A search for evidence for a paternal age effect independent of a maternal age effect in birth certificate reports of Down's syndrome in New York state.

The discovery that in 20% to 30% of Down's syndrome cases the extra chromosome is of paternal origin, and the recent independent report of two groups that maternal age-specific rates are two-fold greater for livebirths to couples in which the father is aged 55 years and over prompted this investigation. Analyses were of coded birth certificate reports of Down's syndrome in Upstate New York residents in the years 1963-1974. The expected numbers of cases, on the assumption of no paternal age effect, were determined at each paternal age interval (and at each paternal age minus maternal age interval) adjusting for an effect of maternal age; these were compared with observed values. There was a slightly lower number of observed than expected cases for fathers aged 55 years and over (ratio = 0.76), and the results exclude with 95% confidence an increase of 1.5-fold or greater in rates in this group after correction for maternal age. There was, moreover, no overall evidence for any trend to increasing rates with paternal age. Regression analyses in which the data were first fit to functions of maternal age and subsequently terms involving paternal age were introduced also revealed no evidence that paternal age made a significant independent contribution to the observed rates in contrast to the conclusion of earlier positive reports.