• Addiction;
  • genetics;
  • neurobiology

The paper by Harold Kalant, ‘What neurobiology can not tell us about addiction’[1], rehashes arguments regarding the limitation of neurobiology research in understanding addiction. We agree that addiction research is limited by complexity at all levels of analysis, but we respectfully disagree with several key issues raised by the author.

Let us start with the author's first question: ‘What is addiction?’. The author moves quickly past the DSM-IV-TR and other definitions to a statement by an ad hoc committee of the Royal Society of Canada that: ‘the only elements common to all definitions of addiction are a strongly established pattern of repeated self-administration of a drug in doses that reliably produce reinforcing psychoactive effects, and great difficulty in achieving voluntary long-term cessation of such use, even when the user is strongly motivated to stop’.

The author concludes that the key phrase is ‘self-administration’ and that ‘addiction is not produced by a drug, but by self-administration of a drug’. He supports his conviction by contrasting the relatively low likelihood that a pain patient will become addicted when administered opioids by a health care professional, with the anecdote that Civil War veterans allowed to self-administer opiates often developed ‘soldier's disease’. [2] He acknowledges differences between pain patients and the soldiers, but his conclusion remains that the only important difference is self-administration. He concludes from this that models of addiction not based on self-administration are basically ‘flawed’.

That addiction does not occur in all patients administered opioids by health care professionals is clear, but some do become addicted. At the same time, a great majority of patients who self-administer opioids do not become addicted. That said, we point to the many important factors that account for such differences beyond self-administration such as environmental contexts (e.g. war veterans often experienced severe traumas during and following the war, experienced high unemployment and possibly mental health problems), the likelihood and importance of co-administered substances, high levels of social disruption and displacement, as well as other pre-existing vulnerabilities (genetic and otherwise). We also point to conceptual/integrative models such as proposed by Koob and colleagues that place such factors into a wider context [3].

We agree on the need to look at genetic and environmentally engraved phenotypes as drivers of addictive behavior. Differences between modes of drug administration have been researched extensively at the molecular, cellular and behavioral levels, and researchers are far from naive. For example, Robinson and colleagues demonstrated marked differences between experimenter and self-administered opioids in alteration of neuronal spine density in several brain regions [4]. That experimenter-administered drug, which results in sensitization and place preference, is completely ‘flawed’ for understanding aspects of addiction is not our opinion.

We agree that study of the genetics of addiction has the best chance in the areas of socially accepted drugs such as alcohol and nicotine. Recent identification of a region of chromosome 15 (containing the α5-α3-β4 nicotinic receptor genes) separately in smokers and lung-cancer patients provides evidence that genetic tools will have power to identify aspects of susceptibility [5]. Also, success of treatment options has gained insight from genetics with the finding that naltrexone appears to be more successful in alcoholics with selective alleles of the mu-opioid receptor [6] The numbers needed for genome-wide association studies for relatively straightforward phenotypes has been greater than expected because the effect size for specific genes is often exceedingly small [7]. The implication is that it will be extremely difficult to identify the genetic vulnerability to illicit drugs and complex psychological endophenotypes underlying susceptibility to addictive behavior [8]. The author makes the important point that the control of proteins are at the heart of the issue, and the opportunity for genetic influence at multiple loci to modulate proteins has increased markedly as we begin to understand the complexity of epigenetics, promoters, protein trafficking/turnover and microRNA regulation.

Finally, we agree that reductionistic or analytical approaches are placed into a more appropriate context when corralled within appropriate and integrated conceptual frames. This is true for all areas of science. The challenge is certainly one of making the science ‘fly’, as nothing brings focus like a plane about to crash. We agree that the field needs to think more conceptually and integratively, and we look forward to flying with Dr Kalant in the future.


  1. Top of page
  2. Declarations of interest
  3. References