Pregnancy and obstetric-neonatal outcomes of patients with thin endometrium using three different endometrial preparation protocols in frozen embryo transfer cycles: a historical cohort of 2671 patients.

The study population comprised 2671 women with thin endometrium (endometrial thickness < 8 mm) who fulfilled the inclusion and exclusion criteria. Respectively, they were grouped by different endometrial preparation protocols for frozen embryo transfer: 1908 (71.43%) patients in the AC group, 506 (18.94%) in the GnRH-a + AC group and 257 (9.62%) in the NC group. Table  1 gives an overview of the characteristics of all patients included. There were significant differences ( P  < 0.05) in antral follicle count (AFC), anti-Müllerian hormone (AMH), cause of infertility, maternal age at oocyte retrieval, and FET among the three groups. As for cycle characteristics in Table  1 , gonadotropin dose, number of retrieved oocytes, and MII oocytes significantly differed across the study groups ( P  < 0.05). Similarly, demographic and cycle characteristics of 1104 patients (endometrial thickness < 7 mm) were exhibited in Supplemental Table 1. Table 1 General characteristics and pregnancy outcomes of patients undergoing different endometrial preparation protocols Variable Endometrial preparation protocols P AC ( N  = 1908) GnRH-a + AC ( N  = 506) NC ( N  = 257) Maternal age at oocyte retrieval, years 33 (30, 37) a 34 (31, 37) b 34 (30, 38) b < 0.001 3* BMI, kg/m 2 21.48 (19.92, 23.44) 21.48 (19.98, 23.23) 21.23 (19.80, 23.26) 0.467 3 Baseline FSH, IU/L 7.38 (6.21, 8.89) 7.56 (6.40, 9.21) 7.57 (6.35, 9.06) 0.212 3 AFC 11 (7, 17) a 10 (6, 15) b 9 (6, 13) b < 0.001 3* AMH, ng/ml 3.00 (1.55, 5.39) a 2.56 (1.31, 4.66) b 2.28 (1.29, 3.89) b < 0.001 3* Duration of infertility, years 2 (1, 4) 2 (1, 4) 2 (1, 4) 0.379 3 Infertility type 0.553 1  Primary infertility 882 (46.2%) 222 (43.9%) 113 (44.0%)  Second infertility 1026 (53.8%) 284 (56.1%) 144 (56.0%) Cause of infertility < 0.001 2*  Tubal factor 628 (32.9%) a 195 (38.5%) a 83 (32.3%) a  Male factor 219 (11.5%) a 26 (5.1%) b 40 (15.6%) a  Diminished ovarian reserve 259 (13.6%) a 93 (18.4%) b 42 (16.3%) a, b  Ovulatory disorders 159 (8.3%) a 29 (5.7%) a 0 (0.0%) b  Endometriosis 37 (1.9%) a 31 (6.1%) b 3 (1.2%) a  Unexplained 93 (4.9%) a, b 14 (2.8%) b 17 (6.6%) a  Uterine 139 (7.3%) a 26 (5.1%) a 23 (8.9%) a  Mixed 374 (19.6%) a 92 (18.2%) a 49 (19.1%) a COH protocol 0.105 1  Depot GnRH-a 224 (11.7%) 52 (10.3%) 24 (9.3%)  Long GnRH-a 581 (30.5%) 148 (29.2%) 60 (23.3%)  GnRH antagonist 802 (42.0%) 214 (42.3%) 121 (47.1%)  Other protocols 301 (15.8%) 92 (18.2%) 52 (20.2%) Fertilization method 0.640 1  IVF 1414 (74.1%) 383 (75.7%) 188 (73.2%)  ICSI 422 (22.1%) 100 (19.8%) 56 (21.8%)  IVF + RICSI 72 (3.8%) 23 (4.5%) 13 (5.1%) Gonadotrophin duration, days 10 (9, 11) 10 (9, 11) 9 (8, 11) 0.149 3 Gonadotrophin dose, IU 2400 (1800, 2925) a 2550 (1995, 3150) b 2475 (1950, 3075) a, b 0.008 3* No. of oocytes retrieved 11 (7, 17) a 10 (7, 15) b 10 (6, 15) b 0.001 3* No. of MⅡ oocytes 10 (6, 15) a 9 (5, 13) b 9 (5, 13) a, b 0.007 3* No. of 2PN 7 (4, 10) 7 (4, 9) 6 (4, 10) 0.265 3 Normal fertilization rate 0.73 (0.60, 0.85) 0.75 (0.60, 0.87) 0.75 (0.61, 0.89) 0.167 3 Blastocyst formation rate 0.67 (0.33, 0.86) 0.63 (0.31, 0.86) 0.67 (0.24, 0.86) 0.228 3 Maternal age at FET, years 33 (30, 37) a 35 (32, 38) b 35 (32, 39) b < 0.001 3* Endometrial thickness, mm 7.1 (6.4, 7.5) 7.0 (6.3, 7.5) 7.0 (6.3, 7.5) 0.426 3 No. of embryos transferred 0.170 1  One 1468 (76.9%) 369 (72.9%) 196 (76.3%)  Two 440 (23.1%) 137 (27.1%) 61 (23.7%) Type of embryo transferred 0.487 1  Day 3 (cleavage stage) 530 (27.8%) 150 (29.6%) 79 (30.7%)  Day 5/6 (blastocyst stage) 1378 (72.2%) 356 (70.4%) 178 (69.3%) Pregnancy outcomes  Biochemical pregnancy loss 165 (8.6%) a 35 (6.9%) a, b 10 (3.9%) b 0.020 1*  Clinical pregnancy 691 (36.2%) 178 (35.2%) 87 (33.9%) 0.721 1  Live birth 512 (26.8%) 128 (25.3%) 71 (27.6%) 0.729 1  Multiple pregnancies 74 (3.9%) 20 (4.0%) 15 (5.8%) 0.325 1  Ectopic pregnancy 11 (0.6%) 1 (0.2%) 0 (0.0%) 0.303 2  Miscarriage 176 (25.5%) 49 (27.5%) 16 (18.4%) 0.262 1 The data are presented as median (Q1, Q3) or N (%) BMI  Body mass index,  FSH  Follicle stimulation hormone, AFC  Antral follicle count, AMH  Anti-müllierian hormone, COH  Controlled ovarian hyperstimulation, IVF  In-vitro fertilization, ICSI  Intracytoplasmic sperm injection, RICSI  Rescue ICSI, MII  Metaphase II, PN  Pronucleus, FET  Frozen embryo transfer, AC  Artificial cycle, GnRH-a  Gonadotropin-releasing hormone agonist, NC  Natural cycle * P < 0.05 (without Bonferroni correction) 1 Chi-Square p-value; 2 Fisher Exact p-value; 3 Kruskal-Wallis p-value a, b Statistically significant differences between groups after Bonferroni correction. Groups sharing the same letter are not significantly different General characteristics and pregnancy outcomes of patients undergoing different endometrial preparation protocols The data are presented as median (Q1, Q3) or N (%) BMI  Body mass index,  FSH  Follicle stimulation hormone, AFC  Antral follicle count, AMH  Anti-müllierian hormone, COH  Controlled ovarian hyperstimulation, IVF  In-vitro fertilization, ICSI  Intracytoplasmic sperm injection, RICSI  Rescue ICSI, MII  Metaphase II, PN  Pronucleus, FET  Frozen embryo transfer, AC  Artificial cycle, GnRH-a  Gonadotropin-releasing hormone