No add-on therapeutic benefit of at-home anodal tDCS of the primary motor cortex to mindfulness meditation in patients with fibromyalgia

Objective This study investigated the efficacy of combining at-home anodal transcranial direct current stimulation (tDCS) of the left primary motor cortex (M1) with mindfulness meditation (MM) in fibromyalgia patients trained in mindfulness. Methods Thirty-seven patients were allocated to receive ten daily sessions of MM paired with either anodal or sham tDCS over the primary motor cortex. Primary outcomes were pain intensity and quality of life. Secondary outcomes were psychological impairment, sleep quality, mood, affective pain, mindfulness level, and transcranial magnetic stimulation (TMS) measures of cortical excitability. Outcomes were analyzed pre-and post-treatment, with a one-month follow-up.


Introduction
Fibromyalgia (FM) is a chronic pain condition characterized by widespread musculoskeletal nociplastic pain, accompanied by mood, psychological, cognitive, and sleep impairments (Kosek et al., 2021;Wolfe et al., 1990Wolfe et al., , 2010)).The debilitating impact of FM on patients' physical, social, and psychological well-being results in a reduced quality of life (QOL) (Birtane et al., 2007;Lee et al., 2017;Schaefer et al., 2011).Repeated anodal transcranial direct current stimulation (tDCS) (Nitsche & Paulus, 2000;Paulus, 2011) over the left primary motor cortex (M1) for two weeks or longer has proved effective in pain relief and showed improvements in associated symptoms and QOL in FM with Level A recommendations for pain management (Baptista et al., 2019;Fregni et al., 2021), but not in all studies (Cheng et al., 2023;Conde-Antón et al., 2023;Fregni et al., 2006;Khedr et al., 2017;Molero-Chamizo et al., 2023;Moshfeghinia et al., 2023;Teixeira, 2020).These heterogeneous findings ranging from no discernible improvement to long-lasting pain reduction up to three months post-stimulation could be due to variations in stimulation duration and intensity and methodological loopholes regarding blinding, sham, and sample size.Previous studies showed that coupling anodal tDCS with other non-pharmacological interventions can boost its therapeutic benefits (Arroyo-Fernández et al., 2022;Mendonca et al., 2016;Pinto et al., 2018).
Recently, combining tDCS with mindfulness interventions in chronic pain has gained attention (Divarco et al., 2023;Rebello-Sanchez et al., 2022).Mindfulness involves the cultivation of a non-judgmental, non-attached, and intentional focus on the present moment (Kabat-Zinn, 1982), with mindfulness meditation (MM) training one's attention to attain a mindful state (Matko & Sedlmeier, 2019).Regular mindfulness practitioners with FM reported enhanced sleep quality, mental health, and resilience (Amutio et al., 2015;Cejudo et al., 2019;Van Gordon et al., 2017).Pairing MM with anodal M1-tDCS in knee osteoarthritis showed greater pain relief and reduction of symptoms than the sham group (Ahn et al., 2019;Pollonini et al., 2020).In our pilot study, FM patients reported clinically meaningful improvement in QOL following combined MM and anodal M1-tDCS compared to control groups (Ramasawmy et al., 2022).The local, network, and global brain states immediately before and during stimulation play a pivotal role in shaping the online and the enduring offline neuronal response to tDCS (Bergmann, 2018;Stagg et al., 2018;Vergallito et al., 2022).The first mechanistic exploration of MM paired with repeated M1-anodal tDCS showed reduced pain in patients with knee osteoarthritis and demonstrated increased activation of the somatosensory and superior motor cortices, measured using functional near-infrared spectroscopy, without any significant activations in the sham group (Pollonini et al., 2020).The combination of MM and tDCS in chronic pain is believed to further enhance the adaptive changes in the brain caused by the individual interventions, producing augmented therapeutic benefits (Ahn et al., 2019).Our rationale for the combined intervention lies in the potential for MM to synergistically bolster the analgesic effects of tDCS in FM through the modulation of brain state during stimulation.
Compared to the aforementioned studies (Ahn et al., 2019;Pollonini et al., 2020;Ramasawmy et al., 2022), we trained the participants in mindfulness by including a compulsory 4-week mindfulness training tailored for FM.In this study, we paired 10 sessions of M1-anodal tDCS with MM in FM patients trained in mindfulness practice.Our main objective was to determine if MM paired with anodal tDCS was more effective at reducing pain and improving QOL and core symptoms in FM than sham.Despite differences in the cortical excitability between FM patients and healthy controls (Cardinal et al., 2019;Mhalla et al., 2010;Pacheco-Barrios et al., 2022), the impact of repeated anodal tDCS on transcranial magnetic stimulation (TMS) measures of corticospinal excitability in chronic pain remains unclear, primarily due to variations in the number of tDCS sessions and inconsistent findings (Antal et al., 2010;Nitsche et al., 2002;Portilla et al., 2013).The sole meditation study using TMS inferred that meditation may be associated with intracortical inhibition in healthy individuals (Guglietti et al., 2013).Therefore, we aimed to explore the mechanistic corticospinal and intracortical effects of combining M1-anodal tDCS and MM using single-pulse and paired-pulse TMS (Burke et al., 2019;Hallett, 2007).

