fragmentation ≤ 25% and basically symmetrical blastomeres were viable for vitrification. The remaining cleavage-stage embryos were cultured in blastocyst medium. The viable blastocyst criteria for frozen were at least ratings of above CC, according to Gardner criteria [6] consisting of the size of blastocyst cavity, inner cell mass and trophoblast cells. The vitrification and thawing procedures were performed as previously described [7]. Embryo quality grading The survival standard of cleavage-stage embryo after thawing was that more than half of blastomeres survived. A high-quality cleavage-stage embryo was scoring I/8 and IIa/7–14 within 10% fragmentation [8]. The survival standard for thawed blastocysts was re-expansion with less than half of the inner cell mass or trophoblast cells showing morphological damage. With reference to the Gardner criteria, the standard for good-quality blastocysts was defined as on D5, an embryo development stage ≥ 3 with ratings of BB above; on D6, stage ≥ 4 with ratings of BB above. Study variables and outcome measures Patient baseline demographics are listed as follows: female age, male age, infertility duration, body mass index (BMI), basal follicle-stimulating hormone levels (FSH), luteinizing hormone levels (LH), prolactin levels (PRL), testosterone levels (T), anti-Müllerian hormone levels (AMH), primary/secondary infertility, ovulation induction protocols, insemination methods. Other variables in this study were measured including endometrial parameters: endometrial thickness and type on the transfer day, endometrial preparation protocol; and embryology parameters: number of embryos warmed and transferred, type of embryos transferred (embryos at cleavage embryo/extended culture of cleavage embryo/blastocyst/others), number of good-quality embryos, survival embryos. The serum hCG ≥ 25 IU/L at 14 days after FET was considered positive for pregnancy. An ultrasound examination was conducted 28 days after transfer to distinguish clinical pregnancy from biochemical pregnancy, indicated by the presence of a gestational sac (GS). Multiple pregnancy was diagnosed via ultrasound if more than one intrauterine GS with a visible heartbeat or one GS with two or more visible heartbeats. Ectopic pregnancy was determined by the ultrasound showing a GS outside uterine cavity. Miscarriage was defined as pregnancy termination before 28 weeks, with early miscarriage (before 12 weeks) and late miscarriage (at 12–28 weeks). Live birth was defined as a delivery of at least one live-born infant. Preterm birth (< 37 weeks) and full-term birth (37–42 weeks) were distinguished based on gestational age. Low birth weight (LBW) was defined as a full-term birth weight < 2,500 g. Maternal complications of gestational hypertension disease (ICD-10 code O13-15), gestational diabetes mellitus (ICD-10 code O24.4), placenta accreta spectrum disorders (ICD-10 code O43.2) and postpartum hemorrhage (ICD-10 code O72), preterm delivery and cesarean section were analyzed. Neonatal outcomes included incidence of LBW, newborn gender, neonatal birth weight and length, congenital anomalies. The primary observation indicator was CPR. The secondary observation indicators included embryo IR, LBR, biochemical pregnancy rate, miscarriage rate, LBW incidence and congenital anomaly incidence. Statistical analysis All statistical analyses were performed by Empower (R: http://www.r-project.org; version 4.3.0) (www.empowerstats.com, X&Y solutions, Inc. Boston, Massachusetts, USA). The continuous variables presented as mean ± standard deviation (X ± S) and the categorical variables were expressed in numbers and percentages (n/N, %). Comparisons of continuous or categorical variables between groups were performed using Student’s t test or Wilcoxon rank-sum test and Pearson’s chi-square test or Fisher’s exact test when appropriate. Given the inherent differences between eligible patients in two groups, propensity-score matching (PSM) was applied for data processing. We calculated a propensity score (PS) for the potential confounding variables: female age, infertility type, type and number of embryos transferred, number of FET cycles and year and month of FET. Matching was based on the