Open AccessArticle Real-World Safety and Tolerability of Low-Intensity Repetitive Transcranial Magnetic Stimulation in Fibromyalgia: A Multicenter Observational Cohort Study by Nazario Felix-Gonzalez Nazario Felix-Gonzalez Scilit Preprints.org Google Scholar 1,*, Jose-Maria Gomez-Arguelles Jose-Maria Gomez-Arguelles Scilit Preprints.org Google Scholar 2 and Ceferino Maestu-Unturbe Ceferino Maestu-Unturbe Scilit Preprints.org Google Scholar 3 1 Departamento de Arquitectura y Tecnología de Sistemas Informáticos, Escuela Tecnica Superior de Ingenieros Informaticos, Universidad Politecnica de Madrid, 28223 Madrid, Spain 2 Unidad de Neurología, Hospital Universitario Quironsalud Madrid, 28223 Madrid, Spain 3 Centro de Tecnologia Biomedica, Universidad Politecnica de Madrid, 28223 Madrid, Spain * Author to whom correspondence should be addressed. J. Clin. Med. 2026, 15(12), 4452; https://doi.org/10.3390/jcm15124452 (registering DOI) Submission received: 9 May 2026 / Revised: 4 June 2026 / Accepted: 5 June 2026 / Published: 9 June 2026 Abstract Background/Objectives: Low-intensity repetitive transcranial magnetic stimulation (Li-rTMS) is an emerging non-invasive neuromodulation approach for fibromyalgia; however, large-scale data on safety and tolerability in routine clinical settings remain limited. This study aimed to evaluate the safety and tolerability of Li-rTMS in patients with fibromyalgia treated in outpatient practice. Methods: A retrospective multicentre observational cohort study was conducted using routinely collected clinical data from nine outpatient centres (November 2020–March 2026). Safety analyses included all patients initiating Li-rTMS treatment (intention-to-treat cohort, n = 1381). The protocol consisted of eight stimulation sessions delivered using a CE-certified medical device. Patient-reported tolerability and perceived symptom changes in three predefined domains (headache, generalized pain, and sleep) were assessed using a standardized 0–10 scale and analysed descriptively, with stratification by age group. Results: Li-rTMS was generally well tolerated across all age groups, including those >65 years. High-intensity symptom worsening was infrequent (headache 3.8%, generalized pain 6.1%, sleep disturbances 2.1%), predominantly mild and self-limited. No serious adverse events were spontaneously reported or clinically identified during routine clinical follow-up, within the limitations inherent to a real-world passive registry framework. Exploratory observations from patient-initiated recall sessions did not suggest an increased adverse symptom burden with repeated exposure. Conclusions: Li-rTMS demonstrated a favourable patient-reported tolerability profile among patients with available follow-up data, characterized by a low frequency of transient high-intensity symptom worsening and the absence of serious adverse events spontaneously reported or clinically identified during routine, non-systematic follow-up. These findings support the feasibility of repeated Li-rTMS administration in outpatient settings; however, due to the observational design and non-systematic follow-up, no conclusions regarding clinical efficacy can be drawn. Keywords: fibromyalgia; neuromodulation; low-intensity rTMS; real-world evidence; safety; tolerability 1. Introduction Current management strategies for FM remain challenging due to the heterogeneous and fluctuating nature of symptoms. Pharmacological approaches frequently provide limited or inconsistent benefits and are often associated with poor long-term adherence. Consequently, non-pharmacological and multimodal approaches, including exercise-based interventions, cognitive and behavioral therapies, and neuromodulation techniques, have gained increasing attention as complementary therapeutic strategies [ 3, 7]. This gap is especially relevant in chronic conditions such as fibromyalgia, where long-term adherence is essential and tolerability plays a major role in treatment feasibility, as poorly tolerated interventions are more likely to be discontinued. Observational studies conducted in routine care settings can complement controlled trials by providing information on treatment tolerability in heterogeneous patient populations. In this context, the present multicenter observational cohort study was designed to evaluate the safety and tolerability of Li-rTMS in patients with FM. This study reports data from a large multicentre cohort of 1381 patients, providing extensive real-world information on Li-rTMS tolerability. Additionally, exploratory analyses were conducted to describe tolerability during patient-initiated recall sessions. 