Material and methods
This randomized, triple-blinded, and placebo-controlled Phase III clinical trial at the University Medical Center Göttingen, Germany from May 2022 to October 2023.The protocol was approved by the local ethics committee, according to the Declaration of Helsinki.The study was registered at the German Clinical Trials Register under the identification number DRKS00029024 (https://drks.de/search/en/trial/DRKS00029024).All participants signed an informed consent form.

Participants
We recruited participants through advertisements in local and national pain and rheumatology clinics and practices, self-help groups for FM, newspapers, and social media.Efforts were titrated to ensure consistent participant inclusion and minimal waiting time.An experienced pain physician (FP) confirmed participant inclusion.
Patients (30 age 75 years) who met the ACR for FM with a widespread pain index !7 and a symptom severity ! 5 (Wolfe et al., 2010) and had persistent pain for at least six months were eligible for the study.Participants required access to the Internet and basic skills in how to operate online meetings.Participants were excluded if there was evidence of 1) acute or unstable medical, neurological, or psychiatric disorder, 2) severe or uncontrolled comorbid rheumatic disease, 3) untreated and ongoing treatment for cancer, 4) chronic migraine, 5) history of substance abuse or 6) active or contested disability application.Pregnant or breastfeeding women were excluded.Patients with contraindications to TMS and tDCS-a history of unexplained or repeated loss of consciousness, brain surgery, metallic implants in the head, neck, or chest, and contraindicative pain medication-were also excluded.Adjunct medication or non-pharmacological therapies for FM were allowed depending on its duration, dose, and nature, if the dose was stable for at least four weeks before inclusion and stayed stable during the study.Patients were not allowed more than three drugs for FM symptom relief.An exhaustive list of medication and therapies permitted and not allowed for participation is summarized in Supplementary Table S1.Participants were permitted to continue psychotherapy (maximum once per week) and/or physiotherapy ( 3 times per week).
All included participants were trained in mindfulness meditation by completing a four-week mindfulness intervention tailored for FM patients, which was developed by OLGA, an adept mindfulness teacher with more than 10 years of experience, FP and PR by adapting the standard mindfulness-based stress reduction program (Kabat-Zinn, 1982, 2013), while its structure was similar to Moreira et al. (2022).The mindfulness intervention aimed to train participants in mindfulness practices that would provide participants with the appropriate techniques and adequate experience necessary to follow a 20-min guided breathing meditation, thus preparing them for an adequate application of the combination therapy.

Study design
The study had three phases: a one-week baseline period, a twoweek therapy combining at-home tDCS with MM, and a threemonth of follow-up period.Included participants were randomized to receive either anodal tDCS over the left M1 at 2 mA paired with real MM or sham tDCS combined with real MM.A block randomization with stratification for sex and age was conducted by a trained colleague for the allocation of participants to the intervention groups.The colleague received the code for each stimulator and the information regarding the nature of the protocol, which was pre-programmed by the Sooma Medical company.Since experimenters were only aware of the stimulator to be assigned to each participant, both patients and interventionists were blinded to the tDCS protocol.The data collection and data analysis were blinded, making the study triple-blinded.Measurement visits took place in the clinic at baseline (T1), within a week following the last stimulation (T2), and 4 weeks following the last combined intervention (T3), as summarized in Table 1.Self-reported questionnaires were administered for two extra follow-up time points at 8 and 12 weeks post-stimulation (T4 and T5), where patients completed the questionnaires at home and mailed them to the study team.The study took place in a hybrid setup, with some meetings taking place in the hospital and others supervised online via Zoom (Zoom Video Communications Inc., California, USA).