method of 1:2 matching protocol without replacement (greedy-matching algorithm), with a caliper width equal to 0.01 of the standard deviation of the logit of PS. Standardized differences were estimated to verify a cohort of transfer cycles with similar baseline characteristics. The comparison of the differences of binary outcome between matched groups after PSM were used by multivariable logistic regression. Variables with multicollinearity were excluded. Crude odds ratios (ORs) and adjusted ORs with 95% confidence intervals (CIs) for each outcome were calculated. The likelihood ratio test and Wald test were performed after adjustment for subgroup factors. Statistical significance was determined at two-tailed p < 0.05. Results The study collected a total of 13,624 FET cycles from our reproductive medicine center. Of these, 1,829 cycles were excluded described as methods. 19 cycles involving re-cryopreserved embryos mixed with cryopreserved embryos, as well as the cycles mixed with fresh embryo (n = 1) or slow-freezing embryo (n = 1) were further excluded from re-cryopreservation group. A total of 112 re-cryopreserved embryo transfer cycles and 11,661 cryopreserved embryo transfer cycles were included (Fig. 1). The distribution of indications for re-cryopreservation was as follows: 107 cycles were due to the clinical practice of elective single embryo transfer (eSET) followed by cryopreservation of surplus embryos; 3 cycles were due to abandonment of the current transfer cycle caused by the absence of good-quality blastocyst formed after thawing; 1 cycle was due to non-medical cancellation from patient after embryo thawing; 1 cycle was because the patient was suddenly caught by a severe cold. Finally, using PSM, we successfully matched 107 cycles as re-cryopreservation group and 214 cycles as cryopreservation group. The distributions of PS were visualized in this cohort study, and closely matched after matching (Fig. S1). Flow diagram for study enrollment Baseline characteristics The demographics and baseline IVF characteristics of overall observed subjects were displayed before and after matching (Table 1). Significant differences were observed in the type and number of embryos transferred per cycle, the year and the times of FET before matching (p < 0.05). After PSM, the year of FET was comparable, and meanwhile, the disparities in the times of FET, the type and number of embryos transferred per cycle were reduced. There were no notable differences observed in other baseline variables between the intersected cohorts including indicators of ovarian reserve, such as basal FSH and AMH, insemination method, except for ovulation induction protocol and the endometrial type at transfer day (Table 2). Table 1Comparison of demographic characteristics of cryopreservation and re-cryopreservation embryo cycles before and after PSMBefore matchingAfter matchingCryoRecryoSDP valueCryoRecryoSDP valueN11,661112214107Female agea (year)33.5 ± 4.433.7 ± 4.40.0 (− 0.2, 0.2)0.70634.0 ± 5.033.6 ± 4.30.1 (− 0.1, 0.3)0.458Male agea (year)35.5 ± 5.936.4 ± 4.90.2 (− 0.0, 0.3)0.11436.3 ± 6.636.5 ± 4.90.0 (− 0.2, 0.3)0.831Infertility duration(year)3.7 ± 2.74.1 ± 3.10.2 (− 0.0, 0.3)0.0813.5 ± 2.94.1 ± 3.10.2 (− 0.0, 0.4)0.097Infertility Type0.2 (− 0.0, 0.4)0.0750.0 (− 0.2, 0.2)0.934 primary5142 (44.1%)40 (35.7%)73 (34.1%)37 (34.6%) secondary6519 (55.9%)72 (64.3%)141 (65.9%)70 (65.4%)No. of embryosper transfer cycle1.8 ± 0.41.1 ± 0.31.6 (1.4, 1.8) < 0.0011.3 ± 0.71.2 ± 0.40.4 (0.1, 0.6)0.007Type of embryotransferred1.2 (1.0, 1.4) < 0.0010.5 (0.2, 0.7)0.005 Blastocyst4839 (41.5%)98 (87.5%)167 (78.0%)94 (87.9%) Cleavage embryo3401 (29.2%)4 (3.6%)24 (11.2%)3 (2.8%) Extended culture of cleavage embryo1642 (14.1%)9 (8.0%)10 (4.7%)9 (8.4%) Others1785 (15.2%)1 (0.9%)13 (6.1%)1 (0.9%)Transfer date year0.6 (0.5, 0.8) < 0.0010.2(− 0.0, 0.4)0.588 20182983 (25.6%)8 (7.1%)29 (13.6%)8 (7.5%) 20192568 (22.0%)17 (15.2%)33 (15.4%)17 (15.9%) 20201912 (16.4%)17 (15.2%)33 (15.4%)17 (15.9%) 20212210 (19.0%)33 (29.5%)51 (23.8%)30 (28.0%) 20221989 (17.1%)37 (33.0%)68 (31.8%)35 (32.7%)No. of FET1.5 (1.3, 1.7) < 0.0010.8 (0.6, 1.1) < 0.001 16752 (57.9%)5 (4.5%)60 (28.0%)5 (4.7%) 23158 (27.1%)41 (36.6%)68 (31.8%)41 (38.3%) 31175 (10.1%)39 (34.8%)33 (15.4%)39 (36.4%) 4409 (3.5%)17 (15.2%)31 (14.5%)17 (15.9%) 5114 (1.0%)6 (5.4%)15 (7.0%)5 (4.7%) 639 (0.3%)1 (0.9%)4 (1.9%)0 (0.0%) 712 (0.1%)1 (0.9%)2 (0.9%)0 (0.0%) 83 (0.0%)1 (0.9%)1 (0.5%)0 (0.0%) 90 (0.0%)1 (0.9%)––Data are presented as mean ± SD/N (%)Cryo, cryopreservation group; Recry, re-cryopreservation group; SD, standardized difference; No., number; others, mixed extended culture of cleavage embryo with blastocyst in one transfer cycle; FET, frozen–thawed embryo transferaage at retrieval Comparison of demographic characteristics of cryopreservation and re-cryopreservation embryo cycles before and after PSM Infertility duration (year) No. of embryos per transfer cycle Type of embryo transferred Data are presented as mean ± SD/N (%) Cryo, cryopreservation group; Recry, re-cryopreservation group; SD, standardized difference; No., number; others, mixed extended culture of cleavage embryo with blastocyst in one transfer cycle; FET, frozen–thawed embryo transfer aage at retrieval Table 2Comparison of clinical and laboratory indicators between cryopreservation and re-cryopreservation group after matchingCryopreservation groupRe-cryopreservation groupStandardize differenceP valueCycles214107 BMI (kg/m2)21.1 ± 2.721.0 ± 2.70.0 (− 0.2, 0.3)0.688 FSH (mIU/mL)8.2 ± 3.87.6 ± 2.90.2 (− 0.0, 0.4)0.126 LH (mIU/mL)5.1 ± 3.35.0 ± 2.90.1 (− 0.2, 0.3)0.672 PRL (ng/mL)17.9 ± 13.019.0 ± 20.20.1 (− 0.2, 0.3)0.531 T (ng/mL)0.5 ± 0.20.5 ± 0.30.1 (− 0.2, 0.3)0.584 AMH (ng/mL)4.6 ± 3.74.3 ± 3.00.1 (− 0.1, 0.3)0.471Ovulation induction protocol0.6 (0.4, 0.9) < 0.001 Long GnRH agonist72 (33.6%)61 (57.0%) GnRH antagonist99 (46.3%)30 (28.0%) Luteal-phase stimulation8 (3.7%)3 (2.8%) Minimal stimulation8 (3.7%)9 (8.4%) others27 (12.6%)4 (3.7%)Insemination method0.0 (− 0.2, 0.3)0.966 IVF133 (62.1%)68 (63.6%) ICSI68 (31.8%)33 (30.8%) IVF + ICSI13 (6.1%)6 (5.6%)Endometrial thickness(mm)10.9 ± 2.410.5 ± 2.10.2 (− 0.0, 0.4)0.105Endometrial type0.3 (0.1, 0.6)0.036 A33 (15.6%)18 (17.5%) B68 (32.2%)19 (18.4%) C110 (52.1%)66 (64.1%)Endometrium preparedregiment0.2 (− 0.0, 0.4)0.391 NC50 (23.4%)31 (29.0%) HRT86 (40.2%)47 (43.9%) COH45 (21.0%)17 (15.9%) GnRHa–HRT33 (15.4%)12 (11.2%)Pregnancy complicationsNS GDM5 (5.6.%)0 (0.0%) GHD4 (4.5%)1 (3.8%) PAS1 (1.1%)1 (3.8%) Others14 (15.7%)2 (7.7%)No. of deformed fetus0 (0.0%)1 (3.3%)NSData are presented as mean + SD/N (%)NC, natural cycle; HRT, hormone replacement therapy; COH, controlled ovarian hyperstimulation; GnRHa–HRT, hormone replacement therapy with gonadotrophin-releasing hormone analogue down-regulation treatment; GDM, gestational diabetes mellitus; GHD, gestational hypertension disease; PAS, placenta accreta spectrum disorders; NS, no significance; No., number Comparison of clinical and laboratory indicators between cryopreservation and re-cryopreservation group after matching Endometrial thickness (mm) Endometrium prepared regiment Data are presented as mean + SD/N (%) NC, natural cycle; HRT, hormone replacement therapy; COH, controlled ovarian hyperstimulation; GnRHa–HRT, hormone replacement therapy with gonadotrophin-releasing hormone analogue down-regulation treatment; GDM, gestational diabetes mellitus; GHD, gestational hypertension disease; PAS, placenta accreta spectrum disorders; NS, no significance; No., number Pregnancy outcomes The clinical characteristics and outcomes were compared between the cryopreservation and re-cryopreservation group (Table 3). In our center, the overall survival rate was 97.59% for re-cryopreserved embryos and 97.11% for cryopreserved embryos, respectively. After matching, a total of 488 vitrified embryos were thawed, among which 412 embryos were filtered into FET cycles. The embryo survival rates were 99.22% in the re-cryopreserved group and 98.88% in the cryopreserved group. A total of 294 vitrified-warmed human blastocysts were transferred. Among them, 109 had reached the blastocyst stage and been transferred by day 5 after fertilization, 159 by day 6, and 26 by day 7. The re-cryopreservation group showed a higher proportion of D6/D7 blastocysts (53.19% vs. 80.19%; p < 0.05), while the cryopreservation group had a significantly higher proportion of good-quality embryo