Effect of maternal age distribution and prenatal diagnosis on the population rates of Down syndrome--a comparative study of nineteen populations.

We analyzed published data from 1980-85 on maternal age specific rates of Down syndrome in nineteen malformation monitoring systems. In the comparisons, we used maternal age specific Down syndrome baseline rates collected in Sweden using multiple sources and from a period before intense prenatal diagnosis. We supposed that no real maternal age specific risk difference exists between different populations, and that ascertainment of cases is independent of maternal age. With these suppositions, most programs had an ascertainment level of 70-90%, and some had about the same but none had a higher ascertainment level than that in the Swedish baseline. Marked differences in maternal age distribution in the populations studied resulted in a 12.3% standard deviation of the average population Down syndrome rate. The highest theoretical rate (estimated from actual maternal age distribution and Swedish baselines) was found in Spain (20% above average), the lowest in Czechoslovakia (24% below average). The estimated average effect of prenatal diagnosis was a decrease of only 6%, but in Denmark it was 25% and in Sweden and France:Paris 13%.