The Impact Of Assisted Reproductive Technology On Birth Defects: A Systematic Review And Meta-analysis

Background: With the development of assisted reproductivetechnology (ART), the number of ART infants was over three million. Yet,there are still controversies on whether ART is safe enough. This studysystematically reviewed the existing cohort studies looking at the effects ofART on birth defects, and compared whether there were significantdifferences in the rates of birth defects between naturally conceivedchildren and ART infants.Aims: We aimed to examine whether the rate of birth defects in ARTinfants was higher than naturally conceived children, and if yes, we furtheraimed to elucidate whether the higher rate of birth defects was mainly dueto ART or due to the high rate of multiple pregnancies.Methods: Studies quantifying the association between ART and birthdefects published from1978to2012were selected using the followingdatabases: PubMed, Embase, CENTRAL, BIOSIS Previews, ChineseJournal Database, Weipu Database, Wanfang Database, and ChineseBiomedical Medicine Database. Two researchers conducted the literatureinclusion, literature quality assessment and data extraction according to the literature inclusion and exclusion criteria. RevMan5.1and Stata10.0wereused to conduct the meta-analysis and publication bias testing.Results:10English papers with good quality were included in thefinal analysis. We found that the birth defects rate of ART infants wassignificantly higher than the naturally conceived children. We divided thestudies into two subgroups, and found that in the singleton-birth subgroup,the birth defect rate of ART infants increased by34%((RR=1.34，95%CI[1.23，1.41]，P＜0.05）, compared with naturally conceived children.In the two-pregnancy subgroup, the birth defect rate of ART infantsincreased by100%(RR=2.00，95%CI[1.44，2.78]，P＜0.05) compared withnaturally conceived children. The IVF subgroup quantitative analysisrevealed that IVF might increase the risk of birth defects, while the birthdefect rate of ICSI infants was not significantly higher than IVF infants（X2=0.25，P＞0.05）. The ICSI subgroup meta-analysis indicated that ICSIwas weakly correlated with birth defects (RR=1.31，95%CI[1.23，1.40]，P＜0.05）. Funnel plot and Egger regression analysis illustrated that therewas no publication bias.Conclusions: The birth defects risk of ART infants was significantlyhigher than the naturally conceived children. ART was weakly correlatedwith birth defects after controlling for the possible effect of multiplepregnancies, revealing that the rate of multiple pregnancies was an essentialrisk factor of children’s birth defects. Both ICSI and IVF may increase the risk of birth defects, though there was no significant difference between thebirth defect rates of ICSI group and IVF group. The heterogeneity analysis,sensitivity analysis and publication bias results suggested that the result ofthis meta-analysis is robust.