Changes in Neural Processing and Rating of Negative Facial Expressions After Administration of Open-Label Placebo
A total of 109 right-handed women were recruited from a non-clinical sample at a university and via social media. The participants were university students (77%) or white collar workers (13%). Inclusion criteria for participation included female gender and age ≥ 18 years. Exclusion criteria were self-reported diagnoses of mental illnesses, neurological disorders, and psychotropic medications. Data from six participants were removed from the analysis due to excessive EEG artifacts (i.e., less than 70% artifact-free trials). The final sample included 103 women (Mage= 22.48 years, SD = 3.35; all Caucasian, right-handed). Participants were randomly assigned to the OLP group (Mage= 22.46 years, SD = 3.01) or the CG (Mage= 22.49 years, SD = 3.70). The sample was restricted to females due to gender effects in the context of emotional processing and placebo responsiveness32.
The experiment complied with all relevant ethical guidelines and regulations involving human participants and was approved by the ethics committee of the University of Graz (Austria; GZ. 39/98/63 ex 2020/21). All participants gave their informed consent before participating. Eligible participants had to come to the laboratory.
Experimental design and procedure
The study was pre-registered on the Open Science Framework (https://osf.io/cxfgb/?view_only=5f4c6d564) on April 30, 2021, and conducted between May 25 and August 30, 2021 at the University of Graz (Austria) . Additionally, the study has been registered as a clinical trial in the German Clinical Trials Register (DRKS00028129; February 18, 2022).
Participants were invited to an affective processing EEG study (no information about placebos was provided in the invitation). In the preparation room, a researcher used a random number table to assign participants to the OLP or control group. Both groups viewed a presentation (15 PowerPoint slides with figures and text; no audio; fixed duration per slide: 30 s; see supplementary material). Those in the OLP group received information about the neurobiological effects of placebos with an emphasis on affective processing (results from EEG and fMRI studies). The control group watched a presentation on affective neuroscience (results of EEG and fMRI studies). The two presentations were comparable by number of slides, numbers, number of words). Next, participants rated the presentation (0 = not interesting; 10 = very interesting) and their current affective state (0 = not positive; 10 = very positive).
Afterwards, a short video (duration: two minutes; female presenter) introduced the nasal spray. It was clearly stated that the nasal spray did not contain any ingredients other than saline and water. The OLP group was told that the nasal spray could help reduce emotional reactions when viewing images with angry facial expressions, while the CG was told that the spray would improve EEG recording. The information was summarized by the experimenter, who helped deliver the nasal spray once into each nostril. Participants in the PLO group rated their expectations regarding the effectiveness of PLOs (“What do you think? How effective will the PLO be?” 0 = not effective; 10 very effective).
Then the participants were brought to the EEG laboratory. Two experimenters, who were unaware of the group assignment, conducted the EEG experiment. At the end of the EEG experiment, participants were asked to rate the effectiveness of the nasal spray (“How effective do you think the nasal spray was?” 0 = not effective; 10 very effective).
Image viewing task
Participants viewed a total of 60 images of Karolinska’s directed faces33. Thirty images depicted angry facial expressions; 30 images of neutral facial expressions; 50% of the faces were male; 50% women). The images were presented in random order. In each trial, participants first viewed a blank screen (500 ms), a staring cross (500 ms), and then an image with a facial expression (6000 ms). Eight images (four angry/four neutral expressions; 50% male/female) were rated for valence, arousal, and perceived intensity of anger (0-100; 0=I do not feel pleasant, calm, I do not perceive anger; 100: I feel very pleasant, excited, I perceive intense anger). The images for the evaluations had been randomly selected before the experiment; notes were to be provided at random times during the experiment.
Electrophysiological recording and data analyzes
Continuous EEG activity was recorded using the actiCHamp system (actiCHamp, Brain Products GmbH, Gilching, Germany) with 63 active actiCAP pressure electrodes (according to the 10–10 system) and the BrainVision Recorder (version 1.21). The reference electrode was placed on the FCz position, the ground electrode on the FPz position. An electrolytic gel was applied to each electrode to keep electrode impedances below 10 kΩ. The EEG was recorded with a sampling rate of 2500 Hz and a bandwidth of 0.016 at 1000 Hz. For the analysis of the raw data, the BrainVision analyzer (version 2.2.1) was used. The sampling rate was changed to 250 Hz. The data was referenced to the linked mastoid electrodes (i.e. TP9, TP10). Eye movement artifacts were corrected through the implemented ICA eye correction software – only components corresponding to horizontal and vertical eye movements were selected based on their shape, timing and topography matching. Other episodes of artifacts were excluded after visual inspection. Six participants were excluded from the analysis due to a large number of artifacts ( 0.05).
Data were segmented into 6200 ms intervals (200 ms pre-stimulus onset, 6000 ms post-stimulus onset) and corrected to the 200 ms pre-stimulus baseline. An off-line high pass (0.01 Hz) and low pass filter (30 Hz cutoff frequency, 24 dB/octave attenuation) has been applied. Data were averaged for all groups and conditions separately. Based on previous literature and visual inspection of general averaged waveforms, we extracted ERPs for time windows 400-1000 ms (early LPP) and 1000-6000 ms (LPP late) after the start of the picture. Mean amplitudes were aggregated over a centroparietal group (C3, C1, Cz, C2, C4, CP3, CP1, CPz, CP2, CP4, P3, P1, Pz, P2, P4) and a frontal group (AF3, AFz, AF4, F3, F1, Fz, F2, F4, FC3, FC1, FC2, FC4).
After viewing the images, an unannounced memory task was conducted with 16 images (8 angry expressions, 8 neutral expressions; 50% male, 50% female). Eight of the images had been shown in the experiment; the other images were new distractors. The images had been randomly selected before the experiment (and were different from the images assessed (affective ratings) during the experiment). Participants were asked to decide if they had seen the image in the experiment (yes/no).
All statistical analyzes were performed with SPSS (version 28). The investigator who analyzed the data collected in the study was unaware of the treatment applied to the groups. Mixed factor analyzes of variance (ANOVA) were calculated with group (OLP, CG) as between-subjects factor and face (anger, neutral) as within-subjects factor for image ratings, memory performance and EEG data. Cohen’s d is reported as a measure of effect size. Significant interaction effects (group x face) were followed by adjusted t-tests for multiple comparisons (Bonferroni-Holm).
T-tests compared the groups regarding their scores for the PowerPoint presentations (rated interest), emotional state before the experience, and perceived effectiveness of the nasal spray. All tests were two-sided and used a significance level of p