Home-based tDCS intervention
Home-based tDCS was applied for 20 min per session daily for 2 weeks (Monday to Friday) via a battery-driven Sooma tDCS TM stimulator (Sooma Medical, Helsinki, Finland) with a proprietary Sooma headgear and two 25 cm 2 round electrodes made from silicone mixed with Ag/Al.Electrodes maintained contact with the scalp via 0.9% isotonic saline-soaked hydrogel pads.The anode was placed over the left M1 (C3 according to the 10-20 EEG system) and the cathode over the right supraorbital cortex (FP2).For the anodal tDCS, a constant current of 2 mA (19 min 23 s) with 17 s ramp up and 20 s ramp down was applied.Sham stimulation comprised 17 s ramp-up from 0 mA to 2 mA, 17 s ramp-down to 0.3 mA, 0.3 mA constant current for 19 min 23 s, and 3 s rampdown at the end of the session, which was previously validated (Hyvärinen et al., 2016).Each home-based stimulation session conducted in small groups with a maximum of six participants was supervised via Zoom by trained interventionists.Before the selfapplication of home-based tDCS, every participant received two training sessions in the clinic for setting up and operating the device-the first one at T1 and the second on the first day of the combined intervention, which took place in the clinic (Table 1).Each participant received the daily stimulation at the same time of the day, whenever possible to avoid any circadian confounding factors (Salehinejad et al., 2019).
At T3, participants were asked to guess which tDCS protocol they received using the end-of-study guess approach, commonly used in tDCS studies was implemented (Ambrus et al., 2012;Stankovic ´et al., 2022;Turi et al., 2019).The participants were informed about the group they were allocated to after all the follow-up assessments were completed.

Mindfulness meditation intervention
During each tDCS session, participants practiced MM simultaneously with tDCS.The meditation was delivered via an audioguided recording.The meditation instructions were created by OLGA and recorded in German by SES, with five years of experience in teaching meditation.The MM exercise focused on developing attention to the breath as the object of awareness.This attention was then expanded to involve a non-attached, non-judgmental, and open observation of any emotional, cognitive, or sensory experiences (Kabat-Zinn, 1982).Participants were instructed to meditate in a comfortable sitting position with their eyes closed and were encouraged to focus on breathing from the diaphragm or lower abdomen, as opposed to the chest, to aid initial relaxation and encourage a natural breathing rhythm.The same recording was used daily for the 10 days of combined intervention.
From T2 to T3, all participants followed a regular mindfulness routine comprising 20 min guided using a provided audio recording and a mindful activity such as mindful walking, mindful eating, etc. three times per week.We reduced but did not stop their meditation routine as it would be unethical to ask the participants to stop a treatment that has shown some effectiveness in FM.Moreover, adding tDCS to the MM practice would only be sensible if there would be improvements beyond meditation.

Outcome measures
The primary outcomes were self-reported pain intensity measured using the 11-point numerical rating scale (NRS, 0: no pain; 10: worst imaginable pain) (Bolton & Wilkinson, 1998;Euasobhon et al., 2022) and QOL using the German version of the Fibromyalgia Impact Questionnaire (FIQ, 0: no impact; 100: maximal impact) (Bennett, 2005); Offenbaecher et al., 2000).Patients were asked to rate their pain twice a day (immediately after waking up and just before going to sleep) in a pain diary throughout the study from T1 to T3.The pain scores were averaged to give one NRS mean value for each study week.The FIQ was administered at T1 and each follow-up time point.All secondary clinical outcomes were measured, as part of a battery of questionnaires, at T1, then at each follow-up time point.
The German version of the Five-Facet Mindfulness Questionnaire (FFMQ) (Baer et al., 2006;Michalak et al., 2016) was used to quantify the mindfulness level of the participants.The mean of the five facets was computed to obtain an overall mindfulness score.
The German version of the Pittsburgh Sleep Quality Index (PSQI) was used to assess sleep quality (Buysse et al., 1989;Tietze et al., 2014), with lower scores indicating better sleep quality.Healthy sleep quality is represented by a score from 0 to 5.
The safety of the paired tDCS and MM intervention was assessed by monitoring the occurrence of side and adverse effects on a 4-point Likert scale (0: none, 1: mild, 2: moderate, and 3: strong) after each treatment session.
The feasibility of the home use of tDCS was self-reported by patients by rating two statements on a 10-point Likert scale (0: strongly disagree, 10: strongly agree) -1) In general, it was easy to use the device and 2) It was easy to prepare the device and its accessories.This method was adapted from Ahn et al. (2019).