2. Materials and Methods 2.1. Study Design and Setting A retrospective multicenter observational cohort study was conducted using routinely collected clinical data from nine outpatient centers specialized in chronic pain management between November 2020 and March 2026. No experimental procedures or protocol-driven interventions were introduced beyond standard clinical care. The study was conducted and reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) recommendations for observational cohort studies. Given the observational design, the study was not based on formal hypothesis testing but aimed to descriptively evaluate the safety and tolerability profile of Li-rTMS in real-world clinical practice. 2.2. Participants and Eligibility Criteria Eligible participants were adults aged 18 years or older with a clinical diagnosis of FM according to the 2010/2016 American College of Rheumatology (ACR) criteria [ 2]. Patients were included if they initiated the Li-rTMS treatment protocol during the study period and completed the initial pre-treatment assessment before stimulation. Patients were excluded from treatment according to routine clinical safety criteria, including epilepsy or seizure disorders, severe uncontrolled psychiatric conditions, pregnancy or breastfeeding, implanted electronic medical devices, or other contraindications for magnetic stimulation established by the treating physician. The initial database included 1446 clinical records. Data cleaning procedures were performed to ensure record integrity and remove duplicated or incomplete entries. The final intention-to-treat (ITT) cohort used for safety analyses included 1381 unique patients. Patients attending at least one patient-initiated recall session following completion of the induction protocol were additionally identified for exploratory descriptive analyses related to repeated Li-rTMS exposure under routine clinical practice conditions. 2.3. Li-rTMS Intervention Protocol Li-rTMS was delivered using a commercially available CE-certified Class IIa medical device (MINESTIM ପ୍ପ, CTB, Madrid, Spain) intended for low-field transcranial magnetic stimulation in fibromyalgia-associated pain management. The device operates according to MDR Rule 9 and incorporates predefined manufacturer-programmed stimulation parameters. From an engineering and verification standpoint, the electrical performance of the device was characterized through direct measurements conducted during functional testing by the technical responsible team at the manufacturer. These measurements confirmed that the stimulation system delivers a peak-to-peak current of approximately 800 µA following a square waveform at 8 Hz. The current flows through a network of 32 microcoils integrated into a helmet-based applicator, spatially distributed over the scalp according to a modified international 10–10 system, providing global and symmetrical cranial coverage. The coils are electrically connected in series configuration that ensures homogeneous current distribution under standardized operating conditions, with device tolerances maintained within ±10%. Although direct measurement of magnetic flux density was not performed, the magnetic field generated by each microcoil was estimated from the measured current and coil geometry using standard electromagnetic formulations. Based on these calculations, the magnetic flux density at the coil level is in the order of approximately 1–2 µT (≈1.5 µT). These values represent theoretical engineering estimates under standardized operating conditions rather than direct in vivo biophysical measurements. This level is several orders of magnitude lower than the millitesla (mT) to tesla (T) range typically used in conventional high-intensity rTMS systems. Consequently, these values should be interpreted strictly as order-of-magnitude engineering estimates rather than direct measurements in biological tissues. The authors recognize that future physical investigations incorporating direct laboratory magnetic field measurements are required to conclusively validate these calculations and further strengthen the technical characterization of the device. Participants underwent an induction protocol consisting of eight Li-rTMS sessions of 20 min each, administered under routine outpatient