The notion of response invariance in trade-off studies of self-stimulation

Trade-off profiles, displaying the co-variations of 2 electrical parameters required to maintain a constant magnitude of brain stimulation reward (BSR), have been used extensively in order to characterize the self-stimulation (SS) neurons. It has often been assumed that constancy in the magnitude of BSR can be achieved more accurately by holding SS at a constant proportion of the maximum rate, rather than at a constant rate. The validity of this assumption was tested in 2 experiments using rats. In Exp. 1, we first computed the function that relates SS barpressing rate to pulse frequency (RF function) for two different pulse intensities, separately. The peak SS rate was found to be lower in the low current than in the high current RF function. The rats were then placed in a 2-lever box and were allowed to select either a fixed frequency of the high current pulses or a variable frequency of the low current pulses. In Exp. 2, the RF function was first computed for 2 different lever weights, separately. The peak SS rate was found to be lower in the heavy-lever RF function than in the light-weight lever RF function. The rats were then allowed to select either a fixed pulse frequency delivered by the heavy lever or a variable pulse frequency delivered by the light-weight lever. Isopreference was noted in both experiments, for pulse frequencies which, in the single-lever box, elicited the same proportion of the maximum SS rate, not the same SS rate. The data thus validate the idea that a constant magnitude of BSR is translated into a constant proportion of the maximum SS rate, not a constant SS rate.