Perinatal outcome of 12 021 singleton and 3108 twin births after non-IVF-assisted reproduction: a cohort study

Perinatal outcome of pregnancies caused by assisted reproduction technique (ART) is substantially worse when compared with pregnancies following natural conception. We investigated the possible risks of non-IVF ART on perinatal health. We conducted a retrospective cohort study with two exposure groups: a study group of pregnancies after controlled ovarian stimulation (COS), with or without artificial insemination (AI), and a naturally conceived comparison group. We used the data from the regional registry of all hospital deliveries in the Dutch-speaking part of Belgium during the period from January 1993 until December 2003 to investigate differences in perinatal outcome of singleton and twin pregnancies. 12 021 singleton and 3108 twin births could be selected. Naturally conceived subjects were matched for maternal age, parity, fetal sex and year of birth. The main outcome measures were duration of pregnancy, birth weight, perinatal morbidity and perinatal mortality. Our overall results showed a significantly higher incidence of prematurity (<32 and <37 weeks), low and very low birth weight, transfer to the neonatal intensive care unit and most neonatal morbidity parameters for COS/AI singletons. Twin pregnancies resulting from COS/AI showed an increased rate of neonatal mortality, assisted ventilation and respiratory distress syndrome. After excluding same-sex twin sets, COS/AI twin pregnancies were at increased risk for extreme prematurity and very low birth weight. In conclusion, COS/AI singleton and twin pregnancies are significantly disadvantaged compared to naturally conceived children.