conditions. All sessions were supervised by trained healthcare personnel previously instructed in device operation and standardized treatment procedures. No anesthesia, sedation, or analgesic premedication was required. The selection of stimulation frequency (8 Hz) and session duration (20 min) follows the manufacturer-defined protocol and is consistent with previous experimental and clinical studies on low-intensity magnetic stimulation. Frequencies in the low alpha range (8–10 Hz) have been associated with modulation of cortical network activity and functional brain rhythms without inducing direct neuronal depolarization. In the context of Li-rTMS, such frequencies are considered suitable for producing neuromodulatory effects while maintaining a favourable safety profile. The session duration of 20 min was selected to ensure adequate exposure time for cumulative neuromodulatory effects under low-intensity conditions, while remaining well within established safety limits for repetitive magnetic stimulation. Given that Li-rTMS operates at substantially lower field strengths than conventional high-intensity rTMS, the applied frequency and duration are not associated with the safety constraints typically linked to suprathreshold stimulation protocols, such as seizure risk or motor threshold considerations. Treatment sessions were conducted under controlled environmental conditions using non-ferromagnetic examination beds and low-electromagnetic-noise environments routinely employed for Li-rTMS administration. Standard sensory isolation measures, including ear protection and visual shielding, were applied according to the device operating protocol. Following completion of the induction phase, additional recall sessions were available upon patient request as part of routine clinical management. Recall visits were not protocolized and were analyzed descriptively as exploratory observations. 2.4. Data Collection Clinical data were collected within a post-market clinical follow-up (PMCF) program across specialized outpatient centres. All participating sites followed a standardized data collection workflow to ensure consistency in treatment delivery, assessment timing, and data recording. Data were collected using a unified electronic questionnaire implemented on a secure digital platform (Google Forms). The questionnaire included predefined mandatory fields and was identical across all sites. The questionnaire was administered by trained nursing personnel, who were specifically trained in both the Li-rTMS procedure and the PMCF data collection protocol. Data entry was performed using password-protected tablet devices with restricted access. Patients completed the questionnaire prospectively during routine visits prior to selected treatment sessions. Responses were subsequently reviewed and validated by the treating physician before inclusion in the centralized PMCF database, ensuring data completeness and clinical oversight. Assessments were conducted at predefined milestones: (1) baseline (prior to the first session), (2) intermediate (prior to the fourth session), and (3) end-of-treatment (prior to the eighth session). Additional data were collected during optional patient-initiated recall sessions. The baseline assessment was used to characterize the study cohort but did not include treatment-related symptom evaluation, as no Li-rTMS exposure had occurred. Subsequent assessments captured patient-reported responses to treatment. To enhance transparency and reproducibility, the PMCF workflow, questionnaire, and a representative anonymized dataset are publicly available in an open-access repository [ 17]. Data were stored in a centralized encrypted database hosted on institutional servers at the Universidad Politécnica de Madrid (UPM), with restricted access controlled by the project’s Technical Director. Data cleaning, preprocessing, and statistical analyses were performed using a standardized automated pipeline to ensure consistency and reproducibility. 