Risk of placenta previa in assisted reproductive technology: A Nordic population study with sibling analyses

BACKGROUND: A higher risk of placenta previa after assisted reproductive technology (ART) is well established. The underlying mechanisms are poorly understood, but may relate to embryo culture duration, cryopreservation, and cause of infertility. Within-mother analyses, where each woman is her own control (i.e., sibling design), help disentangle treatment contributions from maternal confounders that are stable between pregnancies. We aimed to investigate the risk of placenta previa in pregnancies achieved after ART according to embryo culture duration, cryopreservation, and infertility factors while accounting for stable maternal factors using within-mother analyses.

METHODS AND FINDINGS: We used linked nationwide registry data from Denmark (1994 to 2014), Finland (1990 to 2014), Norway (1988 to 2015), and Sweden (1988 to 2015). All women who gave their first birth during the study period at age 20 years or older were eligible and contributed up to 4 deliveries (singleton or multifetal) occurring between 22 and 44 weeks of gestation, excluding deliveries where maternal age exceeded 45 years. We used multilevel logistic regression to compare risk of placenta previa after ART (n = 139,694 deliveries) versus natural conception (n = 5,614,512 deliveries), both at the population level and within mothers, adjusting for year of delivery, maternal age, parity, and country. We categorized ART according to culture duration, embryo cryopreservation, and infertility factors. Population level risk of placenta previa was higher for ART versus natural conception (odds ratio [OR], 4.16; 95% confidence interval [CI], 3.96-4.37). Controlling for stable maternal factors, the association attenuated, but risk remained higher for ART versus natural conception (OR within mothers, 2.64; 95% CI, 2.31-3.02). Compared to naturally conceived, a larger difference in risk was seen for pregnancies from fresh embryos than for pregnancies from frozen embryos. Further categorization by culture duration showed the largest risk difference after fresh blastocyst transfer, and the smallest after frozen cleavage stage embryo transfer, which persisted in sensitivity analyses (including restriction to singletons). When stratified according to infertility factors at the population level, women with endometriosis conceiving by ART had the highest risk of placenta previa (OR, 9.35; 95% CI, 8.50-10.29), whereas women with polycystic ovary syndrome (PCOS) conceiving by ART had the lowest risk (OR, 1.52; 95% CI, 1.12-2.09), compared to natural conception. Within mothers, we found a higher risk of placenta previa after ART compared to natural conception for women with endometriosis (OR, 2.08; 95% CI, 1.50-2.90), but not for women with PCOS (OR, 0.88; 95% CI, 0.41-1.89 [unadjusted due to sparse data]). However, within-mother analyses are restricted to multiparous women with deliveries after different conception methods. Therefore, findings from these analyses might not generalize to all women undergoing ART.

CONCLUSIONS: The risk of placenta previa in pregnancies conceived by ART differed by embryo culture duration, cryopreservation, and underlying infertility. The highest risk was seen after fresh embryo transfer and especially fresh blastocyst transfer. Women with endometriosis had a higher risk than women with other infertility factors, and within mothers, their risk was higher after ART than after natural conception. Identifying the responsible mechanisms might provide opportunities for prevention.