transfers compared to the re-cryopreservation group (43.01% vs. 25.81%; p < 0.05). The IR (36.10% vs. 25.00%; p < 0.05) declined significantly after repeated cryopreservation. The re-cryopreservation group had lower rates of clinical pregnancy (45.33% vs. 27.10%; p < 0.05) and live birth (38.32% vs. 18.69%; p < 0.05) than the cryopreservation group. The prevalence rates of other pregnancy outcomes, including miscarriage, ectopic pregnancy and biochemical pregnancy were comparable between two groups. Table 3Comparison of embryo viability and pregnancy outcomes between cryopreservation and re-cryopreservation group after matchingCryopreservation groupRe-cryopreservation groupP valueNo. of embryo warmed360128Embryo survival (%)98.88 (356/360)99.22 (127/128)0.867No. of embryo transferred288124Blastocysts transferred188106 < 0.001D588 (46.81%)21 (19.81%)D686 (45.74%)73 (68.87%)D714 (7.45%)12 (11.32%)No. of good-quality embryo transferred124320.005Cleavage embryos30 (10.42%)0 (0.00%)Blastocysts94 (32.64%)32 (25.81%)Serum hCG after 14d0.014 Negative (< 25mIU/mL)107 (50.00%)69 (64.49%) Positive (> = 25mIU/mL)107 (50.00%)38 (35.51%)Embryo implantation (%)36.11 (104/288)25.00 (31/124)0.037Clinical pregnancies97290.016 Singletons91 (42.52%)27 (25.23%) Twins5 (2.34%)2 (1.87%) Triples1 (0.47%)0 (0.00%)Live Birth8220 < 0.001 Singletons78 (36.45%)19 (17.76%) Twins4 (1.87%)1 (0.93%)Biochemical pregnancies10 (4.67%)8 (7.48%)0.303Miscarriages0.687 Early (< 12 weeks)11 (5.14%)7 (6.54%) Late (12–28 weeks)3 (1.40%)2 (1.87%)Ectopic pregnancies0 (0.00%)1 (0.93%)0.333 Comparison of embryo viability and pregnancy outcomes between cryopreservation and re-cryopreservation group after matching Given the significant differences in female age, infertility type, FSH level, and insemination method, we put these possible confounders as well as the number of ART and FET cycles, the quality, type and number of embryos transferred per cycle into multivariate models sequentially. Multi-variable logistic regression analyses demonstrated an independent effect of the frozen–thawed times of embryos on CPR (aOR, 0.48; 95% CI 0.28–0.84) even after these confounders were controlled for (Table 4). Embryos with re-cryopreservation had higher risk of clinical pregnancy failure than did embryos with cryopreservation in FET cycles. In addition, we found persistent differences in the LBR (aOR, 0.35; 95% CI 0.19–0.65) (Table 5). The miscarriage and biochemical pregnancy rates of repeated cryopreservation embryos were also investigated. Compared with crude regression analyses, the associations between these pregnancy outcomes and the frozen–thawed times of embryos did not change markedly after adjusting for above confounders (Table 5). Table 4Multivariable model and clinical pregnancy in re-cryopreservation vs. cryopreservation embryo transferCI, confidence interval; ART, assisted reproduction treatment; FET, frozen–thawed embryo transfer Multivariable model and clinical pregnancy in re-cryopreservation vs. cryopreservation embryo transfer CI, confidence interval; ART, assisted reproduction treatment; FET, frozen–thawed embryo transfer Table 5Multivariable logistic regression analysis for pregnancy outcomes between cryopreservation and re-cryopreservation group using the final adjusted modelAdjusted for female age, infertility type, FSH, insemination methods, number of ART (< = 3, > 3), number of FET (< = 3, > 3), quality of embryo transferred, type of embryo transferred, number of embryo transferred per cycleCryo, cryopreservation group; Re-cryo, re-cryopreservation group; CI, confidence interval Multivariable logistic regression analysis for pregnancy outcomes between cryopreservation and re-cryopreservation group using the final adjusted model Adjusted for female age, infertility type, FSH, insemination methods, number of ART (< = 3, > 3), number of FET (< = 3, > 3), quality of embryo transferred, type of embryo transferred, number of embryo transferred per cycle Cryo, cryopreservation group; Re-cryo, re-cryopreservation group; CI, confidence interval A series of stratified analyses were performed accordingly (Table 6). For female age ≥ 35 years, CPR did not differ between the two groups (aOR, 0.55; 95% CI 