Dose response assessment in a behavior learning curve experiment

Abstract

A new animal model of diseases affecting cognitive functioning is set up. A pharmacological deficit is induced by the administration of compound X. A good understanding of the effect of compound X on cognitive functioning is therefore crucial.Two dose response studies with compound X were pooled and a meta-analysis was performed. Longitudinal data was available on the proportion of correctly executed trails from an animal behavior experiment from 96 male wistar rats during a period of 14 days. The aim of the report is to study and quantify the dose effect of compound X on learning behavior. First, the complete profiles were modeled by a non-linear model. The time until proportion 0.7 was reached (T70) was the most important characteristic of learning behavior to study the dose effect of compound X. T70 prolongs with increasing dose level. Secondly, a time-to-event analysis was performed. T70 was estimated at different dose levels with less precision compared to the non-linear model.