Temporal summation of sensory intensity was investigated in normal subjects using novel methods of thermal stimulation. A Peltier thermode was heated and then applied in a series of brief (700 ms) contacts to different sites on the glabrous skin of either hand. Repetitive contacts on the thenar or hypothenar eminence, at interstimulus intervals (ISIs) of 3 s, progressively increased the perceived intensity of a thermal sensation that followed each contact at an onset latency > 2 s. Temporal summation of these delayed (late) sensations was proportional to thermode temperature over a range of 45-53 degrees C, progressing from a nonpainful level (warmth) to painful sensations that could be rated as very strong after 10 contacts. Short-latency pain sensations rarely were evoked by such stimuli and never attained levels substantially above pain threshold for the sequences and temperatures presented. Temporal summation produced by brief contacts was greater in rate and amount than increases in sensory intensity resulting from repetitive ramping to the same temperature by a thermode in constant contact with the skin. Variation of the interval between contacts revealed a dependence of sensory intensity on interstimulus interval that is similar to physiological demonstrations of windup, where increasing frequencies of spike train activity are evoked from spinal neurons by repetitive activation of unmyelinated nociceptors. However, substantial summation at repetition rates of > or = 0.33 Hz was observed for temperatures that produced only late sensations of warmth when presented at frequencies < 0.16 Hz. Measurements of subepidermal skin temperature from anesthetized monkeys revealed different time courses for storage and dissipation of heat by the skin than for temporal summation and decay of sensory intensity for the human subjects. For example, negligible heat loss occurred during a 6-s interval between two trials of 10 contacts at 0.33 Hz, but ratings of sensory magnitude decreased from very strong levels of pain to sensations of warmth during the same interval. Evidence that temporal summation of sensory intensity during series of brief contacts relies on central integration, rather than a sensitization of peripheral receptors, was obtained using two approaches. In the first, a moderate degree of temporal summation was observed during alternating stimulation of adjacent but nonoverlapping skin sites at 0.33 Hz. Second, temporal summation was significantly attenuated by prior administration of dextromethorphan, a N-methyl-D-aspartate receptor antagonist.