agonist, NC  Natural cycle * P < 0.05 (without Bonferroni correction) 1 Chi-Square p-value; 2 Fisher Exact p-value; 3 Kruskal-Wallis p-value a, b Statistically significant differences between groups after Bonferroni correction. Groups sharing the same letter are not significantly different Table  1 exhibited the pregnancy outcomes and live birth calculation of patients who underwent different endometrial preparation protocols (endometrial thickness < 8 mm). In the AC group, 691 (36.2%) patients realized clinical pregnancies, and 512 (26.8%) patients achieved live births. Patients in the GnRH-a + AC group reached 178 (35.2%) clinical pregnancies and 128 (25.3%) live births. In the NC group, there were 87 (33.9%) clinical pregnancies and 71 (27.6%) live births. The biochemical pregnancy loss rate in NC group 3.9% (10/257) was significantly lower than that in AC group 8.6% (165/1908) ( P  < 0.05). There were no statistical differences between groups in clinical pregnancy, live birth, miscarriage, multiple pregnancy, and ectopic pregnancy. As the incidence of ectopic pregnancy is so low, the power to detect a small difference was limited. In Table  2 , clinical pregnancy and live birth outcomes were analyzed by multivariate logistic regression model, in which AOR and 95% CI were calculated by adjusting predictors (significance level < 0.2), including maternal age at FET, BMI, cause of infertility, number of retrieved oocytes, infertility type, and duration, etc. Endometrial preparation protocol and pregnancy outcome were no association founded, except for biochemical pregnancy loss in the NC referred to AC (OR [95%CI]: 0.428 [0.223,0.821]; AOR [95%CI]: 0.444 [0.230,0.856]). Table 2 Crude and adjusted odds ratios of pregnancy outcomes Variable AC vs. NC AC vs. GnRH-a + AC NC vs. GnRH-a + AC Crude OR (95% CI) Adjusted OR (95% CI) Crude OR (95% CI) Adjusted OR (95% CI) Crude OR (95% CI) Adjusted OR (95% CI) Biochemical pregnancy loss a 0.428 (0.223 , 0.821) 0.444 (0.230 , 0.856) 0.785 (0.538,1.146) 0.795 (0.541,1.170) 1.835 (0.894,3.769) 1.791 (0.866,3.702) Clinical pregnancy b 0.901 (0.685,1.186) 1.067 (0.795,1.434) 0.956 (0.779,1.173) 1.133 (0.907,1.415) 1.060 (0.773,1.455) 1.061 (0.755,1.492) Live birth b 1.041 (0.778,1.393) 1.264 (0.921,1.735) 0.923 (0.738,1.156) 1.117 (0.876,1.426) 0.887 (0.632,1.245) 0.884 (0.611,1.277) Multiple pregnancy c 1.536 (0.868,2.719) 1.775 (0.941,3.349) 1.020 (0.616,1.688) 1.022 (0.593,1.762) 0.664 (0.334,1.320) 0.576 (0.271,1.224) Ectopic pregnancy d NA NA 0.341 (0.044,2.651) 1.141 (0.682,1.909) NA NA Miscarriage e 0.659 (0.373,1.165) 0.616 (0.342,1.111) 1.111 (0.767,1.610) 0.955 (0.645,1.413) 1.686 (0.894,3.179) 1.549 (0.799,3.002) Referred to AC Referred to AC Referred to NC Odds ratios (ORs) and 95% confidence intervals (CIs) are based on the univariate analysis. Adjusted odds ratios (AORs) and 95% CIs are based on the multiple logistic regression model NA: Insufficient data for statistical analysis AC Artificial cycle, NC Natural cycle, GnRH-a Gonadotropin-releasing hormone agonist, OR Odds ratios, CI Confidence interval, FET Frozen embryo transfer, BMI Body mass index, COH Controlled ovarian hyperstimulation a Adjusted for maternal age at FET, BMI, cause of infertility, number of retrieved oocytes, infertility type, duration of infertility, and controlled ovarian hyperstimulation (COH) protocol b Adjusted for maternal age at FET, BMI, cause of infertility, fertilization method, number of retrieved oocytes, COH protocol, type of embryos transferred, and endometrial thickness c Adjusted for maternal age at FET, BMI, cause of infertility, number of retrieved oocytes, COH protocol, endometrial thickness, and number of embryos transferred d Adjusted for maternal age at FET, BMI, cause of infertility, and number of retrieved oocytes e Adjusted for maternal age at FET, BMI, cause of infertility, fertilization method, number of retrieved oocytes, infertility type, COH protocol, number of embryos transferred, and endometrial thickness Crude and adjusted odds ratios of pregnancy outcomes Odds ratios (ORs) and 95% confidence intervals (CIs) are based on the univariate analysis. Adjusted odds ratios (AORs) and 95% CIs are based on the multiple logistic regression model NA: Insufficient data for statistical analysis AC Artificial cycle, NC Natural cycle, GnRH-a Gonadotropin-releasing hormone agonist, OR Odds ratios, CI Confidence interval, FET Frozen embryo transfer, BMI Body mass index, COH Controlled ovarian hyperstimulation a Adjusted for maternal age at FET, BMI, cause of infertility, number of retrieved oocytes, infertility type, duration of infertility, and controlled ovarian hyperstimulation (COH) protocol b Adjusted for maternal age at FET, BMI, cause of infertility, fertilization method, number of retrieved oocytes, COH protocol, type of embryos transferred, and endometrial thickness c Adjusted for maternal age at FET, BMI, cause of infertility, number of retrieved oocytes, COH protocol, endometrial thickness, and number of embryos transferred d Adjusted for maternal age at FET, BMI, cause of infertility, and number of retrieved oocytes e Adjusted for maternal age at FET, BMI, cause of infertility, fertilization method, number of retrieved oocytes, infertility type, COH protocol, number of embryos transferred, and endometrial thickness In patients with endometrial thickness < 7 mm, there