Here's a link to an interesting article from Biological Psychiatry this month:
It is based on the notion that the decision to engage in an addiction is often made based on a short-term, possibly impulsive, analysis of benefits and risks; consequently, longer-term risks or benefits associated with the behaviour are undervalued. This phenomenon is termed "delay discounting." Resistance to delay discounting could be considered a cognitive faculty that would help, on an intellectual level, with making a healthy decision in the face of strong impulses in the moment.
The authors note a relationship between addictive disorders and increased delay discounting. They also note a previously described relationship between delay discounting and reduced working memory function.
Based on these relationships, they did a controlled study of persons with stimulant addiction, in which the active group did a set of memory training exercises for 1-2 months. They found that the memory exercises led to improved (reduced) delay discounting.
This study does not show that memory exercises directly improve the course of addictive disorders; but it does present a promising therapeutic idea which I think is currently underutilized in the therapeutic community, not only for addictions but for other types of problems.
Cognitive exercises could have a variety of benefits for various psychological problems:
1) the improvement one would see with practice could help with self-esteem
2) arguably, the exercises would favourably alter the balance between executive function and visceral, limbic emotional drives (which could often be turbulent or disruptive)
3) the exercises could be an introduction to the various mental and physical disciplines required to effect psychological or behavioural change
In terms of the specific exercises used in this study, I do think that the number of practice sessions was far too small. I believe that most psychologically beneficial activities start to show substantial results after 50-100 hours of practice. This study used only a maximum of 15 training sessions. The memory practice itself could have been organized in a more engaging, game-like manner. I think of some quite unique working memory games from the lumosity.com website, which tap into a type of activity most people would rarely work on directly, but yet are quite entertaining and allow gradual progress.
In summary, this was an interesting article looking at the promising theme of using cognitive training exercises as part of the treatment of a psychological problem. This is a relatively new idea, showing up only a few other times so far in the research literature.