Patient and cycle characteristics
This study included 6096 frozen-thawed blastocyst transfer cycles. The female age, duration of infertility, endometrial thickness and body-mass index were 20-45 years, 0-25 years, 8-14 mm and 13-42 kg/m² respectively. The detailed cycle characteristics of the different groups are shown in Table S1, where no significant differences were observed among the < 35, 35-39 and > 39 age groups. The differences of body-mass index, natural cycle rate and duration of infertility were found to be significant in the total age group (P values were all less than 0.01).
Pregnancy outcomes
Table 1 presents the effects of quality and number of transferred blastocysts on pregnancy outcomes. The results indicated that the quality and number of transferred blastocysts had significant effects on pregnancy outcomes. In the total age group, IR (75.3%) and LBR (62.8%) were observed to be highest in group G, while CPR (77.2%) and MPR (56.5%) were the highest in group GG. Additionally, AR was relatively high in groups P (17.2%) and PP (16.9%). Compared to groups without good-quality blastocysts (groups P and PP), CPR and LBR were higher and AR was lower in groups with good-quality blastocysts (groups G, GG and GP). In the groups with good-quality blastocysts, Group G had higher IR (P /OR/CI = 0.00/2.33/1.99-2.74 and 0.00/2.91/2.57-3.29) and lower MPR (P /OR/CI = 0.00/0.02/0.02-0.03 and 0.00/0.05/0.04-0.06) than groups GG and GP, and their CPR, AR and LBR were similar. In the groups without good-quality blastocysts, group P also had higher IR (P /OR/CI = 0.00/1.80/1.43-2.25) and lower MPR (P /OR/CI = 0.00/0.02/0.01-0.08) than group PP, and their CPR, AR and LBR possessed no significant differences. These results indicated that increasing the number of transferred blastocysts classified as same grade does not obviously increase CPR and LBR, but it would significantly increase MPR. For double blastocyst transfer (groups GG, GP and PP), the better the quality of blastocysts transferred, the higher the IR, CPR, MPR and LBR would be, and these rates were found to decrease significantly in group PP (GG vs PP, P /OR/CI = 0.00/2.07/1.74-2.47, 0.00/2.22/1.68-2.94, 0.00/2.62/1.94-3.55 and 0.00/2.02/1.56-2.62 respectively). For single blastocyst transfer, group G had a higher IR, CPR and LBR (P /OR/CI = 0.00/2.69/2.17-3.33, 0.00/2.69/2.17-3.33, 0.00/2.70-/2.06-3.54 respectively) and a low AR (P /OR/CI = 0.17/0.76/0.52-1.11) compared to group P.
In different age groups, the results indicated that IR, CPR and LBR of same blastocyst transfer groups decreased gradually with age, and these rates decreased significantly when female age was more than 39 years old (P values were all less than 0.05). MPR of groups GG, GP and PP were found to decrease gradually with age (P = 0.00, 0.00 and 0.00 respectively). Compared to group < 35, AR was found to be significantly increased in group > 39 (P values were all less than 0.05). In each age group, the changing trends of IR, CPR, MPR, AR and LBR were similar to that in total age group.
A logistic regression model for predicting live birth is given in Table 2. Variables that were statistically significant in Table S1, or those that were deemed to be clinically significant, were included in order to build the model. This model indicated that the most significant predictors for live birth were female age, type of infertility and transferred blastocysts. LBR of the same blastocyst transfer strategies groups decreased gradually with age, and groups with good-quality blastocysts (groups G, GG and GP) had a higher LBR than those without good-quality blastocysts (groups PP and P) (P /OR/CI = 0.00/2.10/1.81-2.43).
Neonatal characteristics
Tables 3 and 4 present the neonatal characteristics of live born singletons and twins in the total blastocyst transfer group and different blastocyst transfer groups. The results indicated that the singleton group had a higher average gestational age (P /OR/CI = 0.00/2.71/2.47-2.94) and birthweight (P /MD/CI = 0.00/826.62/778.92-874.32) as well as a lower cesarean section rate (P /OR/CI = 0.00/0.09/0.06-0.14), preterm labor rate (P /OR/CI = 0.00/0.06/0.04-0.07) and low birthweight rate (P /OR/CI = 0.00/0.05/0.04-0.07) than the twin group. Moreover, no significant differences were present in regard to sex ratio and congenital malformation rate between the two groups (P /OR/CI = 0.67/1.04/0.88-1.23 and 0.47/0.79/0.42-1.49). Table 4 illustrates that blastocyst transfer strategies had significant effects on gestational age, cesarean section and birthweight in the singleton group (P =0.00, 0.00 and 0.01 respectively). Monozygotic twins had a lower gestational age (34.9 ± 3.0 vs 35.9 ± 2.1, P /MD/CI = 0.01/-0.99/-1.77–0.20) and birthweight (2228.8 ± 502.8 vs 2485.6 ± 406.2, P /MD/CI = 0.00/-256.8/-406.6–107.1) as well as a higher preterm birth rate (75.8% vs 57.1%, P /OR/CI =0.01/3.12/1.25-7.83) and low birthweight rate (66.7% vs 44.3%,P /OR/CI = 0.00/2.52/1.48-4.30) compared to dizygotic twins. The sex ratio of good-quality embryo groups (groups G and GG) (1.35, 1349/998) was highest, followed by the GP group (1.16, 321/276) and poor-quality embryo groups (groups P and PP) (0.98, 219/223). Additionally, the difference was significant between good-quality and poor-quality embryo groups (P /OR/CI = 0.00/1.38/1.12-1.69).