was no significant difference in pregnancy outcomes and live birth rate among the three groups (Supplemental Tables 1–2). Demographic characteristics of patients with singleton live birth were detailed in Table  3 (endometrial thickness < 8 mm). A total of 652 patients were included, with 473 (72.55%) in the AC group, 117 (17.94%) in the GnRH-a + AC group, and 62 (9.51%) in the NC group. AMH level was prominently different ( P  < 0.05) in the NC group compared to the AC group. For AC, GnRH-a + AC, and NC groups, the median birth weights were 3.25 kg, 3.20 kg, and 3.25 kg, and the median gestational ages were 38.7 weeks, 38.6 weeks, and 38.6 weeks. Singleton obstetric-neonatal outcomes of patients who underwent disparate endometrial preparation protocols were summarized in Table  3 (endometrial thickness < 8 mm). There was no significant difference in cesarean delivery, low birth weight, macrosomia, small for gestational age, large for gestational age, preterm birth, gestational diabetes mellitus, hypertensive disorder of pregnancy, and fetal malformation in comparison between groups. Table 3 General characteristics and obstetric-neonatal outcomes of patients with singleton live births Variable Endometrial preparation protocols P AC ( N  = 473) GnRH-a + AC ( N  = 117) NC ( N  = 62) Maternal age at oocyte retrieval, years 31 (29, 34) 32 (30, 35) 32 (29, 36) 0.248 3 BMI, kg/m 2 21.32 (19.68, 23.44) 20.83 (19.65, 23.19) 21.14 (20.28, 24.08) 0.371 3 Baseline FSH, IU/L 7.23 (6.05, 8.52) 7.32 (6.28, 8.29) 7.39 (5.99, 8.39) 0.627 3 AFC 12 (8, 19) 10 (7, 18) 7 (10, 16) 0.101 3 AMH, ng/ml 3.41 (2.01, 5.86) a 2.88 (1.65, 5.18) a, b 2.29 (1.45, 4.25) b 0.005 3* Duration of infertility, years 2 (1, 4) 2 (1, 4) 2 (1.5, 4) 0.361 3  Infertility type 0.037 1*  Primary infertility 236 (49.9%) a, b 46 (39.3%) b 36 (58.1%) a  Second infertility 237 (50.1%) a, b 71 (60.7%) b 26 (41.9%) a Cause of infertility < 0.001 2*  Tubal factor 153 (32.3%) a 44 (37.6%) a 25 (37.3%) a  Male factor 74 (15.6%) a 7 (6.0%) b 12 (19.4%) a  Diminished ovarian reserve 32 (6.8%) a 11 (9.4%) a 9 (14.5%) a  Ovulatory disorders 61 (12.9%) a 10 (8.5%) a 0 (0.0%) b  Endometriosis 7 (1.5%) a 10 (8.5%) b 0 (0.0%) a, b  Unexplained 33 (7.0%) a 4 (3.4%) a 7 (11.3%) a  Uterine 27 (5.7%) a 7 (6.0%) a 3 (4.8%) a  Mixed 86 (18.2%) a 24 (20.5%) a 6 (9.7%) a COH protocol 0.261 1  Depot GnRH-a 52 (11.0%) 16 (13.7%) 8 (12.9%)  Long GnRH-a 165 (34.9%) 39 (33.3%) 12 (19.4%)  GnRH antagonist 207 (43.8%) 52 (44.4%) 32 (51.6%)  Other protocols 49 (10.4%) 10 (8.5%) 10 (16.1%) Fertilization method 0.789 2  IVF 344 (72.7%) 85 (72.6%) 45 (72.6%)  ICSI 118 (24.9%) 27 (23.1%) 16 (25.8%)  IVF + RICSI 11 (2.3%) 5 (4.3%) 1 (1.6%) Gonadotrophin duration, days 10 (9, 11) 10 (9, 11) 10 (8, 10) 0.093 3 Gonadotrophin dose, IU 2250 (1650, 2785) a 2550 (1950, 3135) b 2300 (1850, 2700) a, b 0.046 3* No. of oocytes retrieved 14 (8, 19) 11 (8, 16) 11 (7, 18) 0.103 3 No. of MⅡ oocytes 11 (7, 17) 9 (6, 15) 6 (10, 16) 0.055 3 No. of 2PN 8 (5, 12) 7 (5, 10) 5 (7, 11) 0.279 3 Normal fertilization rate 0.75 (0.62, 0.85) 0.77 (0.67, 0.88) 0.75 (0.60, 0.85) 0.556 3 Blastocyst formation rate 0.73 (0.50, 0.89) 0.71 (0.35, 0.86) 0.75 (0.49, 1.00) 0.538 3 Maternal age at FET, years 32 (29, 35) 33 (30, 35) 33 (30, 37) 0.079 3 Endometrial thickness, mm 7.2 (6.7, 7.6) 7.3 (6.8, 7.6) 7.2 (6.5, 7.6) 0.897 3 No. of embryos transferred 0.357 1  One 398 (84.1%) 92 (78.6%) 52 (83.9%)  Two 75 (15.9%) 25 (21.4%) 10 (16.1%) Type of embryo transferred 0.181 1  Day 3 (cleavage stage) 63 (13.3%) 19 (16.2%) 4 (6.5%)  Day 5/6 (blastocyst stage) 410 (86.7%) 98 (83.8%) 58 (93.5%) Obstetric-neonatal outcomes Delivery mode 0.064 1  Cesarean delivery 414 (87.5%) 102 (87.2%) 48 (77.4%)  Natural labor 59 (12.5%) 15 (12.8%) 14 (22.6%) Gender 0.624 1  Male 240 (50.7%) 65 (55.6%) 34 (54.8%)  Female 233 (49.3%) 52 (44.4%) 28 (45.2%) Gestational age, weeks 38.7 (37.9, 39.3) 38.6 (37.9, 39.1) 38.6 (37.9, 39.1) 0.991 3 Birth weight, kg 3.25 (3.00, 3.54) 3.20 (2.94, 3.50) 3.25 (3.00, 3.60) 0.608 3 Low birth weight  4, 000 g 23 (4.9%) 5 (4.3%) 1 (1.6%) 0.521 2 Small for gestational age 21 (4.4%) 7 (6.0%) 1 (1.6%) 0.421 2 Large for gestational age 73 (15.4%) 12 (10.3%) 6 (9.7%) 0.208 1 Preterm birth 63 (13.3%) 15 (12.8%) 10 (16.1%) 0.813 1 Gestational diabetes mellitus 14 (3.0%) 3 (2.6%) 1 (1.6%) 0.872 2 Hypertensive disorders of pregnancy 17 (3.6%) 5 (4.3%) 5 (8.1%) 0.251 2 Placenta previa 12 (2.5%) 4 (3.4%) 1 (1.6%) 0.730 2 Fetal malformation 2 (0.4%) 2 (1.7%) 0 (0.0%) 0.197 2 AC Artificial cycle, GnRH-a Gonadotropin-releasing hormone agonist, NC Natural cycle, FET Frozen embryo transfer, BMI Body mass index, FSH Follicle stimulation hormone, AFC Antral follicle count, AMH Anti-müllierian hormone, COH Controlled ovarian hyperstimulation, PN  Pronucleus * P < 0.05 (without Bonferroni correction) 1 Chi-Square p-value; 2 Fisher Exact p-value; 3 Kruskal-Wallis p-value a, b Statistically significant differences between groups after Bonferroni correction. Groups sharing the same letter are not significantly different General characteristics and obstetric-neonatal outcomes of patients with singleton live births AC Artificial cycle, GnRH-a Gonadotropin-releasing hormone agonist, NC Natural cycle, FET Frozen embryo transfer, BMI Body mass index, FSH Follicle stimulation hormone, AFC Antral follicle count, AMH Anti-müllierian hormone, COH