Fathers of children conceived using ART have higher cognitive ability scores than fathers of naturally conceived children

STUDY QUESTION Does paternal cognitive ability differ for children conceived with and without assisted reproductive technology (ART)? SUMMARY ANSWER Young fathers of ART conceived children tend to score cognitively below their same-age natural conception (NC) counterparts and older (above 35) fathers of ART conceived children tend to score above. WHAT IS KNOWN ALREADY Cognitive ability is a genetically and socially transmitted trait, and If ART and NC children have parents with different levels of this trait, then this would in itself predict systematic differences in child cognitive outcomes. Research comparing cognitive outcomes of children with different modes of conception finds conflicting results, and studies may be influenced by selection and confounding. STUDY DESIGN, SIZE, DURATION This is a population-based study based on Norwegian data, combining information from the Medical Birth Registry (births through 2012), military conscription tests (birth cohorts 1955–1977) and the population registry. These data allow us to compare the cognitive ability scores of men registered as the father of an ART-conceived child to the cognitive abilities of other fathers and to average scores in the paternal birth cohorts. PARTICIPANTS/MATERIALS, SETTINGS, METHODS The population level study included 18 566 births after ART (5810 after ICSI, 12 756 after IVF), and 1 048 138 NC births. It included all Norwegian men who received a cognitive ability score after attending military conscription between 1973 and 1995. This constituted 614 827 men (89.4% of the male birth cohorts involved). An additional 77 650 unscored males were included in sensitivity analyses. MAIN RESULTS AND THE ROLE OF CHANCE Paternal cognitive level was assessed using intelligence quotients (IQ) converted from stanine scores on a three-part cognitive ability test with items measuring numeracy, vocabulary and abstract thought (Raven-like matrices). ART fathers averaged 1.95 IQ points above the average of their own birth cohort (P-value < 0.0005) and 1.83 IQ points above NC fathers in their own birth cohort (P < 0.0005). Comparisons of the IQ of ART fathers to those of NC fathers of similar age and whose children were born in the same year, however, found average scores to be more similar (point estimate 0.24, P = 0.023). These low average differences were found to differ substantially by age of fatherhood, with young ART fathers scoring below their NC counterparts and older ART fathers scoring above their NC counterparts. LIMITATIONS, REASONS FOR CAUTION We do not have information on maternal cognition. We also lack information on unsuccessful infertility treatments that did not result in a live birth. WIDER IMPLICATIONS OF THE FINDINGS Paternal cognitive ability of ART children differs from that of NC children, and this difference varies systematically with paternal age at child birth. Selection effects into ART may help explain differences between ART and NC children and need to be adequately controlled for when assessing causal effects of ART treatment on child outcomes. STUDY FUNDING/COMPETING INTEREST(S) This research has also been supported by the Research Council of Norway through its Centres of Excellence funding scheme, project number 262700 (Centre for Fertility and Health). It has also been supported by the Research Council of Norway’s Project 236992 (Egalitarianism under pressure? New perspectives on inequality and social cohesion). There are no competing interests. TRIAL REGISTRATION NUMBER N/A