ABSTRACT
In most developmental toxicity studies, exposure is administered to the dam, rather than directly to the developing foetuses. Because of genetic similarity and the same treatment conditions, offspring of the same mother behave more similar than those of another mother. This has been termed “litter effect” and is one important form of clustering. There are several ways to handle clustering. While dose-response modeling is relatively straightforward in uncorrelated settings, it is less so in the clustered context. Of course, one can ignore the clustering altogether by treating the littermates as if they were independent. However, this will in general be too strong an assumption. Also, the litter effect issue can be avoided by modeling the probability of an affected cluster via, e.g., a logistic regression model. Such models are generally too simplistic but there is a multitude of models which do consider clustering.
