When do placebo and nocebo work? The role of time on placebo analgesia and nocebo hyperalgesia.

Verbal suggestions are strong modulators of one’s expectations and they can be used to induce placebo and nocebo responses. Research so far has investigated the magnitude (i.e. stronger or weaker) and the direction (i.e. increase or decrease of pain) of verbal suggestions, while no attention has been given to the dimension of time. Relying on three main experiments, which investigated the influence of temporal verbal suggestions in modulating the onset of action of placebo analgesia and nocebo hyperalgesia, this thesis seeks to address this shortcoming. In Study 1, pain was induced experimentally on healthy participants via short- lasting, medium-to-low intensity electrical stimuli. After each noxious stimulus participants rated their pain from 0 (no pain) to 10 (unbearable pain). Partic- ipants were assigned to one of three placebo groups, three nocebo groups, a no expectancy (NE) group, or a natural history (NH) group. An inert cream was ad- ministered to all participants, except from those in the NH group, while different verbal suggestions were given according to group allocation. Participants in the placebo groups were told that the cream had analgesic properties setting in after 5 (Placebo Group 5, P5), 15 (Placebo Group 15, P15) and 30 (Placebo Group 30, P30) minutes from cream application. Participants in the nocebo groups were told that the cream had hyperalgesic properties setting in after 5 (Nocebo Group 5, N5), 15 (Nocebo Group 15, N15) and 30 (Nocebo Group 30, N30) minutes from cream application. Participants in the NE group were told that the cream only had hydrating properties and that would not influence pain perception, while those in the NH group did not receive the cream and served to control for pain natural fluctuations over time. Participants repeated the pain test at baseline, after 10, 20 and 35 minutes after the cream application. Mixed-method analysis of variance showed a significant interaction between group and time, indicating that pain ratings varied between time-points and between groups. As expected, post hoc comparisons revealed that placebo and nocebo groups began to show a significant change in pain ratings than the NE group at the expected time point but not earlier. Interestingly, once triggered, the analgesic effect remained stable over time, while the hyperalgesic effect increased over time. In Study 2 and 3, the influence of temporal suggestions on placebo analgesia (Study 2) and nocebo hyperalgesia (Study 3) onset was investigated using a long lasting, high-intensity, tonic pain model, induced with the Cold Pressor Test (CPT). Heart Rate (HR) was measured to assess whether it correlated with placebo analgesia and nocebo hyperalgesia. In Study 2, participants were assigned to one of two placebo groups, or to the No Expectations (NE) group. In Study 3, participants were allocated to one of two nocebo groups, while the control group (NE) was taken from the previous study (Study 2). In this case participants also received an inert cream and those in the placebo groups were told that the cream had analgesic properties that would set in after 5 (placebo 5, P5) and 30 (placebo 30, P30) minutes from its application. Participants in the nocebo groups were told that the cream had hyperalgesic properties setting in after 5 (nocebo 5, P5) and 30 (nocebo 30, N30) minutes from application, while those in the NE group were told that the cream only had hydrating properties. All the participants repeated the CPT at baseline and after 10 and 35 minutes from cream application. Percentage change in exposure time (pain tolerance) from baseline to Test 10 (∆10) and to test 35 (∆35) and changes in HR during CPT were compared between the three groups. In both studies, data were non- parametric and non-parametric statistics were used accordingly. In Study 2, ∆10 was greater in P5 than in NE and P30, indicating analgesia only in the group expecting cream early onset effect. ∆35 was greater in P5 and P30 compared to NE, showing a delayed onset of analgesia in P30 and maintained analgesia in P5. The same results, but in the opposite direction, were reported in Study 3, where hyperalgesia onset followed the temporal verbal suggestions that partici- pants received. HR differences between groups were not significant in Study 2 nor 3. In conclusion, the experiments demonstrated that both placebo analgesia and nocebo hyperalgesia follow the temporal information provided. In addition, it was shown that once triggered, both placebo analgesia and nocebo hyperalgesia endure over time (at least for the duration of the experimental session). These data apply to experimentally induced pain both of a phasic nature with medium- low intensity and of a tonic nature, reaching high intensities. The important role of verbal suggestions in modulating the onset of action of a given (inert-) intervention could not only aid the clinical use of placebo treatment (e.g., in open-label placebo), but also support the efficacy of active drugs. Indeed, further research is needed to extend these results from healthy participants to patients and from placebos to active interventions.