Rationale: Previous research in humans suggests a relationship between drug abuse and impulsivity as shown by selection of a smaller immediate reward over a larger delayed reward. However, it is not clear whether impulsivity precedes drug abuse or drug abuse influences impulsivity. Objective: The hypothesis of the present experiment was that rats selected for choosing smaller, immediate over larger, delayed food would acquire IV cocaine self-administration faster than those choosing larger, delayed food rewards. Methods: Female rats were screened for locomotor activity and trained on a delay discounting procedure that allowed them access to two response levers and a food pellet dispenser. Under a fixed-ratio (FR) 1 schedule, responding on one lever resulted in immediate delivery of one 45 mg pellet, while responding on the other lever resulted in delivery of three 45 mg pellets after a variable delay that increased after responses on the delay lever and decreased after responses on the immediate lever. For each rat, a mean adjusted delay (MAD) was calculated for each daily session, and stability was defined as MADs varying less than 5 s across 5 days. Based on their average MADs, rats were separated into low impulsive (LoI) and high impulsive (HiI) groups, implanted with an indwelling IV catheter, and trained to lever press for cocaine (0.2 mg/kg) under an FR1 schedule. Results: There were no differences in locomotor activity between the LoI and HiI groups; however, a greater percentage of the HiI group acquired cocaine self-administration, and they did so at a significantly faster rate than the LoI rats. Conclusions: Performance on the delay discounting model of impulsivity predicted vulnerability to subsequent acquisition of cocaine self-administration.
Bibliographical noteFunding Information:
Acknowledgements The authors would like to thank Mara Zoe Feliciano, Christina Gremel, Annemarie Loth, Andrew D. Morgan, and Isadora Rosario-Gonzalez for their technical assistance, and Andrew D. Morgan for careful reading of the manuscript. We are also grateful for support contributed by NIDA grants R01 DA03240, K05 DA15267 (M.E.C.), and R03 DA13575 01 (G.J.M.).