2.5. Data Management All clinical data were anonymized at the source prior to analysis using unique patient identification codes assigned by each participating center. No directly identifiable personal information was included in the analytical dataset. Following physician validation, data were transferred to a centralized PMCF database for aggregation and analysis. The anonymized dataset was stored on a secure institutional cloud platform compliant with Regulation (EU) 2016/679 (General Data Protection Regulation, GDPR). Data processing and statistical analyses were performed by the study investigators using the anonymized dataset, ensuring consistent handling of information and preventing access to patient identifiers throughout the analytical process. This structured, multi-level data management approach—including trained staff involvement, physician validation, and centralized data handling—was designed to ensure data integrity, minimize variability across centers, and support the reliability of the study findings. 2.6. Safety and Tolerability Outcomes Safety and tolerability outcomes were derived from structured patient-reported responses collected at predefined clinical assessment timepoints, rather than through a formal adverse event reporting system. Accordingly, outcomes reflect patient-perceived changes in symptoms under routine clinical conditions rather than formally adjudicated safety surveillance data. Assessments focused on three predefined symptom domains: (1) headache, (2) generalized pain, and (3) sleep-related disturbances. These domains were evaluated using a standardized 0–10 numerical rating scale (NRS), designed to assess perceived change in symptoms relative to the pre-treatment condition rather than absolute symptom severity. In this scale, 0 indicates maximum improvement, 5 no change, and 10 maximum worsening. “Generalized pain” refers to overall perceived widespread pain typical of fibromyalgia, expressed as change from baseline. Data were collected at predefined treatment milestones: baseline (prior to the first session), intermediate (prior to the fourth session), and end-of-treatment (prior to the eighth session), as well as during optional patient-initiated recall sessions. The baseline assessment was used to characterize the study cohort and did not include treatment-related evaluation. At subsequent visits, patients rated their symptom changes based on the preceding treatment sessions. Each patient contributed one assessment per domain at each timepoint; therefore, outcomes are reported on a per-patient basis rather than per session. For descriptive analysis, responses were categorized as improvement (0–4), no change (5), or worsening (6–10). Tolerability was operationally defined using transient high-intensity symptom worsening, defined as scores ≥8 in the headache or generalized pain domains at a given assessment. These responses were interpreted as clinically relevant short-term worsening occurring during or shortly after Li-rTMS exposure, based on routine clinical evaluation and physician validation. Data were collected prospectively using a standardized electronic questionnaire administered by trained nursing personnel at predefined visits (Sessions 1, 4, and 8) prior to clinical evaluation. Patients did not complete the tolerability scale at baseline, as no stimulation had yet occurred. At intermediate and end-of-treatment visits, responses reflected the tolerability of the preceding sessions. Any major or unexpected clinical events not captured by the questionnaire were documented if spontaneously reported by patients during subsequent medical follow-up visits. This approach captures patient-reported tolerability outcomes rather than formally adjudicated adverse events derived from active standardized surveillance systems. Accordingly, the absence of serious adverse events within this cohort should be interpreted within the context of routine, non-systematic clinical follow-up, where no clinically significant treatment-related complications were spontaneously reported or clinically identified. Exploratory analyses were performed using data from patient-initiated recall sessions to describe symptom evolution and tolerability during repeated Li-rTMS exposure. As these assessments were non-protocolized and patient-driven, they were analysed descriptively and considered subject to potential selection bias. 2.7. Data Curation and Cleansing Protocol Prior to statistical modeling, data extracted from the federated multicenter database underwent a multi-stage deterministic curation pipeline to ensure data integrity, structural consistency, and longitudinal traceability. The curation process was systematically structured as follows: Data Aggregation and Standardization: Raw electronic data capture (EDC) sheets from all participating clinical centers were consolidated into a unified relational structure. Variables were normalized, and