0.22–1.40). However, a distinct decrease in CPR was observed between the subgroups of female age < 35 years (aOR, 0.40; 95% CI 0.19–0.83). Moreover, when transferring non-good-quality embryos, repeated vitrification was associated with a detrimental effect on CPR (aOR, 0.40; 95% CI 0.20–0.83), which was not evident in cycles, where at least one good-quality embryo was transferred (aOR, 0.55 [95% CI 0.21–1.44]). In addition, in embryo transfer of single embryo, the re-cryopreservation group exhibited a significantly lower aOR for CPR (aOR, 0.51; 95% CI 0.28–0.94), whereas not in embryo transfer of two or more embryos (aOR, 0.39; 95% CI 0.07–2.19). Table 6Stratified analysis categorized by female age and number and quality of embryo in comparison of clinical pregnancy between cryopreservation and re-cryopreservation groupAdjusted for female age, infertility type, FSH, insemination methods, number of ART (< = 3, > 3), number of FET (< = 3, > 3), quality of embryo transferred, type of embryo transferred, number of embryo transferred per cycleCI, confidence interval Stratified analysis categorized by female age and number and quality of embryo in comparison of clinical pregnancy between cryopreservation and re-cryopreservation group Adjusted for female age, infertility type, FSH, insemination methods, number of ART (< = 3, > 3), number of FET (< = 3, > 3), quality of embryo transferred, type of embryo transferred, number of embryo transferred per cycle CI, confidence interval Finally, a corresponding sensitivity analysis was also performed in subgroup of single blastocyst transfer cycles. It showed that the risks of clinical pregnancy failure in two groups were elevated only among the cycles with single non-good-quality blastocyst (aOR, 0.31; 95% CI 0.13–0.76) but similar to those with single good quality blastocyst (aOR, 0.66; 95% CI 0.23–1.91) (Table S1). Maternal and neonatal outcomes To further investigate whether the frozen–thawed times of embryos were associated with adverse maternal and neonatal outcomes, we explored the follow-up records of these patients until childbirth. For obstetrical outcomes, no increased proportion of GDM, GHD, PAS and other obstetrical complications was shown with the comparison of these two groups (Table 2). In re-cryopreservation group, only one case of fetal abortion was conducted due to trisomy 21 syndrome. Among the 97 cases of singleton deliveries, no significant differences were found regarding the gestational age, cesarean section, infant gender ratio, newborn length, birthweight, low birthweight rate, and pre-term birth rate between two groups (Table 7). The proportions of congenital birth defects in the neonates were 0/19 in the re-cryopreservation group, and 2/78 in the cryopreservation group, where one singleton baby born had a left hypdronephrosis and the other was born with angioma. There were no reports of stillbirth or neonatal mortality in either group. Table 7Neonatal outcomes of singleton live birth after frozen–thawed embryo transferCryopreservation groupRe-cryopreservation groupStandardize differenceP valueSingleton delivery7819––Week of delivery38.96 ± 1.5238.74 ± 0.940.219 (− 0.51, 0.94)0.550Birth age type0.3 (− 0.2, 0.8)0.583 Full term birth (> = 37 weeks)74 (94.9%)19 (100.0%) Pre-term birth (< 37 weeks)4 (5.1%)0 (0.00%)Delivery Type0.5 (− 0.0, 1.0)0.1421 CS55 (70.51%)17 (89.47%) Natural labor23 (29.49%)2 (10.53%)Newborn gender0.1 (− 0.4, 0.6)0.676 Girl41 (52.6%)11 (57.9%) Boy37 (47.4%)8 (42.1%)Newborn length (cm)49.94 ± 1.9850.26 ± 1.56− 0.32 (− 1.30, 0.66)0.517Newborn weight (g)3193.38 ± 445.953261.05 ± 328.02− 67.68 (− 284.30, 148.95)0.537Low birth weight (g)0.2 (− 0.3, 0.7)1.000 < 25002 (2.7%)0 (0.0%) > = 250071 (97.3%)19 (100.0%)Newborn anomalies20NSCS, cesarean section; NS, no significance Neonatal outcomes of singleton live birth after frozen–thawed embryo transfer CS, cesarean section; NS, no significance Discussion This is a single-center retrospective cohort study comparing pregnancy and perinatal outcomes after FET with different times of vitrification of embryos. Our results confirmed that the FET cycles with repeated cryopreservation embryos were more likely to experience pregnancy failure and failed live