Showing papers by "Gary M. Brittenham published in 1989"

Lead-related Birth Defects: Some Methodological Issues

Abstract: In the course of reviewing our findings on the topic of foetal lead effects, several methodological problems are noted. The first is the effective measurement and control of important confounding variables. In our studies, we demonstrated that failure to consider the effect of foetal lead exposure with the effects of maternal use of alcohol and cigarettes during pregnancy can lead to discrepant findings in the association of lead exposure and the gestational age, birth weight, length and head circumference of the neonate. We also demonstrated that similar discrepancies in findings for the size measurements might follow from failure to consider parental size. Similarly, the association of lead level and neonatal anomalies might be a function of failure to obtain adequate measurements of maternal alcohol use and to control for this confounding condition in statistical models. The second concern is the effect of adjusting cord blood lead (PbB) data for haematocrits. We demonstrated that effect sizes tend to be reduced when PbB corrected for haematocrits is used in place of uncorrected values. Although the haematocrit measure in our data was correlated with cord PbB, it did not appreciably alter a previously reported association of foetal lead exposure and neonatal status.