Controlled ovarian hyperstimulation, PN  Pronucleus * P < 0.05 (without Bonferroni correction) 1 Chi-Square p-value; 2 Fisher Exact p-value; 3 Kruskal-Wallis p-value a, b Statistically significant differences between groups after Bonferroni correction. Groups sharing the same letter are not significantly different Using the method previously stated, adjusted odds ratios and 95%CI were calculated to analyze the association between FET endometrial preparation protocols and singleton obstetric-neonatal outcomes. All the outcome variables were still retained. No significant difference was found after univariate and multivariate regression analyses (Table  4 ). In Supplemental Tables 3–4, the same study was conducted in patients with endometrial thickness < 7 mm, showing no significant difference in obstetric-neonatal outcomes across the three groups. Table 4 Crude and adjusted odds ratios of obstetric-neonatal outcomes Variable AC vs. NC AC vs. GnRH-a + AC NC vs. GnRH-a + AC Crude OR (95% CI) Adjusted OR (95% CI) Crude OR (95% CI) Adjusted OR (95% CI) Crude OR (95% CI) Adjusted OR (95% CI) Low birth weight a 0.747 (0.258, 2.163) 0.775 (0.259, 2.316) 0.689 (0.300, 1.580) 0.594 (0.250, 1.410) 0.923 (0.259, 3.282) 0.766 (0.206, 2.856) Macrosomia b 0.321 (0.043, 2.418) 0.284 (0.036, 2.254) 0.873 (0.325, 2.348) 0.880 (0.311, 2.486) 2.723 (0.311, 23.840) 3.102 (0.332, 28.984) SGA c 0.353 (0.047, 2.670) 0.365 (0.047, 2.813) 1.370 (0.568, 3.304) 1.222 (0.494, 3.021) 3.882 (0.467, 32.292) 3.344 (0.391, 28.611) LGA d 0.587 (0.244, 1.413) 0.608 (0.247, 1.499) 0.626 (0.328, 1.196) 0.661 (0.336, 1.304) 1.067 (0.380, 2.995) 1.087 (0.371, 3.184) Preterm birth e 1.242 (0.601, 2.570) 1.293 (0.607, 2.757) 0.950 (0.520, 1.737) 0.797 (0.428, 1.485) 0.765 (0.321, 1.820) 0.616 (0.249, 1.527) GDM f 0.534 (0.069, 4.132) 0.621 (0.077, 4.995) 0.857 (0.242, 3.033) 0.757 (0.203, 2.828) 1.605 (0.163, 15.764) 1.220 (0.115, 12.970) HDP c 2.337 (0.831, 6.577) 2.359 (0.798, 6.974) 1.190 (0.430, 3.294) 1.151 (0.410, 3.231) 0.509 (0.142, 1.830) 0.488 (0.129, 1.843) Placenta previa b 0.626 (0.080, 4.896) 0.589 (0.073, 4.752) 1.351 (0.428, 4.268) 1.547 (0.468, 5.115) 2.159 (0.236, 19.749) 2.629 (0.270, 25.626) Fetal malformation b NA NA 4.070 (0.567, 29.198) 5.443 (0.668, 44.347) NA NA Referred to AC Referred to AC Referred to NC Odds ratios (ORs) and 95% confidence intervals (CIs) are based on the univariate analysis. Adjusted odds ratios (AORs) and 95% CIs are based on the multiple logistic regression model NA: Insufficient data for statistical analysis AC Artificial cycle, NC Natural cycle, GnRH-a Gonadotropin-releasing hormone agonist, OR Odds ratios, CI Confidence interval, SGA Small for gestational age, LGA Large for gestational age, GDM Gestational diabetes mellitus, HDP Hypertensive disorders of pregnancy, FET Frozen embryo transfer, BMI Body mass index, COH Controlled ovarian hyperstimulation a Adjusted for maternal age at FET, BMI, cause of infertility, duration of infertility, and number of retrieved oocytes b Adjusted for maternal age at FET, BMI, and cause of infertility c Adjusted for maternal age at FET, BMI, cause of infertility, and number of retrieved oocytes d Adjusted for maternal age at FET, BMI, cause of infertility, COH protocol, number of retrieved oocytes, and number of embryo transferred e Adjusted for maternal age at FET, BMI, cause of infertility, infertility type, and number of retrieved oocytes f Adjusted for maternal age at FET, BMI, cause of infertility, and number of embryo transferred Crude and adjusted odds ratios of obstetric-neonatal outcomes Low birth weight a 0.747 (0.258, 2.163) 0.775 (0.259, 2.316) 0.689 (0.300, 1.580) 0.594 (0.250, 1.410) 0.923 (0.259, 3.282) 0.766 (0.206, 2.856) 0.321 (0.043, 2.418) 0.284 (0.036, 2.254) 0.873 (0.325, 2.348) 0.880 (0.311, 2.486) 2.723 (0.311, 23.840) 3.102 (0.332, 28.984) 0.353 (0.047, 2.670) 0.365 (0.047, 2.813) 1.370 (0.568, 3.304) 1.222 (0.494, 3.021) 3.882 (0.467, 32.292) 3.344 (0.391, 28.611) 0.587 (0.244, 1.413) 0.608 (0.247, 1.499) 0.626 (0.328, 1.196) 0.661 (0.336, 1.304) 1.067 (0.380, 2.995) 1.087 (0.371, 3.184) 1.242 (0.601, 2.570) 1.293 (0.607, 2.757) 0.950 (0.520, 1.737) 0.797 (0.428, 1.485) 0.765 (0.321, 1.820) 0.616 (0.249, 1.527) 0.534 (0.069, 4.132) 0.621 (0.077, 4.995) 0.857 (0.242, 3.033) 0.757 (0.203, 2.828) 1.605 (0.163, 15.764) 1.220 (0.115, 12.970) 2.337 (0.831, 6.577) 2.359 (0.798, 6.974) 1.190 (0.430, 3.294) 1.151 (0.410, 3.231) 0.509 (0.142, 1.830) 0.488 (0.129, 1.843) Placenta previa b 0.626 (0.080, 4.896) 0.589 (0.073, 4.752) 1.351 (0.428, 4.268) 1.547 (0.468, 5.115) 2.159 (0.236, 19.749) 2.629 (0.270, 25.626) Fetal malformation b 4.070 (0.567, 29.198) 5.443 (0.668, 44.347) Odds ratios (ORs) and 95% confidence intervals (CIs) are based on the univariate analysis. Adjusted odds ratios (AORs) and 95% CIs are based on the multiple logistic regression model NA: Insufficient data for statistical analysis AC Artificial cycle, NC Natural cycle, GnRH-a Gonadotropin-releasing hormone agonist, OR Odds ratios, CI Confidence interval, SGA Small for gestational age, LGA Large for gestational age, GDM Gestational diabetes mellitus, HDP Hypertensive disorders of pregnancy, FET Frozen embryo transfer, BMI Body mass index, COH Controlled ovarian hyperstimulation a Adjusted for maternal age at FET, BMI, cause of infertility, duration of infertility, and number of