Supraphysiologic estradiol is an independent predictor of low birth weight in full-term singletons born after fresh embryo transfer

Is supraphysiologic estradiol (E-2) an independent predictor of low birth weight (LBW) in singletons born after fresh IVF-embryo transfer (ET) cycles? Our results suggest that E-2 > 2500 pg/ml is an independent predictor for LBW in full-term singletons born to normal responder patients undergoing fresh IVF-ET cycles. The pathogenesis of LBW in IVF singletons remains unknown. However, recent studies have suggested that the hyperestrogenic milieu generated during ovarian stimulation may create a sub-optimal peri-implantation environment, leading to placental dysfunction, and therefore, LBW. Retrospective cohort study of normal responder patients, < 40 years old, undergoing fresh IVF-ET cycles resulting in live singleton births between January 2005 and June 2014. A total of 6419 patients had live births after fresh IVF-ET during the study period, of which 2348 (36.6%) patients were excluded due to multiple gestation, vanishing twins or incomplete records. Perinatal outcomes recorded for all patients included birth weight, gestational age (GA) at delivery, mode of delivery and gender. Term birth, preterm birth (PTB) and LBW incidence proportions were plotted against E-2 level on the day of trigger. The term LBW incidence proportion (i.e. singletons born at GA 37 weeks with birth weight < 2500 g) was considered the primary outcome of interest. A total of 4071 patients with live singleton births were included. The median age, BMI, E-2 level and birth weight for the study cohort was 36 (33-39) years, 22.3 (20.4-25.0) kg/m(2), 1554 (1112.7-2179) pg/ml and 3289 (2920-3628) g, respectively. The incidence proportion of LBW rose from 6.4% (E-2 2001-2500 pg/ml) to 20.7% (E-2 3501-4000 pg/ml), without a corresponding rise in the incidence proportion of PTB. The odds of term LBW with E-2 > 2500 pg/ml were 6.1-7.9 times higher compared to the referent E-2 group. Multivariable logistic regression analysis revealed that E-2 was an independent predictor for term LBW, even after adjusting for age, BMI, race, parity, infertility diagnosis, duration of ovarian stimulation, gonadotropin dosage and method of insemination (adjusted odds ratio 10.8, 95% CI 9.2-12.5). Receiver operating characteristic analysis generated an AUC estimate of 0.85 for E-2 level as a predictor of LBW. The current study did not include analyses of hypertensive disorders of pregnancy or placental abnormalities. Furthermore, all patients were normal responders and of normal BMI, possibly limiting the overall generalizability of the study. Finally, as with any retrospective study, prospective data are required to validate the role of E-2 in predicting LBW. Our results emphasize the importance of minimizing the supraphysiologic elevations of E-2 levels during ovarian stimulation in fresh IVF-ET cycles. This, in turn, can optimize the early peri-implantation environment and mitigate adverse perinatal outcomes such as LBW.