TMS mechanistic outcomes
Monophasic TMS was delivered using a Magstim Bistim 2 module (Magstim Co., Whiteland, Dyfed, UK) applied to the 'hotspot' of the right first dorsal interosseous muscle (FDI) through a figure-ofeight 70 mm magnetic coil (2.2 T; average inductance,16.35mH).The coil was held tangentially to the scalp and positioned in the posterior-anterior direction with 45°from the midline.Motorevoked potentials (MEPs) were recorded via surface electrodes placed using the classical belly-tendon montage by measuring the electromyographic activity of the relaxed FDI.The electromyography signals were band-pass filtered (2 HzÀ3 kHz) amplified (D360, Digitimer, Ltd), digitized (sampling rate 5 kHz) using an CED 3001 A/D converter (MICRA 1401 mk II; Cambridge Electronic Design Limited, Cambridge, UK) and Signal 2.13 software for electromyography recording, and monitored on a computer screen using visual feedback (Rossi et al., 2009).Stimulation intensities were expressed as a percentage of the maximum stimulator output (MSO).The TMS system was connected to a stereotactic neuronavigation system (Brainsight TMS Navigation, Rogue Resolutions Ltd) coupled with a Polaris Vicra infrared camera (NDI, Waterloo, Canada) (Chakalov et al., 2022;Zmeykina et al., 2020).We assessed the individual resting motor threshold (RMT), defined as the stimulator intensity to constantly evoke MEP peak-to-peak amplitudes !0.05 mV (positive MEPs) in the resting FDI, estimated using the maximum likelihood parameter estimation by sequential testing (Awiszus, 2003;Awiszus et al., 1999;Groppa et al., 2012) based on adaptive threshold-hunting in the TMS Motor Threshold Assessment Tool software.Input-output (I-O) curves (Alavi et al., 2021) were measured using six increasing stimulator intensities (90%, 100%, 110%, 120%, 130%, and 140% of RMT), each with 10 single pulses at a frequency of 0.25 Hz.We fitted the mean values for each stimulation intensity with the Boltzmann sigmoid equation (Devanne et al., 1997).In cases of inadequate curve fitting, the Gompertz sigmoidal function was implemented.The parameters included the curve slope and the stimulation intensity required to achieve half-maximal MEP size (S 50 ) that were computed from the curve fit (Kukke et al., 2014).The SI 1mV , defined as the TMS intensity that elicits an MEP with a mean peak-to-peak amplitude of $1 mV in the FDI, was also assessed.We then measured short interval intracortical inhibition (SICI), intracortical facilitation (ICF), and long interval intracortical inhibition (LICI) (Burke et al., 2019;Hallett, 2007).The SICI/ICF protocol (Kujirai et al., 1993) comprised a suprathreshold test stimulus (TS) at 120% of RMT preceded by a subthreshold conditioning stimulus (CS) delivered at 80% of RMT to prevent any possible saturation effect.We tested an inter-stimulus interval (ISI) of 3 ms for SICI and 10 ms for ICF.The LICI protocol consisted of a CS and a TS set 120% RMT, which was delivered with ISIs of 100 ms and 150 ms (Redondo-Camós et al., 2022;Sanger et al., 2001;Valls-Solé et al., 1992).All the ISIs were randomized with the TS.Sixteen trials were delivered per ISIs and also for the TS.The mean peak-to-peak amplitude of the conditioned MEP for SICI, ICF, and LICI was expressed as a percentage of the mean peak-to-peak size of the test pulse.LICI was taken as the mean percentage inhibition at 100 ms and 150 ms ISIs.The raw MEP data was preprocessed using the tmsExplorer software (https://github.com/omeganow/tms-explorer),developed by OM, TTM, and PR to remove stimulation artifacts and noisy baselines.
The TMS measures were performed at T1, T2 and T3.To limit the influence of circadian rhythms on cortical excitability, the patients' appointments were kept at the same time of day (Lang et al., 2011).Patients were asked to maintain the same amount of caffeine intake prior to each measure visit (Kalmar & Cafarelli, 2004;Zulkifly et al., 2021).

