Time trends in the prevalence of birth defects in Texas 1999-2007: Real or artifactual?

Abstract

Few studies have reported time trends for total birth defects or for a comprehensive range of phenotypes. We examined data from the Texas Birth Defects Registry (TBDR) from 1999 through 2007. Poisson regression was used to fit trend lines to birth prevalence over time for total birth defects (each infant/fetus counted once), for every birth defect collected by the TBDR, and for subsets of cases or defects grouped various ways. From 1999 through 2007, birth prevalence of total birth defects in Texas increased 3.6% per year. Increases were observed in all population groups, persisted after adjustment for demographic characteristics, and were strongest in regions of Texas that were more urban. There was a wide variety of different defects showing significant increases. The trends of several defects were driven by their mild cases. Perhaps the most compelling finding was that larger upward trends were observed in defects that had been rated as more susceptible to diagnostic variation. One notable exception to that was gastroschisis, which showed an average increase of over 5% per year, the total birth defects rate in TBDR increased at 3.6% per year, similar to 3.7% per year in birth certificate check boxes. In our opinion, the weight of evidence in our study suggests that the observed increase over time in total birth defects and in many specific birth defects is artifactual. This likely reflects increased awareness, referral, detection, or documentation in health care facilities visited by TBDR staff, resulting in more complete ascertainment by the registry, rather than a true change over time in the occurrence of most birth defects.