Parents who are having trouble conceiving may hesitate to use assisted reproductive technology (ART) because of associations between ART and worse perinatal outcomes in offspring. To address limitations in previous literature in this regard, researchers from Finland analyzed potentially harmful effects of ART using a sibling-comparison model approach.
Published in The Lancet, the findings stem from data collected on a random sample of 20% of Finnish households with at least one child between 0 and 14 years at the end of 2000 (n=65,723). They then analyzed birthweight, gestational age, risk of low birthweight and risk of preterm birth (PTB) among children conceived naturally and through ART (or medically assisted reproduction, the term used by the researchers).
Differences in birth outcomes by mode of conception in the general population were estimated using standard multivariate methods that controlled for factors including multiple births, birth order and parental sociodemographic characteristics. Using a sibling comparison approach, the authors compared children conceived by ART with children conceived naturally as a way to control for all observed and unobserved factors shared by siblings.
The sample included 2,776 (4%) children who were conceived by ART between 1995 and 2000. Of them, 1,245 (natural = 620, ART = 625) were included in the within-family analysis (analyzing birth outcomes by comparing siblings born to the same parents but conceived either naturally or through medical assisted reproduction). In absolute terms, children who were conceived through ART had worse perinatal outcomes than children who were conceived naturally.
Using between-family analysis (analyzing birth outcomes by comparing different families), the authors found that 356 of the 2,776 children (13%)conceived through ART had low birthweight (< 2500 g at birth) compared with 2189 of the 62,947 children (3%) conceived naturally. Offspring of ART were also more likely than their naturally conceived counterparts to be firstborn (1710 [62%] vs 23738 [38%]) and almost 10 times likelier to be from a multiple birth (580 [21%] vs 1365 [2%]).Using within-family analysis, the authors noticed that differences in birth outcomes were reduced but not completely eliminated.
In linear model analysis, children conceived by ART had a difference in birthweight of -60g (95% CI -86 to -34) and a 2.15 percentage point (95% CI 1.07 to 3.24) increased risk of PTB. In the sibling comparison, the difference in birthweight was -31 g (95% CI -85 to 22) and a 1.56 percentage point (95% CI -1.26 to 4.38) increase in risk of PTB.
The authors noted several strengths to their study. The large dataset allowed sibling comparison, data were not prone to self-selection, and the methodological approach allowed the authors to account for unobserved parental characteristics shared by siblings. Limitations include inability to test whether the effects of ART on birth outcomes vary according to length of infertility, medication dose, or number of treatment cycles or birth order.
Ultimately, the authors believe that it is important for parents who are considering ART to understand the risks involved. But their results, they said, suggest that while children born through ART are at an absolute higher risk for adverse birth outcomes, those risks might be attributable to factors other than the treatment.