Much ado about something: The weight of evidence for PCB effects on neuropsychological function

D.V. Cicchetti, A.S. Kaufman, and S.S. Sparrow (this issue) use six criteria to evaluate the published findings from seven different studies of PCB exposure and neuropsychological function. They point out a number of weaknesses or flaws in each study and conclude that these weaknesses make the overall conclusion that PCB exposure negatively impacts neuropsychological function untenable. While we agree that all of the studies have weaknesses, we take issue with some of the specific criticisms Cicchetti and colleagues make including their comments on exposure assessment, statistical control of multiple comparisons, validity of test instruments and clinical relevance of the findings. We argue that - despite the weaknesses of each individual study - the consistency of effects across studies is remarkable, and we conclude that the weight of evidence for a negative association between PCB exposure and intellectual function in childhood is very strong.