birth deliveries compared to those using once-cryopreserved embryos, possibly due to impaired embryo viability. No difference in the rate of maternal complications was found between groups that had been cryopreserved once or twice. As for the neonatal outcomes, repeated cryopreservation showed no significantly negative effects on preterm labor, birthweight and gender ratio of newborn in women with singleton pregnancy. The techniques of embryo cryopreservation, especially vitrification, are constantly improved through decades of efforts. The effectiveness and safety of transfer of vitrified–thawed embryos have been confirmed by a great deal of research [9]. In clinical practice, repeated cryopreservation may be unforeseen, particularly with the growing adoption of eSET strategies, which could potentially result in surplus good-quality embryos after thawing [10, 11]. Previous studies have shown that re-vitrified cryopreservation not only resulted in a lower CPR and LBR [12] but also allowed a higher miscarriage rate, despite a comparable neonatal complication rate [13]. Coinciding with those findings, this study demonstrated the adverse effect of additional cryopreservation on embryo development potential in FET, except for a comparable miscarriage rate. Whereas current clinical studies on the impact of embryo cryopreservation times on clinical outcomes have yielded inconsistent conclusions. Some studies have reported that repeated cryopreservation did not adversely affect pregnancy rates in unbiopsied blastocyst transfer [5, 14, 15]. Li et al. showed that only re-vitrification at the cleavage stage, not blastocysts, achieved similar pregnancy outcomes [16]. Concerning the safety of re-cryopreservation for both women and their offspring, we followed up the maternal and neonatal outcomes and this aspect had also been investigated by other authors with same results; it appeared that no significant association was observed between cryopreservation times and perinatal complications, congenital anomaly, and neonatal death [5, 14, 17]. However, it is still too early to decide the issues of safety, which remains to be proved by the further and long-term assessment. The higher risk of pregnancy failure in FET with re-vitrified embryos could be mainly due to the elements from cryopreservation procedures, such as vitrification, osmotic shock, and cryoprotectants, which may lead to a disruption of (epi-)genome integrity and transcript stability [9, 16, 18]. The extent of DNA fragmentation and aneuploidy increased because of the usage of embryotoxic agents during vitrification and the freezing damage was suggested cumulatively [4, 16]. Wang et al. reported that the expression of genes involved in human blastocyst implantation were altered after re-cryopreservation and the stress of endoplasmic reticulum (ER) was abnormally activated and ignited the apoptosis‐associated caspase cascade reaction [19]. Consequently, the extra inner cell mass was lost [20] and the function of trophectoderm was impaired with lower secretion of hCG [19]. However, another research verified that re-vitrification influenced the expression of apoptotic genes Bax, Bcl-2 and implantation gene ErbB4 and diminished the expanded blastocyst rate of mouse embryos, only comparing with fresh one, but these differences were not seen between vitrified and re-vitrified embryos [21]. Moving forward, the application of newest multi-omic technologies to the pre-implantation embryo indicated that variants of imprinted gene methylation signatures and histone modification patterns, related to the embryo vitrification and thawing environment exposure, could be transmitted and exert their potential impacts on embryos with unknown consequences [22]. Along with re-vitrification, extended in vitro culture duration of embryos, which is attributed to the two phases of in vitro cultivation process, may also introduce new issues. We noted a significant increase in transfer cycles involving D6/D7 blastocysts in the re-cryopreservation group compared with the cryopreservation group. Prolonged in vitro embryo culture of blastocysts may indicate delayed development associated with cryoinjury and thawing environment exposure. Zheng et