Neurotoxicity Study: Test of Trend in Multiple Endpoints Correlated Data

The data are obtained from a neurotoxicity bioassay. Such studies often center on determining the effects of some compound on measures contained in the so-called Functional Observational Battery ( FOB ). In FOB studies, it is assumed that an adverse effect of a substance is manifested as a distinct change in an animal’s physiological or behavioral response. Due to the fact that many outcomes are involved in an FOB , it may be that one or more of these items will reflect the same underlying effect of toxicity.

Data are provided by Dr. Virginia Moser at the U.S. EPA, and are reprinted from Han, et al. (2003). The data are from a study of the acute neurotoxic effects of perchlorethylene (also know as tetrachloroethylene, PCE, and PERC), a manufactured chemical which is most commonly known for its use in the dry cleaning industry, as well as for textile manufacturing, vapor degreasing, metal cleaning operations, and the production of rubber coatings.

In the PERC experiment considered here, forty animals were randomized to one of five dose levels, 0, 0.15, 0.5, 1.5, or 5 g/kg of perchlorethylene, resulting in eight animals per dose group. Four hours after exposure, each animal was classified for presence or absence of an adverse reaction with respect to different outcomes of Central Nervous System Excitability. The Table below summarizes the number of animals in each dose group that exhibited an adverse reaction to PERC for each of three outcomes comprising the Central Nervous System Excitability domain: Handling reactivity, clonic movement, and arousal.

The analysis of the basic data of the 40 animals, in StatXact 7, produces the following results.

The p-value indicates that the trend in multiple end points adverse reaction with respect to increasing dose levels is highly significant.