retrieved oocytes b Adjusted for maternal age at FET, BMI, and cause of infertility c Adjusted for maternal age at FET, BMI, cause of infertility, and number of retrieved oocytes d Adjusted for maternal age at FET, BMI, cause of infertility, COH protocol, number of retrieved oocytes, and number of embryo transferred e Adjusted for maternal age at FET, BMI, cause of infertility, infertility type, and number of retrieved oocytes f Adjusted for maternal age at FET, BMI, cause of infertility, and number of embryo transferred Throughout this analysis, the NC group contained both true natural cycles and modified natural cycles, which were different in HCG trigger and luteal phase support. To minimize the potential confounding effects of such differences, we categorized the true natural cycle and the modified natural cycle into distinct groups and compared both baseline characteristics and pregnancy as well as obstetric-neonatal outcomes. The results are summarized in Supplementary Tables 5–6. It revealed that the pregnancy rate and live birth rate were comparable, and no significant differences were observed in obstetric-neonatal complications.

This retrospective cohort study was conducted among women with a thin endometrium (defined as an endometrial thickness < 8 mm on the first day of progesterone administration) who underwent their first FET at the Reproductive Medicine Center of a university hospital between January 2018 and August 2022. Follow-up for all patients continued until August 2023. Exclusion criteria included a history of recurrent spontaneous abortion, the presence of uterine abnormalities (such as uterine malformations, severe adenomyosis, untreated submucosal fibroids, endometrial polyps, or uterine effusion), oocyte donation, other endometrial preparation protocols, canceled cycles, chronic hypertension or diabetes, endometrial thickness ≥ 8 mm on the trigger day during controlled ovarian hyperstimulation, and incomplete data (e.g., missing values for endometrial thickness, number of embryo transfers, or type of embryo transfers). A flowchart of this study is presented in Fig.  1 . Fig. 1 Flowchart of this study. Notes: Inclusion and exclusion criteria, statistical analysis methods, and outcome indicators. FET = frozen embryo transfer; PGT = pre-implantation genetic testing; AC = artificial cycle; NC = natural cycle; GnRH-a = gonadotropin-releasing hormone agonist Flowchart of this study. Notes: Inclusion and exclusion criteria, statistical analysis methods, and outcome indicators. FET = frozen embryo transfer; PGT = pre-implantation genetic testing; AC = artificial cycle; NC = natural cycle; GnRH-a = gonadotropin-releasing hormone agonist The study protocol was approved by the Declaration of Helsinki by the Institutional Review Board (IRB) of Tongji Hospital, affiliated with Tongji Medical College, Huazhong University of Science and Technology (approval number TJ-IRB20230921). Patients’ informed consent was waived. The study compared three common endometrial preparation protocols for frozen embryo transfer: the AC group, the AC group with GnRH agonist pretreatment (GnRH-a + AC), and the NC group [ 41 , 42 ]. These protocols are clearly illustrated in Fig. 2 . In the AC group, estradiol (Progynova, Bayer Schering Pharma AG, Germany) was administered orally, starting with 4 mg/day from day 2 to 5, increasing to 6 mg/day from day 6 to 9, and then to 8 mg/day from day 10 to 13. Endometrial thickness and serum progesterone levels were monitored on day 14 of the menstrual cycle. Once the endometrial thickness reached the desired level and serum progesterone was below 1.2 ng/ml, the embryo was transferred following daily progesterone administration for three or five consecutive days. Luteal phase support consisted of 20 mg of orally administered progesterone (Abbott Biologicals B.V., Netherlands) in conjunction with a 90 mg vaginal gel (Merck Serono Limited, Germany) until 8 to 10 weeks of pregnancy. In the GnRH-a + AC group, patients received a single 3.75 mg injection of a GnRH agonist (triptorelin or leuprorelin) on day 2 of their menstrual cycle. After a 28-day suppression period, they commenced the same AC protocol as described, followed by embryo transfer. The NC group monitored endometrial thickness, follicular development, and serum hormone levels from day 10 to 12 of the menstrual cycle until the dominant follicle measured 18 mm or detected an LH surge. Ovulation may occur naturally in the true natural cycle or be induced by the injection of human chorionic gonadotropin (HCG; Ovitrelle 250 µg, Merck, Rahway, NJ, USA) in the modified natural cycle. Additionally, all patients were routinely administered low-dose aspirin (Bayer, 100 mg/day) orally, beginning at the onset of each endometrial preparation protocol. Fig. 2 Timeline of the three endometrial preparation protocols. Notes: In the AC group, estradiol valerate was administrated, endometrial thickness was monitored, and estradiol dosage increased gradually. In the GnRH-a + AC group, patients received a single 3.75 mg injection of a GnRH agonist on day 2 of the menstrual cycle. In the NC group, endometrial thickness, follicular development, and serum hormone measurements were monitored starting on days 10–12 of the menstrual cycle. P = progesterone; AC = artificial cycle; NC = natural cycle; GnRH-a = gonadotropin-releasing hormone agonist; D3 = cleavage stage embryo; D5/6 = blastocyst; HCG = human chorionic gonadotropin Timeline of the three endometrial preparation protocols. Notes: In the AC group, estradiol valerate was administrated, endometrial thickness was monitored, and estradiol dosage increased gradually. In the GnRH-a + AC group, patients received a single 3.75 mg injection of a GnRH agonist on day 2 of the menstrual cycle. In the NC group, endometrial thickness, follicular development, and serum hormone measurements were monitored starting on days 10–12 of the menstrual cycle. P = progesterone; AC = artificial cycle; NC = natural cycle; GnRH-a = gonadotropin-releasing hormone agonist; D3 = cleavage stage embryo; D5/6 = blastocyst; HCG = human chorionic gonadotropin Embryos were vitrified using standard protocols and subsequently stored in liquid nitrogen. On the transfer day, the embryos were thawed and placed in a recovery medium at 37℃ for 2 to 3 h before assessing their viability. Only those embryos or blastocysts deemed viable were selected for transfer. The embryo transfer procedure was conducted under ultrasound guidance using a soft catheter, irrespective of the endometrial preparation protocol. Typically, a single high-quality blastocyst was transferred into the uterine cavity, positioned approximately 1.5 to 2 cm from the uterine fundus to enhance implantation potential. In our center, the strategy of elective single embryo transfer was advocated for most patients to mitigate complications associated with multiple pregnancies. However, in cases of thin endometrium, where clinical pregnancy rates are lower and miscarriage rates are higher compared to non-thin endometrium cases, no more than two embryos or blastocysts were transferred. This study evaluated both pregnancy and obstetric-neonatal outcomes. Data on pregnancy outcomes were collected from medical records, including clinical pregnancy rate (CPR, confirmed by the presence of a gestational sac on ultrasound), biochemical pregnancy loss rate (determined by a positive serum β-hCG test, but did not result in a clinical pregnancy), live birth rate (LBR, defined as the birth of a live neonate beyond 24 weeks gestation), multiple pregnancy rate (the presence of two or more gestational sacs), and the ectopic pregnancy rate (where the embryo implanted outside the uterus, diagnosed via ultrasound or clinical evaluation). The primary outcome was the LBR. To minimize bias from vanishing twin pregnancies, the analysis included only live births from singleton pregnancies for obstetric-neonatal outcomes. These outcomes included low birth weight (LBW, birth weight  4, 000 g), small-for-gestational-age (SGA, birth weight below the 10th percentile for gestational age), large-for-gestational-age (LGA, birth weight above the 90th percentile), preterm birth (birth before 37 completed weeks of gestation), gestational diabetes mellitus (diagnosed according to the criteria of the International Association of Diabetes and Pregnancy Study Groups), hypertensive disorders of pregnancy (HDP, which includes pregnancy-induced hypertension and preeclampsia), placenta previa (placenta implanted over or near the inner opening of the cervix, diagnosed by ultrasound) and fetal malformation (any structural or functional abnormality identified at birth). All statistical analyses were performed using SPSS 27.0. Categorical variables were compared pairwise among different endometrial preparation protocols (AC, GnRH-a + AC, and NC) using the Chi-squared test or Fisher’s exact test, as appropriate. Continuous variables were assessed for normality using the Shapiro-Wilk test. As none of the continuous variables followed a normal distribution, they were summarized as medians with interquartile ranges (Q1, Q3), and the Kruskal-Wallis test was used for pairwise comparisons. To assess the associations between endometrial preparation protocols and pregnancy and obstetric-neonatal outcomes, logistic regression models were used to estimate crude odds ratios (OR) and adjusted odds ratios (AOR) with corresponding 95% confidence intervals (CIs). Univariate logistic regressions were initially conducted between each outcome and all covariates to identify the factors influencing outcomes. Statistically significant covariates with P < 0.2 were considered in subsequent multivariate logistic regressions, accounting for the exposure of interest and clinically relevant variables. Age, body mass index (BMI), infertility type, and endometrial preparation protocols were included as fixed variables in the multivariate logistic regression analyses. Given the multiple comparisons conducted across different clinical and obstetric-neonatal outcomes, Bonferroni correction was applied to adjust for the increased risk of type I errors due to multiple hypothesis testing. Both crude and adjusted ORs were reported alongside their corresponding 95% CIs to assess the strength of the associations. Statistical significance was determined with two-sided P values, and a threshold of P < 0.05 was considered statistically significant for all tests. The sample size was calculated using R based on differences in the incidence of preeclampsia (3.7% versus 11.8%) and HDP (2.8% versus 11.4%) between the NC and AC groups [ 43 ]. To provide a two-tailed significance level of 0.05 and a power of 80%, 116 participants were required in the natural cycle group.

