T.-J. Chiang

Theory and method in the quantitative analysis of ”impulsive choice” behaviour: implications for psychopharmacology

Abstract  Impulsive choice refers to the selection of small immediate gains in preference to larger delayed gains, or the selection of large delayed penalties in preference to smaller immediate penalties. Current theoretical interpretations of impulsive choice are reviewed, and a synthesis of these ideas, the ”multiplicative hyperbolic model of choice”, is presented. The model assumes that the value of a positive reinforcer increases as a hyperbolic function of its size, and decreases as a hyperbolic function of its delay and the odds against its occurrence. Each hyperbolic function contains a single discounting parameter which quantifies the organism’s sensitivity to the variable in question. The hyperbolic discounting functions combine multiplicatively to determine the overall value of the reinforcer. Equivalent functions are postulated to govern the (negative) value of aversive events, the net value of an outcome reflecting the algebraic sum of the positive and negative values. The model gives rise to a quantitative methodology for studying impulsive choice, based on a family of linear indifference (null) equations, which describe performance under conditions of indifference, when the values of the reinforcers are assumed to be equal. This methodology may be used to identify individual differences in sensitivity to the magnitude, delay and probability of reinforcement. The methodology is also suitable for the quantitative evaluation of the effects of some pharmacological interventions on discounting parameters. Recent psychopharmacological studies of impulsive choice are reviewed, and the utility of indifference equations for extending this work, and developing a quantitative psychopharmacology of impulsive choice is discussed.

Effect of central 5-hydroxytryptamine depletion on inter-temporal choice: a quantitative analysis

Abstract Rationale: It has been proposed that the ascending 5-hydroxytryptaminergic (5-HTergic) pathways are involved in ”impulse control”. Previous experiments have shown that rats whose 5-HTergic pathways have been destroyed are more liable than intact rats to select a smaller, immediate reinforcer rather than a larger, delayed reinforcer (impulsive choice). However, it remains unclear whether this effect of central 5-HT depletion reflects a change in the rate of time discounting (i.e. a change in the rate at which reinforcers become devalued as a function of delay) or a change in sensitivity to reinforcer size. Objective: We examined the effect of central 5-HT depletion on time discounting using a quantitative model of inter-temporal choice (multiplicative hyperbolic model), which enables effects on time discounting to be differentiated from effects on sensitivity to reinforcer size. Methods: Rats received injections of 5,7-dihydroxytryptamine into the dorsal and median raphe nuclei or sham lesions. They were trained to press two levers for food-pellet reinforcers in a discrete-trials adjusting-delay schedule. In free-choice trials, selection of lever A resulted in a brief fixed delay (dA) followed by delivery of one pellet; selection of lever B resulted in a longer variable delay (dB) followed by delivery of two pellets; dBwas adjusted in accordance with the subject’s choices. The value of dA was varied (0.5–8.0 s) in successive phasesof the experiment, and the indifference value of dB was determined in each case. Results: In both groups, the indifference value of dB was linearly related to the value of dA,in accordance with the multiplicative hyperbolic model. The lesioned group showed shorter indifference delays than the sham-lesioned group, this being reflected in a parallel displacement of the linear indifference function. In both experiments, the levels of 5-HT and 5-hydroxyindole-acetic acid were reduced in the brains of the lesioned rats, but the levels of noradrenaline and dopamine were not altered. Conclusions: According to the multiplicative hyperbolic model, parallel displacement of the linear indifference function uniquely specifies a change in time discounting. Thus these results indicate that central 5-HT depletion results in an increase in the rate of time discounting for food reinforcers.

Effect of central 5-hydroxytryptamine depletion on performance in the free-operant psychophysical procedure: facilitation of switching, but no effect on temporal differentiation of responding

Abstract This experiment examined the effect of destroying the ascending 5-hydroxytryptaminergic (5-HTergic) pathways on timing and switching behaviour in the free-operant psychophysical procedure. Rats received injections of 5,7-dihydroxytryptamine into the dorsal and median raphe nuclei or sham lesions. They were trained to press levers for sucrose reinforcement; sessions consisted of fifty 50-s trials in which reinforcers were available on a variable-interval 30-s schedule. In the first 25 s, of each trial, reinforcement was only available for responses on lever A; in the last 25 s, it was available only for responses on lever B. In phase 1 (70 sessions) repetitive switching between the levers was prevented by withdrawal of lever A after the first response on lever B in each trial; in phase 2 (40 sessions) this constraint on switching was removed; in phase 3 (40 sessions) the constraint was reinstated. Data were collected from probe trials (four per session) in which no reinforcers were delivered, during the last ten sessions of each phase. In all phases, both groups showed declining response rates on lever A and increasing response rates on lever B as a function of time from the onset of the trial. Response rate on lever B, expressed as percentage of overall response rate, could be described by a two-parameter logistic function. Removal of the constraint on switching reduced the slope of the function without changing the indifference point (time corresponding to 50% responding on lever B). The parameters of the timing function did not differ between the groups in any of the phases. However, the lesioned group showed a greater enhancement of switching rate during phase 2 than the control group. The levels of 5-HT and 5-hydroxyindoleacetic acid were reduced in the brains of the lesioned rats, but the levels of noradrenaline and dopamine were not altered. The results provide further evidence for the involvement of the ascending 5-HTergic pathways in switching between response alternatives, but cast doubt on our previous suggestion that the effects of 5-HT depletion on temporal differentiation of behaviour are mediated by facilitated switching.

