Graduation Year

2004

Document Type

Thesis

Degree

M.S.P.H.

Degree Granting Department

Public Health

Major Professor

Yiliang Zhu

Keywords

benchmark dose, modeling, methodology, neurotoxicity, dose response

Abstract

In its latest neurotoxicity guidelines released by the US EPA Office of Prevention, Pesticides and Toxic Substances (OPPTS) in 1998, it is recommended that in a neurobehavioral testing, at a minimum, for acute studies, observations and activity testing should be made before the initiation of exposure, at the estimated TOPE (time of peak effects) within 8 hrs of dosing, and at 7 and 14 days after dosing. It is recommended that estimation of TOPE be made by dosing pairs of rats across a range of doses and making regular observations of gait and arousal. However it is well known that TOPE may vary with end points or exposure conditions. In order to derive quantitative safety measures such as the benchmark doses (BMD), dose-time-response modeling must be done first and a model-based estimate is then implied. In many cases, the overall BMD corresponds to a TOPE estimate.

In such cases a substantial variation in the TOPE estimate in turn may result in substantial variation in BMD estimate. Therefore a reliable statistical estimate of TOPE is crucial to the correct determination of BMD. We therefore performed simulation studies to assess the impact of the experiment-based TOPE on the statistical estimation of the true TOPE on the basis of a fitted dose-time-response model. The simulation allows for the determination of the optimal timing range for the 2nd testing. The results indicated that given only four repeated observations, the optimal second testing time was at about midway between time zero and the true TOPE. Choosing the second testing time at the TOPE may not generate statistical estimates closer to the true TOPE.

Share

COinS