The repeated failure of animal models to yield findings that translate into humans is a serious threat to the credibility of preclinical biomedical research. The use of animals in research that lacks translational validity is unacceptable in any ethical environment, and so this problem needs urgent attention. To reproduce any human illness in animals is a serious challenge, but this is especially the case for psychiatric disorders. Yet, many authors do not hesitate to describe their findings as a 'model' of such a disorder. More cautious scientists describe the behavioural phenotype as 'disorder-like', without specifying the way(s) in which the abnormal behaviour could be regarded as being analogous to any of the diagnostic features of the disorder in question. By way of discussing these problems, this article focuses on common, but flawed, assumptions that pervade preclinical research of depression and antidepressants. Particular attention is given to the difference between putative 'models' of this illness and predictive screens for candidate drug treatments, which is evidently widely misunderstood. However, the problems highlighted in this article are generic and afflict research of all psychiatric disorders. This dire situation will be resolved only when funders and journal editors take action to ensure that researchers interpret their findings in a less ambitious, but more realistic, evidence-based way that would parallel changes in research of the cause(s), diagnosis and treatment of psychiatric problems in humans.
Keywords: Forced Swim Test; animal model; depression and antidepressants; endophenotype; predictive drug screen; psychiatry; psychopharmacology; translation; validity.