Animal model

An animal model is a non-human organism — most often a rat or mouse, occasionally a primate or zebrafish — used in drug research to study mechanisms that cannot be safely or ethically measured in people. In the evidence hierarchy, animal studies sit before human trials: they generate the hypotheses and safety estimates that justify moving to clinical research.

Most of the pharmacological vocabulary on these pages — tolerance curves, reward-threshold shifts, downregulation — was first traced in animal work. That makes animal models foundational to what is known, while also making it important to understand exactly what the design can and cannot say.

What this design can establish

A well-designed animal study can establish that a mechanism operates in that species under those conditions. Self-administration paradigms — where an animal presses a lever to receive a drug — establish whether a compound has reinforcement, the closest behavioural proxy for abuse liability that preclinical research offers.

Animal studies also establish neurotoxicity: whether a compound damages neurons, which brain regions are affected, and how the pattern relates to exposure level within that species. These signals inform what researchers carry into early human trials, setting the biological plausibility for the next stage of evidence.

What it cannot

What an animal model cannot establish — regardless of how rigorously it is run — is that the finding transfers to humans. Species differ in metabolism, density, blood-brain barrier permeability, and environmental context. A compound that is neurotoxic in rodents may or may not produce the same damage in a person.

The inference a reader is most likely to make — and should resist — is that animal-derived data directly predicts human risk. It establishes a signal in the model species, not a verdict for any other. Many substances flagged as harmful in animal literature are used by people without those predicted outcomes; others proved more dangerous than the models suggested.

The honest reading of animal data is: this mechanism can operate in this organism. Whether it operates in humans, and at what level of exposure, requires human evidence to answer. An absent human study is a stated gap in the record — not a reason to treat the animal finding as settled.