Cocaine users 1) are more impulsive based on self-report measures, 2) have more difficulty inhibiting behavioral responses, 3) make poorer decisions in decision-making tasks, and 4) discount the value of future rewards at a faster rate than non-cocaine users, suggesting that impulsivity plays a role in cocaine abuse. After collecting data on these 4 domains in current cocaine abusers (n = 200) and matched controls (n = 180), we will conduct confirmatory factor analyses testing a second-order factor model of Impulsive Choice. We hypothesize that 3 first-order latent factors (personality/behavioral inhibition, risk taking, reward/punishment value) will account for the significant relations among the measures within each factor, and a single latent second-order factor (impulsive decision-making) will account for the relationships among the 3 first-order factors. We will test the abilit of our model to predict real world cocaine use and clinical outcome.
Aim 1. Obtain self-report and task-based measures of 1) impulsivity~ 2) response inhibition~ 3) decision- making~ and 4) temporal discounting in cocaine abusers and matched controls in order to examine the correlational structure among the measures and test our model of Impulsive Choice.
Aim 2 a. Compare the structural model and the scores for each factor between groups. We hypothesize that the 4 latent factors of our model of Impulsive Choice will be observed in both groups, but cocaine abusers will have greater scores on all factors.
Aim 2 b. Evaluate the predictive validity of our factors with respect to cocaine use. We hypothesize a positive relationship between our factors and current cocaine use.
Aim 3. Compare the structural model and the scores for each factor between sexes. We hypothesize that the 4 latent factors of our model of Impulsive Choice will be observed in both sexes with the following rank order of factor scores: Female Cocaine Abusers > Male Cocaine Abusers > Male Controls > Female Controls.
Aim 4 a. Evaluate the predictive validity of our factors in 35 cocaine users with respect to neural measures of response inhibition and temporal discounting.
Aim 4 b. Evaluate whether the factors predict neural activity associated with decisions to smoke cocaine measured using a novel fMRI task (CNET) in 35 cocaine users. We hypothesize that our factors, with incorporation of neural measures of response inhibition and temporal discounting, will predict neural activity related to decisions to smoke cocaine and self-reported cocaine use.
Aim 5. Evaluate the predictive validity of our factors with respect to treatment outcome for 100 participants for whom we will provide cognitive-behavioral treatment (CBT) for their cocaine use. We hypothesize that there will be a negative relationship between our factors and clinical outcome. Impact: A better understanding of the construct of impulsivity and which factors of impulsivity best predict clinica outcome will provide critical information to guide the development of therapies specifically targeting cognitive processes associated with those factors in drug taking.
We propose a model that will identify four latent factors related to Impulsive Choice derived from neurobehavioral domains that are commonly used to assess impulsive choice. We will compare cocaine abusers to healthy controls on our measures of impulsive choice and then determine the validity of our model in predicting actual cocaine use and success of cocaine treatment. If our model is confirmed, we can then design studies to target these neurobehavioral domains in order to decrease impulsive choice in a manner that could have important public health consequences across a range of 'impulsive' behaviors: binge eating, drug and alcohol use, risky sexual behavior.
