Prenatal diagnosis, pregnancy terminations and prevalence of Down syndrome in Atlanta

BACKGROUND: The impact of prenatal diagnosis on the live birth prevalence of Down syndrome (trisomy 21) has been described. This study examines the prevalence of Down syndrome before (1990-1993) and after inclusion of prenatally diagnosed cases (1994-1999) in a population-based registry of birth defects in metropolitan Atlanta. METHODS: We identified infants and spontaneous fetal deaths with Down syndrome (n = 387), and pregnancies electively terminated after a prenatal diagnosis of Down syndrome (n = 139) from 1990 to 1999 among residents of metropolitan Atlanta from a population-based registry of birth defects, the Metropolitan Atlanta Congenital Defects Program (MACDP). Only diagnoses of full trisomy 21 were included. Denominator information on live births was derived from State of Georgia birth certificate data. We compared the prevalence of Down syndrome by calendar period (1990-1993, 1994-1999), maternal age (<35 years, 35+ years), and race/ethnicity (White, Black, other), using chi-square and Fisher's exact tests. RESULTS: During the period when case ascertainment was based only on hospitals (1990-1993), the prevalence of Down syndrome was 8.4 per 10,000 live births when pregnancy terminations were excluded and 8.8 per 10,000 when terminations were included. When case ascertainment also included perinatal offices (1994-1999), the prevalence of Down syndrome was 10.1 per 10,000 when terminations were excluded and 15.3 when terminations were included. During 1990-1993, the prevalence of Down syndrome was 24.7 per 10,000 among offspring to women 35+ years of age compared to 6.8 per 10,000 among offspring to women <35 years of age (rate ratio [RR] = 3.65, 95% confidence interval [CI] = 2.53-5.28). During 1994-1999, the prevalence of Down syndrome was 55.3 per 10,000 among offspring to women 35+ years compared to 8.5 per 10,000 among offspring to women <35 years (RR = 6.55, 95% CI = 5.36-7.99). There was no statistically significant variation in the prevalence of Down syndrome by race/ethnicity within maternal age and period of birth strata. During 1994-1999, the proportion of cases that were electively terminated was greater for women 35+ years compared to women <35 years (RR = 5.10, 95% CI = 3.14-8.28), and lower for Blacks compared to Whites among women 35+ years of age (RR = 0.33, 95% CI = 0.16-0.66). CONCLUSIONS: In recent years, perinatal offices have become an important source of cases of Down syndrome for MACDP, contributing at least 34% of cases among pregnancies in women 35+ years of age. Variation in the prevalence of Down syndrome by race/ethnicity, before or after inclusion of cases ascertained from perinatal offices, was not statistically significant. Among Down syndrome pregnancies in mothers 35+ years we found a lower proportion of elective termination among Black women compared to White women. We suggest that future reports on the prevalence of Down syndrome by race/ethnicity take into account possible variations in the frequency of prenatal diagnosis or elective termination by race/ethnicity.