To our knowledge, this is the first research to compare the obstetric-neonatal outcomes among patients with thin endometrium based on different endometrial preparation regimens: natural cycle versus artificial cycle with or without GnRH-a pretreatment. According to our findings, whether endometrial thickness < 7–8 mm, the three strategies were comparable in terms of pregnancy and obstetric-neonatal outcomes after FET. But the biochemical pregnancy loss was significantly lower in NC than that in AC in patients with endometrial thickness < 8 mm (3.9% versus 8.5%; AOR [95%CI]: 0.444 [0.230, 0.856]). However, the statistical power for these comparisons was limited due to the low incidence of obstetric-neonatal complications in each group, possibly reducing the ability to detect a true difference. Managing patients with thin endometrium remains an inevitable obstacle to IVF. Many promising treatment regimens to enhance endometrial thickness have recently focused on FET cycles, including growth hormone, intrauterine autologous platelet-rich plasma infusion, tamoxifen, and granulocyte colony-stimulating factor [ 44 – 46 ]. However, these controversial approaches for managing thin endometrium still need to be clarified. This investigation of thin endometrium found no statistical differences in the incidence of obstetric-neonatal complications among endometrial preparation protocols, which have not been reported before. In the general population, however, it remains uncertain whether the more physiological hormonal environment of NC reduces such complications compared with AC [ 11 , 15 , 47 – 49 ]. Based on data from a single-center randomized controlled trial of 1428 women, no differences were found among NC, modified NC, and AC, comparable to other literature results [ 15 ]. In contrast, a recent systematic review and meta-analysis involving 30 studies and 113,676 patients (NC: N = 56, 445; AC: N = 57, 231) concluded that singleton deliveries from NC decreased the risk of adverse obstetric-neonatal outcomes compared to AC [ 11 ]. The corpora lutea in NC has been estimated to be associated with safer outcomes in FET [ 50 – 52 ]. However, this does not seem to be true regarding thin endometrium. In this present study of thin endometrium, obstetric-neonatal outcomes were not influenced by the different endometrial preparation protocols. This may be because a thin endometrium itself can lead to various obstetric and neonatal complications, including SGA, LBW, HDP, placental disease, and preterm birth [ 32 – 35 ]. These risks are difficult to mitigate through endometrial preparation protocols alone. A cohort study of 1057 deliveries following IVF examined placental pathology [ 35 ]. It revealed a reduced placental thickness, a higher rate of placental abruption, and alterations in the frequency of maternal hypoperfusion in the thin endometrium group. Studies have shown that fresh embryo transfer with more corpora lutea has a lower risk of HDP, suggesting that the number of corpora lutea may mediate placental perfusion [ 9 , 10 ]. Therefore, more intensive interventions may be necessary to enhance obstetric outcomes in patients with thin endometrium. Our team is currently conducting a study on the impact of the number of corpora lutea on the outcome of FET, and the results are worth attention. Another reason for this inconsistency was that obstetric-neonatal complications have decreased in FET cycles due to advancements in IVF technology. The variation between study populations, embryo culture medium, cryoprotectant choice, and slow freezing versus vitrification may contribute to substantial inter-study heterogeneity. Regarding pregnancy outcomes, including clinical pregnancy and live birth rate, previous studies have explored which endometrial preparation protocol could create a more receptive uterine environment to embryo implantation. Similar to our research, numerous studies have found no superiority in which treatment regimen provides the best pregnancy rates and clinical outcomes in general populations [ 12 , 13 , 15 ]. Our center’s three most common endometrial preparation strategies (NC, AC, and GnRH-a + AC) failed to offer a satisfactory solution for the challenging problem of thin endometrium. However, three studies have been conducted on the relative effectiveness of stimulated protocols in women with thin endometrium, which seemed promising. In a recent historical cohort study of thin endometrium, 129 women received the AC regimen, and 249 received the ovarian induction following GnRH agonist pretreatment [ 37 ]. The investigators found that clinical pregnancy and live birth rates in FET cycles were significantly higher in the stimulated cycles, similar to a case report [ 38 ]. Another case-control study involved 40 women for the stimulated protocol and 40 age-matched women for the AC protocol in patients with thin endometrium. The clinical pregnancy rate was 35% and 12.5%, respectively ( P = 0.017) [ 39 ]. These studies showed the beneficial effects of stimulated cycles and deserved further study. In this study, the biochemical pregnancy loss rate was lower in NC (3.9%) than that in AC (8.6%) and GnRH-a + AC (6.9%) ( P = 0.02). The risk of pregnancy loss was also lower in NC but not significant (NC: 18.4%; AC: 25.5%; GnRH-a + AC: 27.5%). As indicated by previous research, ovulatory disorders, mainly caused by PCOS, was associated with biochemical pregnancy loss [ 53 ]. In this research, we performed a multivariate logistic regression analysis, adjusting for the relevant covariates including the cause of infertility. The difference in biochemical pregnancy loss rates between NC and AC remained statistically significant. Therefore, for patients with thin endometrium and normal ovulatory function, using the natural cycle for FET may help reduce the rate of biochemical pregnancy loss. Interestingly, we found a higher biochemical pregnancy loss rate but a similar live birth rate in the AC group than the NC group. A multicenter historical cohort study, involving AC ( N = 8139) and NC ( N = 3126), concluded that AC was associated with higher early pregnancy loss rates (AC: 36.5%; NC: 25.6%; AOR [95% CI]: 1.63 [1.35, 1.97]) and lower live birth rates, but biochemical pregnancy loss rates were similar [ 54 ]. A retrospective study showed significantly higher early pregnancy loss rates (34.2% versus 56.9%) and lower live birth rates (59.7% versus 29.1%) in NC compared to stimulated cycles [ 55 ]. Another retrospective study involving 4470 cycles found comparable clinical pregnancy and live birth rates across all protocols, though AC exhibited higher biochemical pregnancy loss rates [ 56 ]. In contrast, a randomised controlled trial [ 13 ] involving women with normal ovulatory function revealed no significant differences in live birth or pregnancy loss rates between NC, modified NC and AC. These inconsistencies may stem from multiple factors: firstly, differences in study design, where RCTs control for bias through randomisation but may lack generalisability, while observational studies better reflect real-world settings yet remain susceptible to confounding factors; secondly, variations in outcome definitions and measurement timepoints, such as distinguishing early pregnancy loss from biochemical pregnancy loss; thirdly, variations in protocol execution details, such as drug dosages or monitoring frequencies; fourthly, the predominantly low quality of evidence, as noted in systematic reviews [ 14 ], necessitates further high-quality RCTs to clarify these findings. Besides, we provided separate analyses of pregnancy and obstetric-neonatal outcomes from AC with or without GnRH-a. As the AC is currently the most common method of endometrial preparation, finding out populations benefiting from pretreatment with 3.75 mg of GnRH-a was clinically valuable. Molecular mechanism studies have shown that high-dose, long-acting GnRH-a benefits the expression of endometrial receptivity indicators, such as HOXA10, MEIS1, and LIF, improving endometrial receptivity [ 57 ]. Other studies found that GnRH-a enhanced the number of pinocytosis and the expression of integrins [ 58 , 59 ]. Therefore, it was reported that the depot GnRH-a protocol with downregulation of 3.75 mg GnRH-a could improve the outcome of embryo transfer in the general population [ 57 ] as well as in patients with a thin endometrium [ 60 ]. Besides, long-acting GnRH-a can effectively prevent the subsequent occurrence of a dominant follicle in AC. However, our study found that thin endometrium does not seem to require GnRH-a pretreatment, which did not improve the clinical pregnancy and live birth rate, but neither did the pregnancy loss nor obstetric-neonatal complications. Moreover, additional GnRH-a treatment will delay the time for patients to obtain live births and increase the financial burden on patients. One possible explanation may be that the GnRH-a pretreatment in our trial was only one month. Only one small-sample observation historical study has been conducted on women with thin endometrium, and the results showed a promotion in clinical outcomes when treating patients for two months [ 61 ]. Our research has some limitations. The primary constraints of this study stem from its single-centered retrospective historical design and inherent selection bias. However, we included a relatively large sample size and adjusted for confounding variables through multivariate logistic regression to enhance the reliability of our results. Nevertheless, some confounding variables that were overlooked exist inevitably. In addition, the generalizability of our findings may be restricted to the exclusion of stimulated cycles in this study, especially given their potential benefits for thin endometrium, as referenced in prior studies. The small cohort of 40 patients in stimulated cycles with thin endometrium was inadequate for inclusion in our analysis. In our center, natural cycles and artificial cycles with or without GnRH agonist are the most common protocols. The incidence of thin endometrium is low. In order to ensure sufficient enrollment and statistical significance, we screened 2671 patients who met the inclusion criteria from 29,338 FET cycles. As these three common endometrial preparation regimens had no priority in patients with thin endometrium, our team is conducting a study on ovulation induction to induce multiple corpora lutea in the hope of bringing significant benefits.