Effect of 8-OH-DPAT on temporal discrimination following central 5-hydroxytryptamine depletion.

The 5-hydroxytryptamine (5-HT)(1A) receptor agonist 8-hydroxy-2-(di-n-propylamino)tetralin (8-OH-DPAT) alters performance in discrete-trials timing schedules. 5-HT(1A) receptors occur both presynaptically and postsynaptically, but it is not known which receptor population mediates the effects of 8-OH-DPAT on timing. Rats received intra-raphe injections of 5,7-dihydroxytryptamine (n=16) or sham lesions (n=14). They were trained in a discrete-trials psychophysical procedure in which levers were presented at a predetermined time after the onset of each trial (2.5, 7.5,., 47.5 s). A response on lever A was reinforced if lever presentation occurred < 25 s after trial onset; a response on lever B was reinforced if lever presentation occurred >25 s after trial onset. After 70 preliminary sessions, the rats received 8-OH-DPAT (25, 50, 100, 200 microg kg(-1) sc) and saline vehicle. The percentage of responses on lever B (%B) increased as a function of time from trial onset. Under the baseline (vehicle-treatment) condition, performance did not differ between the two groups. 8-OH-DPAT did not alter the indifference point (time corresponding to %B=50%), but dose-dependently increased the Weber fraction in both groups. Forebrain concentrations of 5-HT and 5-HIAA in the lesioned group were approximately 10% of control levels. The results suggest that the effect of 8-OH-DPAT on performance on discrete-trials timing schedules is mediated by postsynaptic 5-HT(1A) receptors.

Effect of central 5-hydroxytryptamine depletion on tolerance of delay of reinforcement: evidence from performance in a discrete-trials “time-left” procedure

Abstract This experiment examined the effect of destruction of the ascending 5-hydroxytryptaminergic (5HTergic) pathways on performance on a new discrete-trials version of the “time-left” procedure. Rats received either injections of 5,7-dihydroxytryptamine into the dorsal and median raphe nuclei or sham lesions. They were trained in a discrete trials schedule in which reinforcers were provided for responding on either of two levers, A and B. At a random time point, t s after the start of each trial, the two levers were inserted into the operant chamber; a response on A resulted in the delivery of one food pellet after dA s, whereas a response on B resulted in the delivery of two pellets after 84-t s. The value of dA was varied between 1 and 12 s in different phases of the experiment. Both groups showed an increasing tendency to respond on lever B as a function of time within the trial. Logistic functions were fitted to the data from each group, and a value of the “indifference point” (T50: the time within the trial at which proportional choice of B attained a value of 50%) was derived for each rat. For each value of dA, the values of T50 were significantly greater in the lesioned rats than in the control rats, reflecting a rightward shift of the logistic function in the lesioned group. The levels of 5HT and 5-hydroxyindoleacetic acid were reduced in the brains of the lesioned rats, but the levels of noradrenaline and dopamine were not significantly altered. The results provide further evidence for the involvement of the ascending 5HTergic pathways in the control of operant behaviour by delayed positive reinforcers.

Effect of central 5-hydroxytryptamine depletion on changeover behaviour in concurrent schedules of reinforcement

Abstract  Rationale: Previous experiments have shown that rats whose 5-hydroxytryptaminergic (5-HTergic) pathways have been destroyed exhibit higher rates of switching between response alternatives on various temporal differentiation schedules. Objective: This paper reports two experiments investigating the effect of central 5-HT depletion on switching between concurrent schedules of reinforcement which do not entail temporal differentiation of behaviour. Methods: Rats received injections of 5,7-dihydroxytryptamine into the dorsal and median raphe nuclei or sham lesions. They were trained to press levers for sucrose reinforcement. In experiment 1, the rats were exposed to concurrent pairs of variable-time (VT) schedules specifying equal inter-reinforcement intervals; responses on a single ”changeover lever” alternated between the two VT schedules. In experiment 2, the rats were exposed to concurrent pairs of variable-interval (VI) schedules specifying equal inter-reinforcement intervals; responses on one lever (”VI lever”) earned reinforcers, while responses on the other lever (”changeover lever”) alternated between the two VI schedules. Results: In experiment 1, both groups showed longer ”dwell-times” (intervals between successive changeover responses) when a reinforcer was delivered in the ”dwell” than when no reinforcer was delivered (”win-stay” effect). The lesioned rats showed higher rates of changeover responding and shorter dwell-times (with and without reinforcer delivery) than the sham-lesioned group. In experiment 2, the rate of responding on the VI lever did not differ significantly between the two groups; however, the lesioned rats showed higher rates of changeover responding, shorter dwell-times (with and without reinforcer delivery) and smaller numbers of inter-changeover responses on the VI lever than the sham-lesioned group. In both experiments, the levels of 5-HT and 5-hydroxyindoleacetic acid were reduced in the brains of the lesioned rats, but the levels of noradrenaline and dopamine were not altered. Conclusions: These results provide further evidence for the involvement of the ascending 5-HTergic pathways in behavioural ”switching”, and indicate that this is not restricted to temporal differentiation schedules.