In search of conditioned pain: an experimental analysis

Abstract

There is an ongoing debate about whether pain can be classically conditioned, but surprisingly, evidence is scarce. Here, we report 3 experiments investigating this idea. In a virtual reality task, healthy participants were approached and touched near or on their hand with a coloured pen (blue or yellow). During acquisition, participants learned that one of the colours of the pen (CS+) was predictive of a painful electrocutaneous stimulus (ECS) whereas the other coloured pen (CS−) was not. During the test phase, more frequent reports of experiencing an US when none was delivered (“false alarm”) for the CS+ vs CS− qualified as evidence of conditioned pain. Notable differences between experiments were that the US was delivered when the pen touched a spot between the thumb and index finger (experiment 1; n = 23), when it virtually touched the hand (experiment 2; n = 28) and when participants were informed that the pen caused pain rather than simply predicting something (experiment 3; n = 21). The conditioning procedure proved successful in all 3 experiments: Self-reported fear, attention, pain, fear, and US expectancy were higher (P < 0.0005) for the CS+ than the CS−. There was no evidence for conditioned pain in experiment 1, but there was some evidence in experiments 2 and 3. Our findings indicate that conditioned pain may exist, albeit most likely in rare cases or under specific situations. More research is needed to understand the specific conditions under which conditioned pain exists and the underlying processes (eg, response bias).