Twenty-seven percent of Swedish women who become pregnant after undergoing in vitro fertilization have a multiple birth, compared with 1% of women overall. Babies conceived via in vitro fertilization are more likely than others to be born preterm, to have a low birth weight and to have congenital malformations. However, the results of a retrospective registry study suggest that these outcomes are mostly attributable to maternal characteristics and the occurrence of multiple births, rather than to in vitro fertilization.1

To calculate the risks of adverse pregnancy and infant health outcomes after in vitro fertilization, the researchers compared data on all 4,517 births following in vitro fertilization in Sweden from 1982 to 1995 with data on all 1.5 million births in the general population during that time period. The data came from national registries of births, malformations, cancers and deaths. For women in the in vitro fertilization group, the researchers also gathered information on their fertility treatment from the clinic that provided it.

Women who underwent in vitro fertilization differed from the overall population of women who gave birth in several ways that may influence pregnancy outcomes: They were only 23% as likely to be younger than 30, were 50% more likely to be having their first child and were only 82% as likely to smoke 10 cigarettes or more per day. Information on the length of time that women who conceived via in vitro fertilization experienced infertility was available for 65% of the women in that group; the median duration of infertility was six years.

Among all births to women who underwent in vitro fertilization, 27% were multiple births--24% twins, 3% triplets and a negligible proportion quadruplets. By contrast, the proportion of multiple births nationwide was only 1%. The investigators observe that the generally increasing frequency of multiple births following in vitro fertilization appears to be contributing to an upward trend in multiple births in all of the Swedish population. Pregnancies resulting from in vitro fertilization accounted for 2% of twin births in 1986-1990; that proportion increased to 13% in 1991-1995.

The duration of gestation was known for all but seven of the 3,305 singleton births resulting from in vitro fertilization. Three percent of these babies were born at less than 32 weeks of gestation (very preterm), and 11% were born at less than 37 weeks (preterm). In the general population, only 1% of babies were born at less than 32 weeks, and 5% were born at less than 37 weeks. Infants who were conceived using in vitro methods were 3.5 times as likely as Swedish infants in general to be very preterm; the increase in risk was reduced when the data were stratified for maternal age, parity and infertility (risk ratio, 1.5).

Birth weight was known for all but 10 singleton infants born after in vitro fertilization. Three percent of these infants weighed less than 1,500 g at birth (very low birth weight), and 9% weighed less than 2,500 g (low birth weight). The corresponding proportions for the general population were 1% and 4%, respectively. Infants born after in vitro fertilization had a substantially increased risk of being very low birth weight (risk ratio, 4.4). Again, the increased risk was lowered when maternal age, parity and infertility were taken into account (risk ratio, 1.5).

Among infants conceived via in vitro fertilization, the rate of perinatal mortality (defined as stillbirths and infant deaths within one week of birth) was eight per 1,000 births; in the general population, this rate was seven per 1,000. The risk ratio for perinatal death among infants conceived using in vitro fertilization (1.6) reached borderline significance before the data were stratified, but it was not significant after stratification for year of birth, maternal age, parity, years of infertility and birth weight.

Of the 5,856 infants who were born after in vitro fertilization, 5% had a congenital malformation. On the basis of this proportion, the analysts calculate that babies born after in vitro fertilization were 1.4 times as likely as Swedish infants in general to have a malformation. The overall likelihood of congenital malformation was significantly increased for singletons (risk ratio, 1.3), but not for multiple births. However, when the investigators examined specific malformations, they found that only three conditions were significantly more common among children born after in vitro fertilization than among others: anencephaly (absence of a major part of the brain), hydrocephalus (increase in the amount of cerebrospinal fluid in the cranial cavity) and atresia (closure) of the esophagus, which had risk ratios of 12.9, 5.7 and 3.9, respectively. Moreover, all of the babies with anencephaly and the majority with hydrocephalus were from sets of twins, suggesting that these conditions were due to the multiple pregnancy rather than to in vitro fertilization.

The number of children born after in vitro fertilization who later developed cancer (four) was similar to the number expected in the general population. The researchers caution that this finding is problematic because their data had limited power to study cancer incidence.

The investigators conclude that children born after in vitro fertilization have elevated odds of experiencing some poor medical outcomes, but that the increase is due largely to maternal risk factors and to the risks inherent in multiple births, rather than to the technique of assisted conception. They believe that "a carefully thought-out strategy is needed to try to lower the rate of multiple births," and that the "selection of one viable embryo" should be the goal of in vitro fertilization practice.

Two researchers commenting on the study agree that the benefits of inserting multiple embryos are outweighed by the risk associated with multiple pregnancies.2 Although the reasoning behind transferring more embryos is that there will be a greater chance of conception, these researchers feel that "data to support this view are scant." They also note that clinicians as well as parents tend to underestimate the risks involved with triplet pregnancies. While advocating that clinical practice should be to transfer only two embryos, they conclude that "the ultimate goal remains single-embryo transfer."

However, a group of clinicians in the field argue the opposite point.3 When the embryo quality is good and the recipient is younger than 38 or will not consider fetal reduction, they also recommend transferring only two embryos. But they have found that older women with poorer embryo quality benefit from a larger number of transfers. If clinical practice were restricted to two embryo transfers, they contend that "in a group of 1,000 women over the age of 40 years, two triplet pregnancies would be avoided at the expense of 44 women who did not achieve a live birth."--L. Gerstein


1. Bergh T et al., Deliveries and children born after in vitro fertilisation in Sweden 1982-95: a retrospective cohort study, Lancet, 1999, 354(9190):1579-1585.

2. Fisk NM and Trew G, Two's company, three's a crowd for embryo transfer, commentary, Lancet, 1999, 354(9190):1572-1573.

3. Abdalla HI, Gearon C and Wren M, Swedish in vitro fertilisation study, letter to the editor, Lancet, 2000, 355(9206):844-845.