Alcohol dependence is a chronic remitting disorder—for some; but clearly not for all, as Cunningham & McCambridge observe in their thought-provoking paper [1]. Several epidemiological studies around the world, at least the western world, confirm that the majority of people who at one point in time fulfil the criteria for alcohol dependence no longer do so a few years later on. It is also evident that this disorder, like most other disorders, varies in severity. Most people seem to develop it in a mild form, typically in young adulthood, and then appear to mature out of their drinking problems. In a recent Swedish study [2], approximately 75% of all subjects with dependence had a mild or moderate form of dependence, with three or four DSM-IV dependence criteria, and a minority, approximately 25%, had a more severe form, with five to seven criteria. How these epidemiological observations connect with neurobiological research is still unclear. In spite of impressive advances in this field we still do not understand why some people do not develop dependence at all, in spite of considerable exposure to alcohol, and how clinical severity and neurobiological measures of impairment correlate [3].

Cunningham & McCambridge make an important point where they observe that most clinical and biological addiction research has been based on clinically selected samples. This limits generalizability, as the vast majority of people with alcohol dependence do not seek or receive treatment [4]. Those who do tend to have more severe forms of addiction and/or more psychiatric comorbidity and other psychosocial trouble. Perhaps it is only in this group that significant structural changes in the brain or changes in biochemistry will be found—and perhaps it is mainly in this group that the chronic remitting form of alcohol dependence is found. More studies are required on the less severe end of the alcohol problems continuum to clarify this.

One serious consequence of present popular conceptions about dependence and its treatment is that they leave the majority of alcohol-dependent people without help from the treatment system. This is troublesome for the individuals concerned and their families, as well as for the wider community. As Cunningham & McCambridge rightly observe, the epidemiological paradox applies to the alcohol field as much as to other health fields, where most of the suffering and costs are generated in the majority with lower severity.

There has been general agreement for at least 20–30 years that treatment for alcohol problems needs a broader base in the community [5]. However, so far little progress has been made in this regard. Despite impressive evidence of effectiveness, screening and brief intervention for alcohol problems has not yet been integrated into routine practice. One reason for this failure is probably the concern that screening will reveal not only hazardous or harmful drinking, but alcohol dependence as well, and that this disorder is considered too complicated to manage in primary care. This view of alcohol dependence is largely a result of clinical research based on the most severe cases.

Clinical research, experiences among practitioners, popular perceptions of addiction, all contribute to a vicious circle. Drinking problems are managed mainly without treatment because treatment is unattractive. The barriers to treatment are overcome only when the trouble becomes very severe, reinforcing the perception that alcohol dependence is a very severe disorder. As this view is shared by general practitioners, other non-specialists and the general population alike, only people with the chronic remitting form of the disorder will be found in treatment—and will form the basis for further studies of the disorder.

What is needed now is, on one hand, neurobiological research on the majority of alcohol-dependent people who do not seek treatment and, on the other hand, research on how people in this group view their condition and what kind of help—if any—they want. A reasonable hypothesis is that many people with alcohol dependence realize that they are drinking too much, but reject help from the formal addiction treatment system because they do not identify with the majority of severely dependent, socially marginalized people who are being treated today. For practitioners, the major challenge is to make treatment for alcohol dependence more attractive to the large majority. Researchers have the responsibility to convey to practitioners a correct picture of alcohol dependence as a mainly time-limited disorder of mild or moderate severity and equip them with adequate tools to manage this disorder.

Declaration of interests