Are animal models of addiction useful?

Abstract

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.

The relevance of animal models of addiction

“…Clinicians and basic researchers share the view that certain life conditions favor expression of predispositions to psychopathology, including addictive behaviors . This context‐dependent expression of vulnerability is compatible with both the network models of psychiatric disorders and the opioid crisis, contradictory to what is stated by Field & Kersbergen . It is also not incompatible with mapping expression of addiction to brain function , together with exploring the role of psychological risk factors contributing to its development.…”

“…sucrose) . Models using choice with alternative reinforcers are therefore not a recent development . An evolution came with the combination of the two types of models (the concomitant choice between punished cocaine use and sucrose) , and what is highly novel is the use of a social interaction as an alternative reinforcer.…”

“…Also, the idea that prevailing animal models portray addiction as a disorder of habit and compulsion is only partially true. Habit has been quickly discarded as a player in cocaine addiction‐like behavior as measured in the multi‐symptomatic model (called here 0/3 criteria) , and as confirmed by the puzzle‐solving procedure.…”

An imperfect model is still useful

“…Field & Kersbergen raise important issues regarding the utility of animal models for understanding and developing new treatments for substance use disorder (SUD). Their argument revolves around four basic tenets: many ‘treatments’ developed in animal models have failed, hence represent a waste of resources; animal models commonly represent drug‐seeking as an inflexible and habitual behaviour, which is misleading; crucial aspects of SUD, such as language, can never be represented in an animal model, and addiction can only be understood in the context of complex society.…”

Lasting translation: how to improve animal models for addiction treatment

“…establishing test consortia. This may effectively reduce the rate of false positives candidates that later fail in the human setting [14,15].…”