Association between peak serum estradiol level during controlled ovarian stimulation and neonatal birthweight in freeze-all cycles: a retrospective study of 8501 singleton live births

Abstract

Is there an association between peak serum estradiol (E2) level during controlled ovarian stimulation (COS) and neonatal birthweight in freeze-all cycles? Peak serum E2 level during ovarian stimulation is not associated with neonatal birthweight in freeze-all cycles. Supraphysiologic E2 levels during COS have been demonstrated to generate a suboptimal peri-implantation endometrial environment and thus lead to adverse neonatal outcomes in fresh embryo transfer cycles. Previous experimental studies also suggested a potential influence of superovulation on oocyte epigenetic programming, but whether it translates into altered phenotypes of fetal growth and development remains unclear in clinical practice. By segmenting the process of COS and embryo transfer, the freeze-all policy provides a novel model to investigate the sole impact of ovarian stimulation on oocytes after ruling out the effects of hyperestrogenic milieu on endometrium in fresh cycles. A retrospective cohort study of 8501 patients who underwent their first COS cycles with a freeze-all strategy and delivered live-born singletons in subsequent frozen-thawed embryo transfer cycles from January 2007 to December 2016 at a tertiary-care academic medical center. Patients were categorized into six groups according to E2 level on trigger day in regular increments of 1000pg/mL: <1000, 1000-1999, 2000-2999, 3000-3999, 4000-4999 and ≥5000pg/mL. Univariable and multivariable linear regression and logistic regression analysis were performed to assess the independent association between peak E2 level and measures of neonatal birthweight including absolute birthweight, Z-score, low birthweight (LBW) and small-for-gestational age (SGA). The six groups did not differ significantly in birthweight, Z-score or the incidence of LBW and SGA. Compared with the E2 <1000pg/mL group, the adjusted mean difference (95% confidence interval [CI]) of stratified higher E2 groups was 17.2 (-31.0-65.5), 12.3 (-35.9-60.5), -4.1 (-51.9-43.7), -0.6 (-48.9-47.8) and -3.6 (-50.0-42.8) g for birthweight, and 0 (-0.11-0.10), 0.02 (-0.08-0.12), 0.04 (-0.06-0.14), -0.01 (-0.11-0.10) and -0.04 (-0.14-0.06) for Z-score, respectively. Regarding the outcomes of LBW and SGA, no increased risks were observed in each E2 category, with the adjusted odds ratio (95% CI) being 1.21 (0.68-2.16), 1.0 (0.58-1.90), 0.90 (0.50-1.63), 0.93 (0.51-1.69) and 1.08 (0.61-1.90) for LBW, and 0.97 (0.58-1.64), 1.06 (0.63-1.77), 0.77 (0.46-1.31), 0.71 (0.41-1.22) and 1.00 (0.60-1.65) for SGA, respectively. The study was retrospective in design, and other unknown confounding factors may not be included for adjustment. Furthermore, the generalization of the study finding could be limited to some extent by the majority of double cleavage-stage embryo transfer and difference in birthweight reference percentiles between Chinese and other populations. Our observations suggest that the hyperestrogenic milieu during COS does not seem to pose adverse effects on neonatal birthweight after frozen-thawed embryo transfer, which provides reassuring information for high ovarian responders in freeze-all cycles concerning their offspring's health. This study was funded by the National Key Research and Development Program of China (SQ2018YFC100163) and National Natural Science Foundation of China (81571397, 81771533). The authors declare no conflict of interest.