The present study enrolled 93,959 oocyte retrieval cycles of 73,794 patients. Groups 1-6 included 20,641, 21,307, 16,613, 11,793, 9,379 and 14,226 cycles, respectively. Furthermore, 45,852 cycles of 42,315 patients underwent fresh embryo transfer. Subsequently, 56,760 FET cycles of 40,674 patients were performed. A total of 45,620 patients performed 49,214 complete cycles (Figure 1).
The baseline characteristics significantly differed among groups (Table 1). Patients with higher E2 tended to possess younger age, lower BMI, basal FSH and lower proportions of diminished ovarian reserve, but the proportions of primary infertility and the use of GnRH-a regimen were higher. The antral follicle counts, total gonadotrophin dose, progestin level on trigger day, and the number of oocytes retrieved were higher with the increase of E2. The rates for specific-stage embryos (0.84±0.24), IVF fertilization rates (0.72±0.35), and ICSI fertilization rates (0.78±0.31) were highest in group 1 (E2<1000 pg/ml). The difference in cleavage rates among groups were limited, even if there were statistical significance (Table 1).
The multivariate analysis of the preimplantation embryo developments of the 6 groups were displayed at Table 2. After adjusting for confounding factors, E2 on trigger day was negatively correlated to the IVF and ICSI fertilization rates, and specific stage embryos rates. Compared with group 1, the aMD (95% CI) for the specific stage embryo rates in groups 2-6 were -0.10 (-0.11, -0.10), -0.13 (-0.14, -0.12), -0.14 (-0.15, -0.13), -0.15 (-0.16, -0.15) and -0.14 (-0.15, -0.13), respectively. The differences of IVF and ICSI fertilization rates between group 1 and groups 2-6 were limited (0.02-0.04), although there were statistic significances. Furthermore, the cleavage rates in group 1 and the higher E2 groups (groups 2-6) had no significant difference in the multivariate model.
The demographic characteristics and treatment information revealed a significant difference among groups for patients who underwent fresh embryo transfer cycles (Table 3), FET cycles (Table 4), and complete cycles (Table 5). The tendency was similar to that of the recruited oocyte retrieval cycle. In addition, the proportion for day 3 embryos decreased as the E2 level increased in both fresh and FET cycles. The live birth rate was highest in group 5 (E2: 4,000-5,000 pg/ml) for fresh embryo transfer cycles (52.8%), and in group 6 (≥5,000 pg/ml) (55.1%) for FET cycles. The CLBR was highest (83.8%) in group 6.
In fresh embryo transfer cycle, the multivariate analysis revealed that the E2 on trigger day was not associated with biochemical pregnancy, clinical pregnancy, and the live birth rate (Table 6). Compared with E2<1,000 pg/ml, the aOR (95% CI) of biochemical pregnancy rates in higher E2 groups were 1.06 (0.94, 1.19), 1.12 (0.99, 1.27), 1.05 (0.91, 1.22), 1.04 (0.89, 1.22), and 0.94 (0.80, 1.11), respectively; the aOR (95% CI) of clinical pregnancy rates in higher E2 groups were 1.06 (0.94, 1.20), 1.11 (0.97, 1.27), 1.04 (0.90, 1.20), 1.07 (0.91, 1.25), and 0.90 (0.76, 1.05), respectively; and the aOR (95% CI) of live birth rates were 1.03 (0.91, 1.16), 1.06 (0.93, 1.21), 1.03 (0.89, 1.19), 1.02 (0.87, 1.20), and 0.88 (0.75, 1.03), respectively.
In FET cycle, the multivariate analysis showed that when E2 was higher than 5,000 pg/ml, E2 on trigger day was positively correlated to the clinical pregnancy rates and live birth rates. Compared to group 1, the aORs (95% CI) of clinical pregnancy rates in groups 2-6 were 1.04 (0.92, 1.17), 1.14 (1.00, 1.30), 1.14 (0.99, 1.31), 1.14 (0.99, 1.32), and 1.21 (1.06, 1.39), respectively; the aORs (95% CI) of live birth rates in groups 2-6 were 1.01 (0.90, 1.14), 1.08 (0.95, 1.22), 1.10 (0.96, 1.27), 1.12 (0.97, 1.29) and 1.20 (1.05, 1.37), respectively. The proportion of biochemical pregnancy cases in group 2 (aOR=1.16, 95% CI=1.02-1.32) and group 6 (aOR=1.21, 95% CI=1.06-1.40) was significantly higher, when compared to group 1. However, the biochemical pregnancy rates were comparable between groups 2, 4-5 and group 1, after adjusting for the confounders (Table 7).
The CLBR for the 49,214 complete cycles was 62.3%. The smooth curve fitted for CLBR on E2 revealed that for every 1,000 pg/ml increase in E2, when E2 was <5,500 pg/ml, the CLBR increased by approximately 10%, when E2 was between 5,500-13,281 pg/ml, the CLBR increased by approximately 1.8%, and when E2 was >13,281 pg/ml, the CLBR decreased by 3% (Figure 2). Based on the smooth fitting curve, the investigators formulated the multivariate analysis with expanded E2 groups. The results revealed that when E2 was within 1,000-14,000 pg/ml, the high level of E2 was correlated to the increase in CLBR. Compared with E2<1,000 pg/ml, the aOR (95% CI) of the CLBR for the E2 groups of 1,000-2,000 pg/ml, 2,000-3,000 pg/ml, 3,000-4,000 pg/ml, 4,000-5,000 pg/ml, 5,000-6,000 pg/ml, 6,000-8,000 pg/ml, 8,000-10,000 pg/ml, 10,000-120,000 pg/ml, and 120,000-140,000 pg/ml were 1.33 (1.13,1.56), 1.56 (1.31,1.85), 1.68 (1.39,2.02), 1.83 (1.50,2.24), 1.86 (1.49,2.33), 1.88 (1.51,2.34), 1.92 (1.46,2.54), 1.82 (1.26,2.65), and 1.92 (1.09,3.41), respectively. When E2 was higher than 14,000 pg/ml, this was irrelevant to the CLBR (Table 8).