Title: Endocrine disruptors and sexually dimorphic behaviors: a question of heads and tails.

Authors: Weiss, B

Published In Neurotoxicology, (1997)

Abstract: Early brain development seems destined to be the territory where questions about the neurotoxicity of endocrine disruptors will be contested, so we ought to clarify how we propose to invade it. Everything else we have learned tells us that the most sensitive indices are likely to be the most subtle, a term we often use to describe the phenomena and domain of neurobehavioral toxicology. It might mean an effect that is small in amplitude, or that requires sensitive measures to detect. One class of effects that might result from environmental exposure to endocrine disruptors reflects both kinds of subtlety. It consists of sexually dimorphic behaviors that seem biologically driven but that could easily masquerade as socially-defined or gender differences. Performances on neuropsychological tests fall into this category. Beyond debates about their origins, such differences tend to be relatively modest in size, especially by the routine criteria adopted by toxicologists. Overwhelmingly, they rely on differences in means accompanied by p values. Such criteria are inappropriate for assessing hypotheses or calculating risks based on the impact of what invariably will be quite low environmental levels. In this arena at least, the quantification of effects should proceed from examinations of distributions. Instead of conventional descriptions based on measures of central tendency, we should instead examine the shapes of distributions, especially the tails. Neuropsychological differences between males and females are amplified and emerge at the extremes, even when sex accounts for no more than a minuscule proportion of the variance. Lead taught a similar lesson about how to quantify population impacts. Furthermore, such differences at the extremes suggest both practical and theoretical implications that can too easily be ignored but that provide paths for fruitful exploration.

PubMed ID: 9291506

MeSH Terms: No MeSH terms associated with this publication