Statistical analysis
Based on our preliminary findings (Ramasawmy et al., 2022), statistical power analysis calculation with an effect size of 0.243 for a statistically significant QOL improvement, statistical power of 0.80, a error rate of 0.05, and correlation among repeated measures of 0.50 using the G*Power 3.1 software required a minimum of 15 participants per group.We added 20% to the sample size required to account for potential dropouts.
Baseline demographics were compared between groups using a two-tailed Student's unpaired t-test for age, body mass index, widespread pain index, symptom severity, and outcomes and a v 2 analysis for biological sex, handedness, medication intake, and non-pharmaceutic therapies.
We implemented an intention-to-treat approach and included all randomized participants (n = 37) who received at least 5 combined intervention sessions (Gupta, 2011;McCoy, 2017).We only included data from T1 to T3 in the statistical analyses as NRS pain and TMS measures were taken until then.Data at follow-up time points T4 and T5 were only reported descriptively.For the NRS pain scores, we used a two-way mixed model ANOVA with WEEK (T1, stimulation week 1 (Stim1), stimulation week 2 (Stim2), and four follow-up weeks (PostStim1, PostStim2, PostStim3, PostS-tim4) as the repeated factor and GROUP ((active + MM, sham + MM) as the independent factor.The data for the FIQ and all secondary outcomes using two-way mixed model ANOVAs with TIME (T1, T2, T3) and GROUP (active + MM, sham + MM) as the repeated and independent factors respectively Bonferronicorrected pairwise comparisons were conducted for statistically significant effects.The effect size was computed as eta squared g 2 and was interpreted based on Cohen's benchmark categorization (small, g 2 = 0.01; medium, g 2 = 0.06 and large, g 2 = 0.14) (Cohen, 1988).Outcome variables at other time points were analyzed and summarized descriptively.The normality of raw data or residuals was verified visually with normality plots and Shapiro-Wilk's test.Non-parametric tests were performed in case of violations of parametric assumptions or inability to transform data.For mixed model ANOVAs, homoscedasticity was tested with Levene's test.Sphericity was tested via Mauchly's test for repeated factors and Greenhouse-Geisser corrections were for sphericity deviations.We compared the number of patients reporting side and adverse effects between the two groups using the Fisher-Freeman-Halton exact test.The total rating of each side or adverse effect over the 10 days of combined intervention was calculated and compared between the groups using a Mann-Whitney U test.
To compare the rating of each feasibility questionnaire between the two groups, we performed a two-tailed unpaired Student's ttest.We assessed blinding success using the Chi-squared (v 2 ) test to test, whether the guessing of the type of stimulation received was different from the chance level (50%).
We followed the iterative Markov chain Monte Carlo imputation method with linear regression for partially missing data.For each outcome, a maximum of 10% imputation was maintained for unbiased analysis (Bennett, 2001).In the case of completely missing data concerning the TMS results, the participant was excluded from the analysis of the TMS outcomes.The statistics plan was approved by the Department of Medical Statistics at the University Medical Center Göttingen.All statistical analyses were conducted using IBM SPSS 28.0 and statistical significance was set at P < 0.05.

Sample description and baseline demographics
Out of 99 interested individuals screened and assessed for eligibility, 45 (45.6%) met the inclusion criteria while the rest either fell under the exclusion criteria or did not wish to participate due to the demanding nature of the study in terms of time and location (Fig. 1).Initially, 38 FM patients who met the inclusion criteria were randomly assigned to the two intervention groups (Fig. 1).

Allocation
Only one patient dropped out after the first combined intervention session because of his health condition.Finally, 37 patients completed the combined intervention.The active + MM (n = 18) and sham + MM (n = 19) groups were comparable in baseline demographics (Table 2), medication intake (Table 3), and all outcome measures (Supplementary Tables S2 and S3).Out of the 37 patients who completed the combined intervention, 28 (75.6%)participated in the TMS experiments, with an equal participant number in each group.Nine patients discontinued the TMS measures after baseline and were not included in the analysis.The reasons for TMS dropout or exclusion were as follows: dizziness after stimulation (n = 1), strong headache (n = 3), overall discomfort (n = 3), did not want to have the TMS measure after demonstration (n = 1), and no MEP was observed despite hotspot being determined properly (n = 1).We could not find the SI 1mV for one patient at baseline in the active + MM group, thus excluding them from the analysis (Supplementary Table S3).When analyzing the paired pulse data, some participants showed inhibition instead of facilitation for ICF and facilitation instead of inhibition for SICI.These participants were excluded from analysis, hence conducive to the inconsistency in the number of participants who completed SICI and ICF (Supplementary Table S3).There were no serious adverse effects because of the TMS stimulation.

Primary outcomes
Both the active + MM and sham + MM groups reported significant reductions in NRS pain intensity (F(3.28,115) = 3.68, P = 0.012, gp 2 = 0.095, Greenhouse-Geisser correction) over time with a medium effect size, without any significant interaction between group and time effect.Fig. 2 illustrates the change in NRS pain scores over time, with no statistically significant pairwise comparisons (P > 0.05).
We found a medium main effect of time for FIQ scores (F (2,70) = 5.39, P = 0.007, gp 2 = 0.13), demonstrating a significant improvement in QOL over the study course.Further investigation of this main effect of time showed a higher QOL at T2 (mean difference (MD) = 7.76, P = 0.003) than at T1, as shown by lower FIQ scores, as illustrated in Fig. 3.However, no further improvement in QOL from T2 to T3 was observed (MD = 2.74, P = 0.872).We did not find any statistically significant interaction effect.

