Impact of corpus luteum number on maternal pregnancy and birth outcomes: the Rotterdam Periconception Cohort

Objective: To investigate associations between assisted reproductive technology (ART)-induced alterations in corpus luteum (CL) number during implantation, and maternal pregnancy and birth outcomes. Pregnancies conceived through ART are associated with increased risks of adverse obstetric and perinatal outcomes, with discrepancies in outcomes between different ART treatment protocols. We hypothesize that this is because the periconceptional hormonal environment regulated by the CL. Assisted reproductive technology protocols affect CL quantity at conception resulting in CL absence or a supraphysiological number of CL. Design: This study is embedded in the Rotterdam Periconceptional Cohort, an ongoing tertiary center prospective cohort at the Erasmus University Medical Center in Rotterdam, the Netherlands. Subjects: Women with a singleton pregnancy with data on CL. Exposure: The number of CL, based on mode of conception: 0 CL (artificial-cycle frozen embryo transfer, n = 72); >1 CL (ovarian stimulated fresh embryo transfer, n = 462), and 1 CL (natural-cycle frozen embryo transfer and natural conceptions, n = 1327). Main Outcome Measures: Hypertensive disorders of pregnancy, gestational diabetes, gestational age at birth and birthweight, derived from medical records. Results: We included 1,861 pregnancies. The results were adjusted for maternal age, maternal body mass index, nulliparity, and obstetric history. In comparison with natural conceptions (1 CL), a pregnancy with CL absence (0 CL) was associated with a higher risk of gestational diabetes (adjusted odds ratio [aOR]: 2.59 [95% confidence interval (CI): 1.31–5.15]), and a higher risk of preeclampsia, albeit non-significantly (aOR: 2.02 [95% CI: 0.91–4.51]). In comparison with pregnancies with >1 CL, the risk of preeclampsia was significantly lower (aOR: 0.36 [95% CI: 0.18–0.72]). Post hoc analyses revealed that in male neonates, >1 CL was associated with a lower birthweight percentile (aβ; –6.18 [95% CI: –11.16 to –1.20]). In contrast, female neonates showed no association with >1 CL, whereas CL absence was associated with a higher birthweight percentile (aβ: 12.93 [95% CI: 2.52–23.34]). Conclusions: Risks of hypertensive disorders of pregnancy, gestational diabetes, and relative birthweight differ between CL groups. These findings support the hypothesis that an aberrant number of CL impacts maternal pregnancy and or birth outcomes. Additional studies need to be conducted to investigate causes and underlying pathophysiology. Clinical Trial Registration Number: NL4115 (https://onderzoekmetmensen.nl/en/trial/25560; date of registration: October 15, 2004; date of enrollment first subject: November 4, 2010).