gms | German Medical Science

The statistical analysis of reproductive toxicity studies with respect to possible developmental effects in the offspring of treated female rodents is challenging due to the possibility of (intra-)litter effects. Endpoints such as pup weight might be dependent not only on the treatment itself, but also on other parameters (e.g. litter size). Hence, there is the possibility of statistical dependence between the pups of the same litter. Possible (intra-)litter dependent effects are not appropriately considered (if at all) in widely used test guidelines which consider the dams, litters and pups as independent experimental units. The use of mixed-effects models, which combine e.g. litter effects as a random effect and treatment effects as fixed effects, has been proposed in the literature. Up to now, mixed-effects modelling is rarely used in a regulatory context for the statistical analysis of developmental effects in reproductive toxicity studies.

The process and application of suitable mixed-effects models to a reproductive toxicity study with oral administration (gavage) of perfluorohexanoic acid (PFHxA) in rodents will be discussed with respect to model selection, model specification, derivation of observed effect levels and model validation.

Results of this study will be compared with the statistical outcome from a published study, in which the litter effect was neglected.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.