Secondary clinical outcomes
We observed a significant main effect of time for the PDL affective pain, PSQI sleep quality, DASS stress, and anxiety, with medium effect sizes and for the DASS depressiveness, PANAS negative affect, and FFMQ mindfulness level with large effect sizes (Table 4).Only PANAS positive affect measures did not show a significant All other variables were analyzed using a a two-tailed Student's unpaired t-test.x -, mean; rM, standard error of the mean.
a Categorical data was analyzed using the Fisher-Freeman-Halton exact test.
b Data was analyzed using Mann-Whitney U test, depending on whether parametric assumptions were met.The percentage of participants taking each categories of medication was calculated as a percentage of patients who took medication, while percentages of participants who were receiving psycho-or physiotherapy were calculated as percentages of the total number of participants in each group.Fisher-Freeman-Halton exact test showed no significant difference in medication intake between the two groups (P = 0.124).
a The anxiolytics used by the two patients was the tricyclic dibenzazepine Opipramol.b .Sedatives used were the phenothiazine derivative Promethazine and the nonbenzodiazepine Zopiclone.main effect of time.No significant time*group interaction was observed for any of the outcome variables.Bonferroni-corrected pairwise comparisons showed improved scores from pre-combined therapy to immediately post-combined therapy in affective pain level (P = 0.023), sleep quality (P = 0.019), depressiveness (P = 0.013), anxiety (P = 0.05), and negative affect (P = 0.006), but no further improvement from T2 to T3 (P > 0.05 for all outcomes).Stress reduction was observed after 4 weeks post-stimulation compared to T1 (P = 0.048), but no immediate effects of the combined therapy were observed.Compared to T1, we observed a significant enhancement in the mindfulness level of all participants after combined therapy (T2; P < 0.001), which was not further increased from T2 to T3 (P = 1.00).Hence, the immediate therapeutic benefits of the combined intervention on most clinical outcomes were observed, irrespective of group.

TMS outcomes
Both the active + MM and sham + MM groups reported a significant increase in the S50 of the I-O curve (F(2,52) = 4.40, P = 0.017, gp 2 = 0.145) over time with a large effect size.Post hoc pairwise comparisons demonstrated only a higher S50 at T3 compared to T1 (MD = 3.50, P = 0.017).We also found a large main effect of group for ICF (F(1,24) = 5.17, P = 0.032, gp 2 = 0.177), with a larger   ICF in the active + MM group, compared to the sham group (MD = 0.250, P = 0.032).Nevertheless, no significant interaction between group and time was noted for any of the TMS measures (Table 5).

Feasibility and side effects
The average rating of the overall easiness of preparation of the device and accessories and easiness of the use of the device was 7.72 (standard error of the mean [r M ] = 0.398) and 9.20 (r M = 0.255) respectively for the whole sample.When comparing the average scores between the two groups, no significant differences were detected (easiness of preparation: t(36) = -0.031,P = 0.975; easiness of use: t(36) = À0.591,P = 0.558).
None of the patients discontinued the stimulation because of side or adverse effects of tDCS or needed a medical intervention following the stimulation.The most common side effect after stimulation was redness on the skin under the electrodes, which was reported by 83.3% of participants in the active + MM group and by 68.4% of participants in the sham group.Hheadache was the most reported adverse effect (active + MM = 61.1% and sham + M M = 68.4%).Also, we did not find any significant differences in side effect occurrences between the two treatment groups over the combined intervention (all P > 0.05; for more information, Supplementary Table S4).Furthermore, no differences in the total rating for each patient over the 10 days between the two groups were found for either side effect after tDCS.Participants did not report any long-lasting adverse effects or any serious adverse event following the combined intervention.For side effects experienced as 'strong' after the stimulation, redness on the skin under the electrodes was the most common with 4 reports in the active group and 2 in sham.

Blinding success
The correct guesses on the type of stimulation received by the participants at (17 out of 37 patients, 45.9%; active + MM: n = 8, sham + MM: n = 9) were at the same level as chance (v 2 = 0.248, P = 0.618), demonstrating successful blinding.Ten patients in the sham + MM believed they got the active stimulation protocol.

