Abstract: Previous researches adopting the Delay Discounting Task (DDT) in behavioral economics suggested that the quick decay for delayed reinforcer is a behavioral process common (Bickel et al., 2007) to a variety of ad-diction such as alcohol (Petry, 2001), nicotine (Mitchell, 1999; Bickel, Odum, & Madden, 1999; Reynolds, Karraker, Horn, & Richards, 2003; Baker, Johnson, & Bickel, 2003; Reynolds, Richards, Horn, & Karraker, 2004; Reynolds, 2004), cocaine (Coffey, Gudleski, Saladin, & Brady, 2003), and opioid or heroin (Madden, Pe-try, Badger, & Bickel, 1997; Madden, Bickel, & Jacobs, 1999; Kirby, Petry, & Bickel, 1999; Odum, Madden, Badger, & Bickel, 2000; Giordano et al., 2002; Zhang, Zhou, Li, & Shen, 2008). It was also discovered that, relative to a general reinforcer (for example money), drug dependence individuals show higher delay discount-ing rate in dependence drug reinforcer such as drink(Petry, 2001), cigarette (Bickel et al., 1999; Field, Santar-cangelo, Sumnall, Goudie, & Cole, 2006; Field, Rush, Cole, & Goudie, 2007), cocaine (Coffey et al., 2003), and heroin (Madden et al., 1997; 1999; Kirby et al., 1999; Odum et al., 2000). However, divergence still exists in researchers’ explanations for the nature of super-quick decay for dependence drug in drug dependence individu-als. A relative previous viewpoint (that may be called the hypothesis of avoidance of withdrawal symptoms) claimed that the super-quick decay for dependence drug is a consequence of escaping from or avoidance of withdrawal symptoms in drug dependence individuals (Madden et al., 1997; Bickel et al., 1999). Alternatively, some other researchers claimed that the steep discounting of drug of abuse compared to monetary reinforcers may occur as part of a general process by which primary reinforcers (e.g. drugs or food) are discounted more steeply than conditioned reinforcers (e.g. money), rather than as a unique aspect of addiction (Odum & Rainaud, 2003). This viewpoint may be called the hypothesis of unconditioned reinforcement effect. However, the reasons underlying the differences in discounting rate between primary and secondary reinforcers are not yet clear, so it is necessary to uncover the mechanisms by which conditioned reinforcers lose value more rapidly with delay than primary reinforcers (Odum & Rainaud, 2003). The present study explores the psychological mechanisms of tendency of super-quick decay in heroin abstainers during different abstinence phases to delayed heroin rein-forcer by the DDT, in which the measurement of craving in cue-exposure situations taken as the validity index of DDT evaluation.
In the present study, 124 male heroin abstainers (17~39 years old, M=29.27, SD=5.00) during different ab-stinence phases (1~29 months, M=12.68, SD=7.58) were selected to administer Chinese computerized version of the DDT which was based on monetary rewards and heroin rewards, respectively. In the DDT based on mone-tary rewards, subjects had to choose between hypothetical monetary rewards available immediately or following a delay. Delayed rewards were ¥10,000 in value; and in the DDT based on heroin rewards, subjects had to choose between hypothetical heroin rewards available immediately or following a delay. Delayed rewards were 250 bags, which was equal to ¥10,000 in value. The immediate-reward amount would be adjusted until choices reflected indifference in the DDT based on monetary rewards or heroin rewards, and these processes were re-peated at each of eight delays (6 hours to 25 years). Moreover, 28 of 124 heroin abstainers were selected to elicit heroin desiring by the cue-exposure paradigm.
The results of the present study showed that (1) Hyperbolic discounting functions provided a good fit for the subjects discounting data of both delayed reinforcement conditions, and median R2 was from 0.91 to 0.90 for monetary and heroin rewards. (2) Relative to monetary rewards, the delay discounting rates (k) of heroin ab-stainers were much higher for heroin rewards along with the abstinence time course, Z = 7.688, p<0.001. (3) The k based on the DDT in heroin abstainers did not improve along with the abstinence time course, c2money=0.603, df=2, p>0.10, c2heroin=0.806, df=2, p>0.10. (4) The delay discounting rates (k) for heroin rewards were correlated with the heroin craving on the cue-exposure paradigm, r=0.47, df=26, p<0.01, and heroin craving of super-quick decay group was much higher than that of low-quick decay group, t=2.343, df=26, p<0.05. (5) The delay dis-counting rates (k) for heroin rewards were correlated with the k for monetary rewards, r=0.52, df=122, p<0.001, and k of higher impulsive group was much higher than that of low impulsive group, U=209.500, p<0.001.
From the present study we concluded that (1) Heroin abstainers are characterized in tendency of su-per-quick decay for delayed reinforcement to heroin in the delay discounting task, and the characteristics in heroin abstainers do not improve along with the abstinence time course. (2) The tendency may be an interac-tional outcome of heroin abstainers’ trait impulsivity and heroin cue eliciting, and their accelerated amount rela-tive to monetary rewards in DDT may be manifested as the heroin cue-eliciting effect.

Key words: heroin abstainers, the delay discounting task, super-quick decay, cue-eliciting, heroin craving