Three decision-making tasks in cocaine-dependent patients: do they measure the same construct?

FRANÇAIS :
La réponse à deux variables, le risque et le délai, est étudiée chez des usagers de drogues afin d'explorer leur importance dans le comportement addictif (différence individuelle et vulnérabilité). Trois tests sont effectués : la DDP (Delay Discounting Procedure) pour laquelle la décision porte entre une récompense petite mais immédiate et une récompense plus grande mais retardée ; la GT (Gambling Task) pour laquelle l'alternative est entre récompense ou punition ; et la RDMT (Rogers Decision-Making Task) pour laquelle le choix est entre des probabilités de jeu plus faibles ou plus grandes. Pour les 32 participants, les performances sur la GT étaient corrélées aux performances sur le DDP. Les temps de réaction sur la RDMT étaient corrélés aux performances sur les GT et DDP. Cependant aucune corrélation n'existe entre les performances et l'impulsivité. Il existe un certain recoupement entre les fonctionnements de la prise de décision mis en évidence par ces tâches, mais les résultats ne permettent pas d'établir le lien avec le risque, le délai, ou la combinaison des deux.
ENGLISH:
Aims. Substance-abusing populations perform poorly on decision-making tasks related to delay and risk. These tasks include: (1) the Delay Discounting Procedure (DDP), in which choices are made between smaller-sooner and later-larger rewards, (2) the Gambling Task (GT), in which choices are made between alternatives varying in pay-off and punishment, and (3) the Rogers Decision-Making Task (RDMT) in which subjects choose between higher or lower probability gambles. We examine the interrelationship among these tasks. Design. A test battery was created which included the DDP, GT and RDMT, as well as measures of impulsivity, intellectual functioning and drug use. Setting. Subjects completed the test battery at an outpatient center, prior to beginning 12 weeks of treatment. Participants. Thirty-two treatment-seeking cocaine dependent individuals (primarily African-American males) participated. Findings. Performance on the GT was significantly correlated with performance on the DDP (r = 0.37; p = 0.04). Reaction times on the RDMT correlated with performance on the GT (r = 0.36, p = 0.04) and DDP (r = 0.33, p = 0.07), but actual choices on the RDMT did not (p > 0.9 for both). While no significant relationships were observed between task performance and impulsivity, IQ estimate was positively correlated with both the GT (r = 0.44, p = 0.01) and RDMT (r = 0.41, p = 0.021). Split half reliability data indicated higher reliability when using only data from the latter half of the GT (r = 0.92 vs. r = 0.80). Conclusions. These data offer preliminary evidence of overlap in the decision-making functioning tapped by these tasks. Possible implications for drug-taking behavior are discussed. (Author' s abstract)