Statistical Aspects of Three-Generation Studies of the Reproductive Toxicity of TCDD and 2,4,5-T

Abstract

Abstract Three-generation studies of the effects of toxic chemicals on mammalian reproduction yield large volumes of data on many small groups of animals. Statistical problems involved in analyzing such data include selection of the appropriate statistical unit, multiple comparisons, identification of dose-related trends, and identification of consistent responses across generations. This paper includes a reanalysis of data from three-generation studies of 2,4,5-T and TCDD in rats. Although the original authors reported establishing “no-observed-effect levels” in these studies, the use of more appropriate statistical techniques shows that the lowest dose level in the TCDD study was an effect level, and that the lowest dose level in the 2,4,5-T study may have been so.