Are animal models of addiction useful?

Background: Pre-clinical research involving non-human animals has made important contributions to our understanding of risk factors for addiction, neuroadaptations that follow chronic drug exposure and to the development of some efficacious pharmacotherapies for addiction. Despite these contributions, we argue that animal models of addiction have impeded progress in our understanding of addiction and its treatment in humans. Argument: First, the majority of pharmacological treatments that were initially developed using animal models have failed to prove effective for the treatment of addiction in humans, resulting in a huge waste of resources. Secondly, we demonstrate that prevailing animal models that portray addiction as a disorder of compulsion and habit cannot be reconciled with observations that psychoactive drug use in humans is a goal-directed operant behaviour that remains under the control of its consequences, even in people who are addicted. Thirdly, addiction may be a uniquely human phenomenon that is dependent on language, which necessarily limits the validity of animal models. Finally, we argue that addicted brains must be understood as one component of broader networks of symptoms and environmental and social factors that are impossible to model in laboratory animals. Conclusions: A case can be made that animal models of addiction have not served us well in understanding and treating addiction in humans. It is important to reconsider some widely held beliefs about the nature of addictive behaviour in humans that have arisen from the zeal to translate observations of laboratory animals.