clinical site codes were cross-verified against institutional registry logs. Deduplication Audit: To avoid artificial statistical inflation and administrative noise, an automated deduplication protocol was applied. Records were flagged as redundant if they shared identical unique Clinical History Numbers (IP) alongside overlapping timestamps or duplicate questionnaire submissions within the same evaluation window. These multi-entry redundancies were systematically reviewed and collapsed into single, validated baseline records. Completeness and Quality Filters: A strict case-complete auditing protocol was enforced. Records missing critical baseline identifiers—such as biological sex, age, clinical centre origin, or core initial safety and symptomatic metrics—were excluded from subsequent analyses. This deterministic filtering approach was chosen to prevent the introduction of imputation biases in the initial demographic characterization. Longitudinal Traceability: Patients who successfully fulfilled the baseline data integrity criteria were assigned a verified, anonymized tracking identifier. This identifier was utilized to dynamically map individuals across all subsequent follow-up intervals—including intermediate assessments, acute treatment termination, and patient-initiated maintenance (recall) sessions—ensuring strict consistency throughout the study timeline. 2.8. Statistical Analysis Descriptive statistics were used to summarize demographic characteristics and patient-reported outcomes. Continuous variables are presented as mean ± standard deviation (SD), and categorical variables as frequencies and percentages. Analyses were performed using all available data at each assessment timepoint. Due to the observational design and variable follow-up, outcomes were evaluated using a complete-case approach with phase-specific denominators. Patient-reported symptom change (categorized as improvement, no change, and worsening) was described across predefined treatment phases (intermediate, end-of-treatment, and recall). Tolerability was assessed based on the proportion of patients reporting high-intensity symptom worsening (NRS ≥ 8) in the headache or generalized pain domains. Comparisons between age groups ( 65 years) were performed at each predefined milestone using Pearson’s chi-square (χ 2) test or Fisher’s exact test when appropriate. Proportions are presented with 95% confidence intervals (CI) calculated using Wilson’s method. Exploratory analyses were conducted in patients who attended recall sessions to describe symptom evolution and tolerability during repeated exposure. These analyses were descriptive and were not subjected to formal hypothesis testing due to their non-protocolized and self-selected nature. Missing data were not imputed. A two-sided p-value 65 Years (Inter: n = 83/Final: n = 61) p-Value (Inter) 2p-Value (Final) 3Headache 0.24 0.58 Intermediate Phase 12/235 (5.1%) [2.9–8.7] 19/503 (3.8%) [2.4–5.9] 3/83 (3.2%) [1.0–9.5] End-of-Treatment Phase 3/144 (2.3%) [0.8–6.3] 10/349 (3.0%) [1.7–5.4] 2/61 (2.4%) [0.5–9.9] Generalized Pain 0.47 0.62 Intermediate Phase 13/235 (5.6%) [3.3–9.3] 32/503 (6.4%) [4.6–8.9] 5/83 (5.9%) [2.5–13.2] End-of-Treatment Phase 7/144 (4.7%) [2.3–9.5] 17/349 (5.0%) [3.2–7.8] 2/61 (3.5%) [1.0–11.5] Sleep Disturbance 0.68 0.49 Intermediate Phase 5/235 (2.2%) [1.0–5.0] 10/503 (1.9%) [1.0–3.5] 2/83 (2.7%) [0.8–8.8] End-of-Treatment Phase 2/144 (1.2%) [0.3–4.6] 5/349 (1.4%) [0.6–3.3] 1/61 (1.2%) [0.2–8.0] Serious Adverse Events (SAEs) Spontaneously Reported or Clinically Identified During Routine Follow-up Intermediate Phase 0/235 (0%) 0/503 (0%) 0/83 (0%) End-of-Treatment Phase 0/144 (0%) 0/349 (0%) 0/61 (0%) 1 Significant transient symptom worsening defined as NRS ≥ 8. The identification of serious adverse events (SAEs) was based on spontaneous patient reports and routine clinical follow-up rather than on a formal active safety surveillance system. 2 p-value derived from the Pearson chi-square test (χ 2), analyzing the intergroup distribution by age ranges in the intermediate phase. 