al. demonstrated that D6 blastocysts maintained implantation potential comparable to D5 blastocysts in both re-cryopreserved and cryopreserved groups, with only a non-significant decrease in CPR observed for D7 blastocysts [23]. However, another study revealed that when controlling for confounders, such as patient age, blastocyst quality, rather than developmental stage (D5/D6), dominated pregnancy outcomes in FET of cryopreserved embryos [24]. Meanwhile, Liu et al. specifically noted that frozen–thawed embryo transfers of D7 blastocyst with delayed expansion yielded lower LBR compared to D5/D6 transfers [25]. The insights of the mechanism about how repeated cryopreservation affects the embryo implantation potential contributing to uncompromised clinical outcomes are continually updated. To further evaluate the influence on pregnancy outcomes caused by the female age, a subgroup comparison was done regarding CPR. There was no statistically significant difference in CPR between the once- and re-cryopreservation group in women ≥ 35 years of age. On the contrary, CPR in the re-cryopreservation group with the female age < 35 years was decreased. This raised an important question: for young couples, should recycling re-vitrified embryos be prioritized, or is initiating a new ovarian stimulation cycle preferable to avoid costly treatments with low success rates? For women under 35, we recommend thawing a single blastocyst per transfer cycle whenever feasible, prioritizing the minimization of simultaneous multiple embryo thawing—especially when good-quality blastocysts are available. It is essential to thoroughly counsel patients that strategies involving multiple embryo thawing to enhance single-cycle success may incur more risks than benefits. Clinicians should appropriately manage patient expectations, underscore the potential impact of repeated cryopreservation on embryo viability, and thereby reduce the likelihood of generating re-cryopreserved embryos. However, for women over 35 years, no harm was observed in clinical pregnancy attainment following re-vitrified embryo transfer. The decline in fecundity in elder women is mostly attributed to higher oocyte aneuploidy rates and poorer ovarian reserve from increasing age. This phenomenon might be attributed to two inter-related mechanisms. For one, re-cryopreserved embryos generally undergo extended in vitro culture, with only morphologically advanced blastocysts selected for repeated cryopreservation. This natural selection process tends to exclude embryos with developmental abnormalities or low implantation potential, which are more common in older patients. For another, older women often face reduced ovarian reserve, and re-cryopreservation offers additional embryo transfer opportunities. Collectively, these factors likely mitigate the potential risks of repeated cryopreservation, especially among women over 35 years. There are no data about the effect of different embryo qualities on pregnancy outcomes of re-cryopreserved embryos transfer currently. The quality of an embryo represents its development potential in part, commonly assessed by embryo morphology under a microscope. A good-quality embryo transferred per FET cycle is thought to be associated with improved PR, even exhibiting developmental potential comparable to that of a fresh embryo [26, 27]. After stratifying the criteria of morphology, we found that the presence of ≥ 1 embryo with good-quality morphology at the freezing and thawing process of repeated cryopreservation may increase the chance of pregnancy similar to that of single cryopreservation in FET. The more times of cryopreservation performed, the greater the damage caused by freezing and thawing would be [4]. Whereas, embryos with poorer quality seemed to perform worse in preserving pregnancy potential during the repeated cryopreservation, suggesting that a poor-quality embryo may be less viable than a good-quality embryo to defend this invisible damage. If occasionally a surplus of embryos were available for re-vitrification after thawing in a FET cycle, embryo quality is necessary to be considered in the selection of re-vitrified embryos to achieve a better outcome to vitrified embryo. On the