Article révisé par les pairs
Résumé : Mechanisms that mediate the transition from occasional, controlled, drug use to the impaired control that characterizes severe dependence are still a matter of investigation. The etiology of substance use disorders (SUDs) is complex, and in this context of complexity, the concept of "endophenotype, " has gained extensive popularity in recent years. The main aim of endophenotypes is to provide a simpler, more proximal target to discover the biological underpinnings of a psychiatric syndrome. In this view, neurocognitive and neurophysiological impairments that suggest functional impairments associated with SUDs have been proposed as possible endophenotypes. Because of its large amplitude and relatively easy elicitation, the most studied of the cognitive brain event-related potentials (ERPs), the P300 component, has been proposed as one possible candidate. However, if a P300 amplitude alteration is a common finding in SUDs, it is also observable in other psychiatric afflictions, suggesting that the associations found may just reflect a common measure of brain dysfunction. On this basis, it has been proposed that a multivariate endophenotype, based on a weighted combination of electrophysiological features, may provide greater diagnostic classification power than any single endophenotype. The rationale for investigating multiple features is to show that combining them provides extra useful information that is not available in the individual features, leading ultimately to a multivariate phenotype.The aim of the present article is to outline the potential usefulness of this kind of "combined electrophysiological procedure" applied to SUDs. We present a review of ERP studies, combining data from people with SUD, family members, and normal control subjects, to verify whether the combination of 4ERPs (P50, MMN, P300, and N400) may produce profiles of cortical anomalies induced by different types of SUD (alcohol vs cocaine vs cannabis vs heroin). © EEG and Clinical Neuroscience Society (ECNS) 2013.