How valuable are animal models in defining antidepressant activity?

Animal models of depression have been utilised to screen novel compounds with antidepressant potential although uncertainty lingers concerning their clinical relevance. In order for a model to be considered of any value, it must possess predictive validity (does drug action in the model correspond to that in the clinic?), face validity (are there phenomenological similarities between the model and the clinic?) and construct validity (does the model possess a strong theoretical rationale?). On the one hand, there are models based on stress such as the learned helplessness model, the forced swimming test and the chronic mild stress model and, on the other hand, models based on neuronal deficits such as the olfactory bulbectomy model. To date, among models more frequently used in depression, none of them meet all these criteria. Moreover, improvements to tests are often poorly validated and estimating time of onset of action of antidepressants remains a major challenge in animal model research. Finally, reproducing the tests outside the laboratory of origin continues to be problematic and leads to variability in results. Although animal models of depression fail to be unequivocally valid, they represent the best tool to define potential antidepressant activity of drugs, to investigate their mechanism of action and, to a greater extent, explore this complex heterogeneous illness. Copyright 2001 John Wiley & Sons, Ltd.