The systematic review by Dom et al (Reference Dom, Sabbe and Hulstijn2005) of studies using behavioural decision-making tasks and/or neuroimaging techniques to investigate orbitofrontal cortex functioning in substance use disorders was comprehensive. Our research article ‘Risk-taking on tests sensitive to ventromedial prefrontal cortex dysfunction predicts early relapse in alcohol dependency’ (Reference Bowden-Jones, McPhillips and RogersBowden-Jones et al, 2005) was published simultaneously and, because of its relevance to the review, we considered it important to bring it to readers’ attention.

We not only used most of the neuro-psychological tests mentioned by Dom et al but, more importantly, rated participants on both the Rogers Cambridge Gamble Task (RCGT; Reference Rogers, Owen and MiddletonRogers et al, 1999) and the Iowa Gambling Task (IGT; Reference Bechara, Damario and DamarioBechara et al, 1994), and on the Barratt Impulsivity Scale and two personality scales: the Structured Clinical Interview for DSM–III–R (Reference Spitzer, Williams and GibbonSpitzer et al, 1989) and the Dimensional Assessment of Personality Pathology–Basic Questionnaire (Reference Livesley and JacksonLivesley & Jackson, 2002).

The 21 participants in our study were in-patients in a residential detoxification unit and we were therefore able to carry out tests at 21 days post-detoxification in the knowledge that they had been substance-free during that period. They were followed up for 3 months post-discharge.

The six patients who relapsed early were significantly younger and more impulsive on the Barratt Impulsivity Scale, they sampled significantly more cards from the bad decks on the IGT and consistently risked more points across all odds on the RCGT. Hence people who had recently undergone detoxification were more likely to relapse within 3 months if they made more choices on a gambling task in which the immediate reward was large but the long-term consequences were disadvantageous.

It is unlikely that these findings reflect alcohol-induced brain damage because these people showed no impairments on a memory test sensitive to the early stages of dementia and on tests of dorsolateral prefrontal cortex functioning, which is particularly affected by long-term alcoholism.

Our results are consistent with the hypothesis that a dysfunctional orbitofrontal prefrontal cortex mediates the inability to resist the impulse to drink. This may lead a person to relapse after treatment despite the ultimately deleterious effects and despite the many hours of psychological input associated with a rehabilitation programme.

Relapse after detoxification is an area in need of further research. If it has a biological basis we need simple tests that are able to predict vulnerability to relapse and treatment programmes which are able to identify those patients at greater risk.