Modulating pain thresholds through classical conditioning

Abstract

Background: Classical conditioning has frequently been shown to be capable of evoking fear of pain and avoidance behavior in the context of chronic pain. However, whether pain itself can be conditioned has rarely been investigated and remains a matter of debate. Therefore, the present study investigated whether pain threshold ratings can be modified by the presence of conditioned non-nociceptive sensory stimuli in healthy participant.

Methods: In 51 healthy volunteers, pain threshold to electrocutaneous stimuli was determined prior to participation in a simultaneous conditioning paradigm. Participants underwent an acquisition phase in which one non-painful vibrotactile stimulus (CS⁺) was repeatedly paired with a painful electrocutaneous stimulus, whereas a second vibrotactile stimulus of the same quality and intensity (CS^-) was paired with a non-painful electrocutaneous stimulus. Stimulation was provided on the lower back with close proximity between the conditioned stimulus and the unconditioned stimulus. In the test phase, electrocutaneous stimuli at the individually-set threshold intensity were simultaneously delivered together with either a CS⁺ or CS^-. Pain intensity ratings were obtained after each trial; expectancy ratings were obtained after each block. The primary outcome was the percentage of test stimuli that were rated as painful.

Results: Test stimuli were more likely to be rated as painful when they were paired with the CS⁺ than when they were paired with the CS^-. This effect was not influenced by contingency awareness, nor by expectancies or mood states.

Discussion: The findings support the notion that the judgement of an event being painful or non-painful can be influenced by classical conditioning and corroborate the possible role of associative learning in the development and maintenance of chronic pain.

Keywords: Allodynia; Associative learning; Classical conditioning; Pain; Pain conditioning; Pavlovian conditioning.

Figure 1. Set-up of vibrotactors and the electrodes on the back (adapted with permission from Madden et al. (2016a)).

Figure 2. The four phases of the experimental procedure.

Dots indicate vibrotactile stimuli, dark dots show which vibrotactor was activated for each trial type. Circles indicate the field on which the two electrodes were placed (adapted with permission from Madden et al. (2016a)).

Figure 3. Scatter boxplot of expectancy ratings by CS type across baseline, acquisition and test phases (mean, standard error), paneled by contingency awareness.

Figure 4. Scatter boxplots of exact mean intensity ratings (error bars = standard error) by CS type during acquisition and test phase.