other hand, in our cohort, the number of embryos transferred per cycle between two groups was strictly matched, and it is worth noting that we observed a significant decrease in CPR after the transfer of a single re-vitrified embryo, which was mitigated by transferring multiple embryos. A previous study proposed the transfer of two or more embryos could raise the CPR but resulted in higher risk of multiple pregnancy [28]. Therefore, clinicians should weigh the benefits of the transfer of two re-vitrified embryos against the complications involved and inform the patients when providing medical consultation. The limitations of this study should be considered especially in the interpretation of these clinical data. The incidence rate of re-cryopreserved embryo thawing and transfer was inherently low, which constrained our stratified analyses of live birth rate in re-cryopreservation group. The subgroups of female age and number and quality of embryos had a relatively small sample size, suggesting that conclusions should be drawn with caution. Furthermore, due to the retrospective design, inherent risk of bias was inevitable in this study. Patients with recurrent implantation failure and poor ovarian response were more likely to undergo the FET using re-cryopreserved embryos after exhausting all once cyropreserved embryos. We performed PS matching to minimize confounding bias. Despite matching, the number of embryos transferred per cycle remained unequal; therefore, we included this confounder in our multivariable logistic regression model to adjust for its effect as much as possible. Consequently, multi-centered introspective studies with larger sample size will be a proper option to further confirm the findings. Conclusions This study suggests that repeated cryopreservation of embryos is associated with a significantly reduced clinical pregnancy rate, while neonatal outcomes remain unaffected. Clinical decisions for re-cryopreserved embryo transfer should prioritize patient age and embryo quality. To minimize embryo wastage and multiple transfers, we recommend strict regulation of the number of thawed embryos per FET cycle and reserving re-vitrification for good-quality surplus embryos. Notably, women under 35 should be advised against transferring re-cryopreserved embryos due to potential impacts on embryo viability. These findings provide a foundation for developing clinical guidelines to optimize decision-making in FET involving re-cryopreserved embryos, emphasizing personalized approaches based on patient demographics and embryo characteristics. Supplementary Information Supplementary Material 1 Acknowledgements Not applicable. Abbreviations Frozen–thawed embryo transfer Assisted reproduction treatment Pre-implantation genetic testing Clinical pregnancy rate Live birth rate Propensity-score matching Propensity score Odds ratios 95% Confidence intervals Author contributions Zengshu Huang and Chunhui Zhang conceptualized the report. Zengshu Huang collected data and wrote the manuscript. Nan Wang and Huiqing He revised the manuscript and drew tables and figures. Song Li revised ethical documents. Changzhong Li supervised the project and reviewed the manuscript. Funding This work was supported by the Shenzhen Science and Technology Program (Grant No.RCBS20231211090710015), Shenzhen High-level Hospital Construction Fund (YBH2019-260), Shenzhen Sanming Project of Medicine (No.SZSM202211043). Data availability The data that support the findings of this study are available from the corresponding author upon reasonable request. Declarations Ethics approval and consent to participate The study involving human participants were reviewed and approved by the Ethics Committee of Peking University Shenzhen Hospital. All procedures performed in the study were under supervision of the institutional research committee, in accordance with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Consent for publication All authors have read and agreed to the published version of the manuscript. Competing interests The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Zengshu Huang and Chunhui Zhang contributed equally to this work. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.