3 p-value derived from the Pearson chi-square test (χ 2), analyzing the intergroup distribution by age ranges at the end of treatment. Percentages are calculated using phase-specific denominators. Notes: Data are presented as n/N (%), with 95% confidence intervals (CI) calculated using Wilson’s score method in brackets. Percentages and confidence intervals correspond to the originally computed values. Counts (n) have been derived from reported percentages and denominators and may not exactly reproduce CI limits due to rounding. Percentages are based on phase-specific denominators (complete-case analysis), corresponding to the number of patients with available data at each assessment phase. Table 3. Patient-reported perceived symptom change and tolerability across treatment phases. Table 3. Patient-reported perceived symptom change and tolerability across treatment phases. Clinical Parameter Intermediate (n = 821) End of Treatment (n = 554) Recall Session f(R1, n = 144) Pain-Related Symptom Assessment Improvement (NRS 0–4) a432/821 (52.70%) 421/554 (76.10%) 103/144 (71.40%) Stable/No Change (NRS 5) b355/821 (43.20%) 120/554 (21.60%) 37/144 (25.50%) Symptom Worsening (NRS 6–10) c34/821 (4.10%) 13/554 (2.30%) 4/144 (3.10%) Headache-Related Symptom Assessment Improvement (NRS 0–4) a485/821 (58.20%) 40/554 (72.40%) 99/144 (68.90%) Sleep-Related Symptom Assessment Improvement (NRS 0–4) a409/821 (49.80%) 357/554 (64.50%) 88/144 (61.20%) Tolerability Indicator Transient Symptom Exacerbations (NRS ≥ 8) d51/821 (6.24%) 12/554 (2.14%) 3/144 (2.08%) [4.79–8.10] [1.23–3.71] [0.71–5.95] Stratified by Age Group 65 years 5/83 (6.40%) 1/61 (2.13%) 1/23 (4.35%) [3.28–12.11] [0.58–7.44] [0.77–21.01] Age-group Comparison ( p-value) ep = 0.868 p = 0.897 p = 0.708 Serious Adverse Events (SAEs) Spontaneously Reported or Clinically Identified During Routine Follow-up 0% 0% 0% a Improvement corresponds to patient-reported scores of 0–4 on the treatment assessment scale (0 = maximum improvement, 10 = maximum worsening). b Stable/No change corresponds to a neutral score of 5 on the assessment scale. c Symptom worsening corresponds to scores of 6–10 on the assessment scale. d Transient symptom exacerbations were defined as acute high-intensity responses (NRS ≥ 8) in headache- or pain-related assessments during the corresponding treatment phase. Serious adverse events reflect the absence of major treatment-related complications spontaneously reported or clinically identified during routine, non-systematic outpatient follow-up. e p-values were computed using Pearson’s chi-square (Χ 2) test to compare the distribution of transient exacerbations across the three age strata ( 65 years) within each specific treatment phase. f Recall sessions were patient-initiated and analysed descriptively as exploratory observations under routine clinical practice conditions. Notes: The assessment scale reflects perceived change relative to pre-treatment symptom status rather than absolute symptom intensity. Percentages are calculated using phase-specific denominators (complete-case analysis), corresponding to the number of patients with available data at each treatment phase (Intermediate: n = 821; End-of-treatment: n = 554; Recall: n = 144), rather than the full intention-to-treat (ITT) cohort. Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. © 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license. Share and Cite MDPI and ACS Style Felix-Gonzalez, N.; Gomez-Arguelles, J.-M.; Maestu-Unturbe, C. Real-World Safety and Tolerability of Low-Intensity Repetitive Transcranial Magnetic Stimulation in Fibromyalgia: A Multicenter Observational Cohort Study. J. Clin. Med. 2026, 15, 4452. https://doi.org/10.3390/jcm15124452 AMA Style Felix-Gonzalez N, Gomez-Arguelles J-M, Maestu-Unturbe C. Real-World Safety and Tolerability of Low-Intensity Repetitive Transcranial Magnetic Stimulation in Fibromyalgia: A Multicenter Observational Cohort Study. Journal of Clinical Medicine. 2026; 15(12):4452. https://doi.org/10.3390/jcm15124452 Chicago/Turabian Style Felix-Gonzalez, Nazario, Jose-Maria Gomez-Arguelles, and Ceferino Maestu-Unturbe. 2026. "Real-World Safety and Tolerability of Low-Intensity Repetitive Transcranial Magnetic Stimulation in Fibromyalgia: A Multicenter Observational Cohort Study" Journal of Clinical Medicine 15, no. 12: 4452. https://doi.org/10.3390/jcm15124452 APA Style Felix-Gonzalez, N., Gomez-Arguelles, J.-M., & Maestu-Unturbe, C. (2026). Real-World Safety and Tolerability of Low-Intensity Repetitive Transcranial Magnetic Stimulation in Fibromyalgia: A Multicenter Observational Cohort Study. Journal of Clinical Medicine, 15(12), 4452. https://doi.org/10.3390/jcm15124452 Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here. Article Metrics Article metric data becomes available approximately 24 hours after publication online.