Discussion
The main objective of this triple-blind sham-controlled randomized study was to underpin the clinical efficacy, primarily pain reduction and improvement in QOL, of pairing anodal M1-tDCS and MM in FM compared to sham in patients formally trained in mindfulness.Contrary to our hypothesis, both intervention groups demonstrated medium reductions in pain intensity and improvements in QOL throughout the study.Immediate reductions in affective pain level and enhancements in sleep quality, depressiveness, anxiety, stress, and negative affect, with medium to large effect size and successful blinding was maintained throughout the study.The lack of group*time interaction failed to show the efficacy of combining MM with home-based supervised tDCS over sham tDCS.We found an increase in S50 over time for both groups and a higher ICF in the active + MM group, without any significant interaction.Our study showed that the combined MM and tDCS intervention was well tolerated and relatively safe with no serious adverse effects.At-home tDCS under supervision was feasible and easy to administer, similar to previous studies (Pilloni et al., 2022).
Despite the Level A recommendation for the use of repeated anodal tDCS over the left M1 in FM (Baptista et al., 2019;Fregni et al., 2021) and the significant pain reduction observed in earlier clinical studies (Fregni et al., 2006;Khedr et al., 2017;Valle et al., 2009), recent systematic reviews and meta-analyses discussed the heterogeneous findings across different studies with variable stimulation protocols (Cheng et al., 2023;Conde-Antón et al., 2023;Coskun Benlidayi, 2020;Lloyd et al., 2020;Molero-Chamizo et al., 2023;Moshfeghinia et al., 2023;Teixeira et al., 2023).The largest clinical trial to date, with a minimum of 30 patients per group and 15 sessions of 20 min tDCS at 2 mA, showed that left M1-anodal tDCS was no more effective than placebo in terms of pain, QOL, and other clinical outcomes in FM (Samartin-Veiga et al., 2022a;Samartin-Veiga et al., 2022b).Our study also failed to replicate the positive findings of previous studies (Ahn et al., 2019;Pollonini et al., 2020;Ramasawmy et al., 2022) showing effectiveness of anodal tDCS and MM in chronic pain management.The significant pain reduction previously observed in patients with osteoarthritis following concurrent MM and tDCS (Ahn et al., 2019;Pollonini et al., 2020) must be interpreted with caution, considering that only a single NRS value was measured as baseline.Furthermore, the sham group in our study pairing sham tDCS with active MM was different from that of the aforementioned studies (Ahn et al., 2019;Pollonini et al., 2020), which combined sham MM with sham tDCS.In our study, the mindfulness level from T1 to T3 increased largely but did reach statistical significance from T2 to T3, despite continuing mindfulness practice.This can be explained by a reduction in the frequency of mindfulness practice from daily between T1 and T2 to 3 days per week from T2 to T3.This is in line with previous studies showing more frequent MM practice to result in increased beneficial therapeutic effects in FM (Adler-Neal & Zeidan, 2017;Haugmark et al., 2019).Furthermore, the significant improvements in QOL and FMassociated symptoms were immediate effects and the symptoms did not improve further from T2 to T3.Our clinical findings and change in mindfulness level hint towards a potential effect of the daily MM practice over the combined intervention weeks, rather than a synergistic effect of stimulation and meditation.Additional long-term impact was limited due to no further change in mindfulness level from T2 to T3.Almost half of the participants in both sham and active groups believed that they received anodal tDCS.The patients' beliefs on the stimulation type received could play a decisive role in determining the clinical outcomes, producing a placebo-like effect (Braga et al., 2021;Rabipour et al., 2018;Ray et al., 2019).Subjective beliefs about receiving the real intervention could even predict the studied outcome better than the actual therapy (Fassi and Kadosh, 2020), though this has not yet been studied in tDCSmediated analgesia.
In our exploration of TMS cortical excitability parameters (Karabanov et al., 2015;Paulus et al., 2008;Ziemann et al., 2008), a noteworthy time effect for S50 of the I-O curve was observed.Unlike the slope of the I-O curve, which is a key measure of the strength of the corticospinal tract (Ridding & Rothwell, 1997), the less reported S50 measure is a measure of the stimulus threshold (Devanne et al., 1997).S50 can be used as an ideal test stimulus intensity to assess basal cortical excitability changes as it produces a half-maximal response, which can be modified experimentally to either demonstrate facilitation or inhibition (Kukke et al., 2014).In this study, the increase in S50 value indicates a higher stimulus threshold at T3 than T1, meaning that a higher stimulation intensity is required to produce an MEP halfway to saturation.We suggest that this change in S50 could be attributed to alterations in the brain activity following regular practice of mindfulness.Nevertheless, no previous study investigated the relation between MM and S50 of the I-O