Intermittent cocaine use patterns, not total intake, predict cue-induced drug seeking in rats

Cocaine-associated cues trigger relapse to drug use in humans and animal models. In rats, long daily cocaine access (4-6 vs. 1-2 h) increases drug self-administration and cue-induced cocaine seeking, suggesting that greater intake promotes relapse. However, prior studies used continuous drug access paradigms, whereas human cocaine use is typically intermittent. Prior studies also used conditioned stimuli (CS), whereas discriminative stimuli (DS) are more effective in triggering drug seeking. Here, we used intermittent access (IntA) self-administration to examine how session length influences CS- and DS-induced cocaine seeking.

Female rats self-administered cocaine intermittently during daily 2- or 4-h IntA sessions (Short-IntA and Long-IntA), with alternating DS+ (cocaine available; 5 min) and DS- (no cocaine; 25 min) periods. Lever pressing during DS+ produced cocaine and a CS+; lever pressing during DS- produced only a CS-. After 4 weeks of abstinence, rats received a test where all cues were presented response-independently, and lever presses-which had no consequence-measured cocaine seeking.

Long-IntA rats took twice more cocaine than did Short-IntA rats. In both groups, DS+ but not CS+ later triggered significant increases in cocaine seeking, with no group differences. Thus, total intake did not predict cue-induced relapse intensity. However, individual cocaine intake patterns did, including hourly consumption, episodes of burst-like self-administration and latency to self-administer.

Under intermittent-access conditions, individual cocaine-use patterns, not cumulative intake, predict vulnerability to cue-induced relapse. This highlights the importance of individual drug-taking profiles in relapse risk assessment.