Manipulating placebo analgesia and nocebo hyperalgesia by changing brain excitability

Abstract

Harnessing placebo and nocebo effects has significant implications for research and medical practice. Placebo analgesia and nocebo hyperalgesia, the most well-studied placebo and nocebo effects, are thought to initiate from the dorsal lateral prefrontal cortex (DLPFC) and then trigger the brain's descending pain modulatory system and other pain regulation pathways. Combining repeated transcranial direct current stimulation (tDCS), an expectancy manipulation model, and functional MRI, we investigated the modulatory effects of anodal and cathodal tDCS at the right DLPFC on placebo analgesia and nocebo hyperalgesia using a randomized, double-blind and sham-controlled design. We found that compared with sham tDCS, active tDCS could 1) boost placebo and blunt nocebo effects and 2) modulate brain activity and connectivity associated with placebo analgesia and nocebo hyperalgesia. These results provide a basis for mechanistic manipulation of placebo and nocebo effects and may lead to improved clinical outcomes in medical practice.

Keywords: dorsolateral prefrontal cortex; expectancy manipulation; mechanistic manipulation; placebo and nocebo effects; transcranial direct current stimulation.

Fig. 1.

Study design. (A) The procedures in the present study. (B) Inert lidocaine, capsaicin, and neutral creams were applied at different spots. Different intensities of heat pain were applied at corresponding spots to manipulate subjects’ expectancy on the analgesic and hyperalgesic effects of the creams. (C) Placebo and nocebo tests. Inert lidocaine, capsaicin, and neutral creams were applied at different spots. Different intensities of heat pain were applied at the three spots in the left column, while the same moderate intensity painful stimuli were applied to the remaining six spots. (D) Timings for a typical trial. (E) tDCS setup. Subjects received 20 min tDCS in Sessions 3, 4, and 5. The anodal electrode was placed over F4 and the cathodal electrode over FP1 for rDLPFC excitability enhancement. The anodal electrode was placed over FP1 and the cathodal electrode over F4 for rDLPFC excitability inhibition. For sham tDCS treatment, stimulation was applied only at ramp-up/ramp-down periods at the beginning and end of sham stimulation to mimic the somatosensory effect of real tDCS for 15 s.

Fig. 2.

Pain and conditioned pain-related brain responses. (A) Heat painful stimuli elicited brain activations in the bilateral insula (INS), caudate (CAU), anterior cingulate cortex (ACC), presupplementary motor area (pre-SMA), and primary somatosensory cortex (S1) and decreased brain activations in the precuneus (PCUN) and visual cortex (VS). (B) During pain experience, the expectancy of pain relief with lidocaine induced increased brain activations in the rDLPFC, CAU, pre-SMA, and ACC and decreased brain activations in the SMA and S1. (C) During pain experience, the expectancy of pain exacerbation with capsaicin induced increased brain activations in the bilateral INS, pregenual and subgenual ACC (pgACC and sgACC), and CAU and decreased brain activations in the PCUN. (D) Pain ratings corresponding to identical moderate painful stimuli applied on different creams. The sham group had both significant placebo analgesia and nocebo hyperalgesia, while both anodal and cathodal groups had significant placebo analgesia and nonsignificant nocebo hyperalgesia. (E) Compared with sham tDCS, cathodal tDCS significantly boosted placebo analgesia, and anodal tDCS significantly inhibited nocebo hyperalgesia. Asterisks indicate two-tail P_tukey < 0.05 for post hoc comparisons of adjusted placebo and nocebo responses by covariates in the ANCOVAs. (F) Subjects who received cathodal tDCS showed significantly higher brain activations in the vmPFC for placebo contrast (lidocaine versus neutral). Cathodal tDCS significantly increased activations in the vmPFC when experiencing painful stimuli on the lidocaine cream, and the fMRI brain response difference (lidocaine neutral) was significantly correlated with placebo analgesia in the cathodal group. (G) The task-based connectivity (measured by PPI) between the rDLPFC and vmPFC was increased when experiencing pain on the lidocaine cream in the cathodal group. (H) Subjects who received anodal tDCS showed decreased brain activations in the left INS for nocebo contrast (“capsaicin versus neutral”). Anodal tDCS significantly inhibited activations in the INS when experiencing painful stimuli on the capsaicin cream and disrupted the significant association between the fMRI brain response difference (lidocaine neutral) and nocebo hyperalgesia observed in the sham group. (I) The task-based connectivity between the rDLPFC and insula was increased when experiencing pain on the capsaicin cream in both sham and anodal groups, but such connectivity in the anodal group was significantly lower than the sham group. Note, results in A–C were corrected for multiple comparisons at the whole-brain level, and results in E were corrected within the mask consisting of typical regions in the DPMS (i.e., the ACC, mPFC, insula, and SMA). Statistical tests between bars were threshold as P < 0.05 and were corrected from multiple comparisons using FDR. Error bars represent SE of mean.

Fig. 3.

Modulatory effects of tDCS on intrinsic brain connectivity. (A) Anodal tDCS significantly decreased functional connectivity between the rDLPFC and insula in Sessions 3 and 5. (B) In the anodal group, the rDLPFC-insula connectivity at the post-tDCS fMRI scan of Session 5 was significantly associated with subsequent brain responses in the insula and pain ratings when experiencing pain on the capsaicin cream but not on the neutral cream. (C) Cathodal tDCS significantly increased functional connectivity between the rDLPFC and vmPFC in Sessions 3 and 5. (D) In the cathodal group, the rDLPFC-vmPFC connectivity at the post-tDCS fMRI scan of Session 5 was significantly associated with subsequent brain responses in the vmPFC and pain ratings when experiencing pain on the lidocaine cream but not on the neutral cream. vmPFC: ventromedial prefrontal cortex; INS: insula; Pre-/Post-tDCS: the resting-state fMRI scan before/after the application of tDCS; tDCS early/late: the simultaneously collected fMRI during the first/last 6 min of tDCS application; rsfMRI: resting-state fMRI. Asterisks indicate two-tail P_FDR < 0.05. Error bars represent SEM.