Prenatal PCB exposure and neurobehavioral development in infants and children: Can the Oswego study inform the current debate?

Abstract

AbstractIn the current paper we describe the methodology and results of the Oswego study, in light of D.V. Cicchetti, A.S. Kaufman, and S.S. Sparrow's (this issue) criticisms regarding the validity of the human health/behavioral claims in the PCB literature. The Oswego project began as a replication of the Lake Michigan Maternal Infant Cohort study. Beyond replication of the Michigan findings, the study sought to extend results and conclusions through more comprehensive behavioral assessment, and improved confounder control and analytic methodology. Results over the past 5 years have demonstrated a convincing replication of the Michigan findings. The Michigan cohort reported findings relating Great Lakes fish consumption to performance impairments on the Neonatal Behavioral Assessment Scale (J. Jacobson, S. Jacobson, P. Schwartz, G. Fein, & J. Dowler, 1984). These findings were also found in the Oswego cohort (E. Lonky, J. Reihman, T. Darvill, J. Mather, & H. Daly, 1996), and the Oswego study extended the association to cord blood PCBS (P.W. Stewart, J. Reihman, E. Lonky, and T. Darvill, 2000). The Michigan cohort reported an association between prenatal PCB exposure and poorer performance on the Fagan Test of Infant Intelligence (S.W. Jacobson, G.G. Fein, J.L. Jacobson, P.M. Schwartz, & J.K. Dowler, 1985). The Oswego cohort found similar results (T. Darvill, E. Lonky, J. Reihman, P. Stewart, & J. Pagano, 2000). The Michigan Cohort reported an association between prenatal PCB exposure and performance impairments on the McCarthy Scales of Children's abilities (J. Jacobson & S. Jacobson, 1997). The Oswego study also found PCB‐related impairments on the McCarthy Scales (P.W. Stewart, J. Reihman, E. Lonky, T. Darvill, & J. Pagano, 2003). The Oswego results used the same exposure metric in every paper, employed conservative statistical design and analysis, and controlled for more than 40 potentially confounding variables. Moreover, while PCBs were related to all the behavioral endpoints outlined above, alternative candidates for effect, including lead, HCB, Mirex, DDE, and MeHg were not. Taken together, these results support the hypothesis that prenatal PCB exposure results in statistically significant predictors of small, but measurable, deficits in cognitive development from infancy through early childhood. Cicchetti et al. argue that these results, generated by independent investigators, be dismissed because they reflect a combination of measurement error, Type I error, and residual confounding. The evidence Cicchetti et al. present in support of their position fails to explain the nearly identical pattern of associations observed in the Oswego and Michigan Cohorts. In light of this replication, the extensive assessment of potential confounders, the effective elimination of alternative contaminants, and the conservative statistical approach employed in the Oswego study, we find that Cicchetti et al.'s claims are not substantiated. © 2004 Wiley Periodicals, Inc. Psychol Schs 41: 639–653, 2004.