The natural cycle might decrease the biochemical pregnancy loss, referred to as the AC cycle for patients with endometrial thickness < 8 mm. The live birth rate and obstetric outcomes were not affected beneficially or detrimentally by endometrial preparation strategies in natural cycles, artificial cycles, or GnRH-a artificial cycle for thin endometrium (endometrial thickness < 7–8 mm on the first day of progesterone administration) women undergoing FET.

In vitro fertilization (IVF) is the mainstay of contemporary infertility therapy. More than 2 million treatment cycles are carried out globally every year [ 1 – 3 ], with a notable increase in frozen embryo transfer (FET) cycles attributed to the advancements in vitrification techniques [ 4 – 7 ]. However, studies show that FET is linked to higher rates of hypertensive disorders of pregnancy (HDP) and preeclampsia compared to fresh embryo transfers. This increased risk may be due to differences in the number of corpora lutea present during the cycle [ 8 – 10 ]. Therefore, it is important to explore ways to improve pregnancy and obstetric-neonatal outcomes after FET. One key approach is investigating different endometrial preparation protocols used before embryo transfer. The various protocols for endometrial preparation remain a topic of debate within the general population [ 11 – 15 ] and in specific groups, such as those with polycystic ovary syndrome (PCOS) [ 16 – 18 ] or endometriosis [ 19 , 20 ]. A key aspect of this debate is the trade-off between convenience and safety. The popularity of the artificial cycle (AC) protocol may be due to its convenience in scheduling embryo transfer while achieving similar pregnancy outcomes to the natural cycle (NC). Accumulating evidence reported that artificial cycles may be associated with increased maternal risks, particularly HDP and preeclampsia [ 21 , 22 ]. In addition, recent studies suggest that the use of natural cycles was associated with reduced risks of miscarriage, as well as increased rates of clinical pregnancy and live birth [ 23 – 25 ]. Evaluating endometrial preparation protocols is crucial for patients with thin endometrium. As insufficient thickness leads to poorer fresh transfer outcomes [ 26 ], a freeze-all strategy is often adopted [ 27 ]. Consequently, FET is often used for individuals with thin endometrium, a known independent predictor of pregnancy outcomes that frequently causes implantation failure [ 28 – 31 ]. In addition to reduced pregnancy rates and increased miscarriage risks, it is also associated with higher rates of obstetric-neonatal complications [ 32 – 35 ]. Guidelines therefore recommend evaluating the impact of endometrial preparation protocols on pregnancy outcomes in these patients [ 36 ]. However, relevant literature remains limited, with no reports on obstetric-neonatal complications. Since numerous studies in general populations have shown that NC decreases the risk of these complications compared to AC [ 11 , 12 ], it is worth investigating whether this benefit extends to patients with thin endometrium. Some studies indicated the positive effectiveness of stimulated protocols in pregnancy outcomes with a thin endometrium [ 37 – 39 ], not to mention the comparison of NC and AC. While the AC with downregulation of gonadotropin-releasing hormone agonist (GnRH-a) improved the clinical outcomes of embryo transfer [ 40 ], its impact on FET with thin endometrium is unclear. The present study aimed to assess the impact of various endometrial preparation strategies on outcomes in women with thin endometrium. We conducted a historical cohort study to evaluate the pregnancy and obstetric-neonatal outcomes among the NC, AC, and GnRH-a + AC groups of patients with thin endometrium.

Supplementary Material 1. Supplemental Table 1. General characteristics and pregnancy outcomes of patients undergoing different endometrial preparation protocols (endometrial thickness < 7 mm). Supplemental Table 2. Crude and adjusted odds ratios of pregnancy outcomes (endometrial thickness < 7 mm). Supplemental Table 3. General characteristics and obstetric-neonatal outcomes of patients with singleton live births (endometrial thickness < 7 mm). Supplemental Table 4. Crude and adjusted odds ratios of obstetric-neonatal outcomes (endometrial thickness < 7 mm). Supplemental Table 5. General characteristics of patients undergoing true natural cycles and modified natural cycles (endometrial thickness < 8 mm). Supplemental Table 6. Crude and adjusted odds ratios of pregnancy and obstetric-neonatal outcomes in true natural cycles and modified natural cycles (endometrial thickness < 8 mm) Supplementary Material 1. Supplemental Table 1. General characteristics and pregnancy outcomes of patients undergoing different endometrial preparation protocols (endometrial thickness < 7 mm). Supplemental Table 2. Crude and adjusted odds ratios of pregnancy outcomes (endometrial thickness < 7 mm). Supplemental Table 3. General characteristics and obstetric-neonatal outcomes of patients with singleton live births (endometrial thickness < 7 mm). Supplemental Table 4. Crude and adjusted odds ratios of obstetric-neonatal outcomes (endometrial thickness < 7 mm). Supplemental Table 5. General characteristics of patients undergoing true natural cycles and modified natural cycles (endometrial thickness < 8 mm). Supplemental Table 6. Crude and adjusted odds ratios of pregnancy and obstetric-neonatal outcomes in true natural cycles and modified natural cycles (endometrial thickness < 8 mm)