curve.The large significant group effect for ICF, with higher facilitation observed in the active + MM group, did not lead to any meaningful interpretation for this investigation.Finally, the lack of significant effect of the combined intervention over time was observed for any of the TMS measures, making the synergic effects of the two therapeutic methods inconclusive.We should also note that the sample size for investigating the cortical excitability alterations was underpowered due to dropouts.Our findings did not replicate previously demonstrated enhancement in ICI, neither regarding enhanced mindfulness levels (Guglietti et al., 2013) nor the normalization of central disinhibition in FM (Pacheco-Barrios et al., 2022).It must be noted that the study by Guglietti et al. (2013) used a 60-min meditation session that was not purely mindfulness-based, and their sample comprised healthy individuals.Antal et al. (2010) is the sole clinical study to have tested the impact of repeated tDCS on cortical excitability and found a reduction in SICI.However, their study involved a heterogeneous patient group with different types of chronic pain and tested M1-anodal tDCS sessions for 5 days at 1 mA.
Our findings must be interpreted in the light of the study's limitations.The lack of monotherapy intervention groups and waitinglist controls prevented us from making meaningful interpretations on the add-on effects of combining MM and tDCS to the effects of MM or tDCS alone in FM.On account of the inter-subject variability arising from internal factors including skull thickness, cortex morphology, genetic profile, and sex (Vergallito et al., 2022), the same dose of stimulation given to each participant may not cause the same effect in the brain, which is a common limitation of all tDCS clinical trials.This issue could be solved by using a within-subject design.Taken together, our findings challenge the combination of anodal tDCS of the left M1 and MM as a potentially effective therapy in FM, despite the blinded, sham-controlled, and randomized design and powered sample size used in the study.However, the negative findings do not exclude the potential therapeutic value of using tDCS per se in FM, and even in a combined protocol, whether tDCS is applied to other single cortical targets such as the prefrontal targets or to multiple targets.
FM is a syndrome comprising a spectrum of somatic and psychological symptoms, with each contributing to different degrees and compositions of functional disability and symptom burden (Bartley et al., 2018;Chandran et al., 2012;Häuser et al., 2008).Future studies should define the target of tDCS based on the more prominent symptom profile of each participant: M1 for somatic and DLPFC for psychological impairments (Lefaucheur et al., 2017).In our study, the exclusion of more than 50% of patients assessed for eligibility demonstrates that the included sample is not representative of the general FM population, a frequent observation made in randomized clinical trials (Sherman et al., 2016).Therefore, findings from clinical trials need to be supplemented with evidence from real-world studies (Luce et al., 2010) to provide a more comprehensive picture of the benefits and drawbacks of therapies as they are used in clinical practice (Blonde et al., 2018).Unlike tDCS which causes changes in cortical excitability (Woods et al., 2016), transcranial alternating current stimulation (tACS) (Brighina et al., 2019;Wandrey et al., 2022) and MM (Day et al., 2021;Lomas et al., 2015) both modulate endogenous brain oscillations in chronic pain.The preliminary efficacy of tailored tACS combined with physiotherapy was tested in FM patients (Bernardi et al., 2021).Despite the heterogeneity of findings across studies exploring the effects of MM on brain state using EEG, there is a concensus on an increased theta activity in frontal midline during MM as a reliable neurophysiological signature and even in the resting state after MM (Aftanas & Golocheikine, 2001;DeLosAngeles et al., 2016;Katyal & Goldin, 2021;Lomas et al., 2015;Takahashi et al., 2005;Tang et al., 2019).Future research is required to test whether combining MM with tACS applied over the frontal cortices to induce theta waves might be a more synergistic intervention than pairing it with tDCS.

Fig. 1 .
Fig. 1.Consolidated Standards of Reporting Trials flow diagram of the participants with fibromyalgia throughout the clinical trial comparing active + MM to sham + MM groups.

Fig. 2 .
Fig. 2. Pain intensity scores as indexed by numerical rating scale (NRS).Higher NRS scores indicate higher pain levels.Bars show standard error of the mean.

Fig. 3 .
Fig. 3. Quality of life as assessed by scores of the Fibromyalgia Impact Questionnaire (FIQ).Higher FIQ scores illustrating lower life quality.Bars show standard error of the mean.** P < 0.01.

Table 1
Participant protocol timeline.

Table 2
Participant demographics at baseline.

Table 3
Treatment profile of participants.

Table 4
Mixed model ANOVA for clinical outcomes and mindfulness level comparing active + MM group (n = 18) to sham + MM group (n = 19).

Table 5
Mixed model ANOVA for measures of cortical excitability using TMS comparing active + MM group to sham + MM.