The study of very early pregnancy loss is impractical in the general population, but possible amongst infertility patients receiving carefully monitored treatments. We examined the association between fetal loss and the risk of birth defects in the surviving co-twin in a retrospective cohort study of infertility patients within an infertility clinic in South Australia from January 1986 to December 2002, linked to population registries for births, terminations and birth defects. The study population consisted of a total of 5683 births. Births from singleton pregnancies without loss were compared with survivors from (1) pregnancies with an empty fetal sac at 6-8 weeks after embryo transfer, (2) fetal loss subsequent to 8-week ultrasound and (3) multiple pregnancy continuing to birth. Odds ratios (OR) for birth defects were calculated with adjustment for confounders. Amongst infertility patients, the prevalence of birth defects was 7.9% for all twin pregnancies without fetal loss compared with 14.6% in pregnancies in which there had been an empty sac at ultrasound, and 11.6% for pregnancies with fetal loss after 6-8 weeks. Compared with singleton pregnancies without loss, the presence of an empty sac was associated with an increased risk of any defect (OR=1.90, 95% confidence intervals (CI)=1.09-3.30) and with multiple defects (OR=2.87, 95% CI=1.31-6.28). Twin pregnancies continuing to birth without loss were not associated with an overall increased prevalence of defects. We conclude that the observed loss of a co-twin by 6-8 weeks of pregnancy is related to the risk of major birth defects in the survivor.
Keywords: Assisted reproductive technologies (ART); congenital abnormalities; fetal loss; infertility; vanishing twin.