Migraine is one of the most prevalent primary headache disorders worldwide and is associated with substantial impairments in quality of life, work productivity, and psychosocial functioning. In Germany alone, nearly 18 million people are affected. Although pharmacological therapy remains a cornerstone of migraine management, non-pharmacological interventions-particularly physiotherapy-play an important role in multimodal treatment concepts and are generally well tolerated.

Recent evidence suggests that physiotherapy for migraine should extend beyond conventional manual techniques and include patient education, aerobic endurance training, and self-management strategies. However, in routine clinical practice, physiotherapeutic approaches vary widely and are often not fully aligned with current evidence-based recommendations. Moreover, there is a lack of pragmatic randomized controlled studies evaluating the effectiveness of structured, evidence-based multimodal physiotherapy compared with conventional physiotherapy under real-world clinical conditions.

This study is a randomized controlled pilot trial designed to compare an evidence-based multimodal physiotherapeutic intervention with conventional physiotherapy in adult patients with migraine. The multimodal intervention consists of evidence-based physiotherapy, structured patient education on migraine and pain mechanisms, guided moderate aerobic endurance training, and relaxation techniques. The control group receives conventional physiotherapy according to German standard clinical practice.

The study is designed as a randomized controlled trial with a pre-post design. Adult patients aged 18 to 60 years with a specialist-confirmed diagnosis of migraine will be recruited from a specialized pain therapy center in Osnabrück, Germany. Participants will be randomly allocated to either the multimodal treatment group or the usual care physiotherapy group. The intervention period lasts three months.

The primary outcome is migraine-related quality of life measured using the Headache Impact Test (HIT-6™). Secondary outcomes include headache frequency and intensity assessed via headache diaries, migraine-related work absenteeism, and functional impairment. Feasibility outcomes will be evaluated through the feasibility and acceptance of the multimodal treatment program from the perspective of the treating physiotherapists.

The results are expected to provide robust evidence on the effectiveness and real-world applicability of multimodal treatment programs compared with usual care for migraine management in Germany.

Embedded to this project there is also the prediction analysis. For that patients will be divided into those who responded to the treatment and those who stay stable or get worse after treatment. Based on that a statistical analysis will be used to assess which factors influenced the improvement of those patients after physical therapy treatment.

2. Scope: Migraine can significantly impair quality of life and pose substantial economic burdens to those affected and society. Migraine, a common neurological disorder that often begins in early life, can lead to developmental limitations, high rates of suicide, and reduced work productivity.

Migraine could be treated with various physiotherapeutic modalities, especially exercises. However, selecting the appropriate treatment approach can be challenging due to the need for personalized, evidence-based interventions.

At centers like Schmerztherapiezentrum Dr. Brau und Dr. Michel and Capu Motion in Osnabrück, interdisciplinary treatments for migraine are being explored, with a strong emphasis on physiotherapy, which offers a non-invasive alternative without the side effects associated with drugs. While recent advancements in evidence-based practices have been made, their implementation in clinical settings remains scarce.

This project aims to optimize physiotherapy interventions by integrating recent evidence, emphasizing a combination of patient education, endurance training, specific exercises and techniques, and self-management. By advancing research and applying new evidence in practice, the project seeks to improve patient outcomes, enhance treatment effectiveness, and reduce healthcare costs, ultimately improving quality of life for individuals affected by migraine. In addition, this project seeks to understand the interactions among biopsychosocial factors, including demographic diversity, physical assessment, and psychological variables, which are crucial for tailoring interventions. Additionally, considering variations in health systems, treatments, infrastructure, and clinical practices, especially within Germany, is essential for developing comprehensive treatment strategies for patients suffering from these complex conditions. This research aims to identify the most effective treatment methods within the German clinical context, potentially serving as a model for other countries and improving global treatment approaches for patients with migraine.

This pilot trial has several objectives:

Primary Objective: To determine whether a three-month structured multimodal treatment program leads to greater improvement in migraine-related quality of life compared with usual care, as measured by the Headache Impact Test (HIT-6™).

Secondary Objectives: 1) To compare changes in headache frequency and pain intensity between groups; 2) To assess differences in migraine-related work absenteeism and functional impairment; 3) To evaluate adherence to and implementation of the multimodal program; 4) In addition, this trial seeks to identify factors (demographic, physical, and psychosocial) related to treatment success.

Since this trial is designed as a pilot feasibility study, our objectives are to estimate recruitment, retention, and adherence, and to assess the acceptability and feasibility of implementing this multimodal treatment in everyday clinical practice.

3. Protocol Design: This is a randomized parallel controlled pilot study. This study will compare a multimodal, evidence-based intervention consisting of education, relaxation therapy, guided endurance training, and specific exercises and techniques with conventional physiotherapy. Participants will be randomly assigned to two intervention groups: group 1: multimodal physiotherapy intervention (MPI) and group 2: standard physiotherapy.

Outcome measures will be recorded at the start of the study (T0) and at the end of treatment (12 weeks) (T1).

In addition, a nested cohort study will be conducted within this RCT to identify factors associated with treatment success.

Recruitment: Participants will be recruited from the Dr. Brau and Dr. Michel Pain Therapy Centre in Osnabrück, Germany. Potential participants will be informed about the study and invited to participate by a physician or physiotherapist. In addition, the study will be advertised on social media platforms and within university settings. Individuals who express interest will be asked to contact the research team, after which a telephone appointment will be scheduled to provide detailed study information and address any questions.

Randomization: Participants will be randomly assigned to one of two intervention groups using a computer-generated block randomization procedure, stratified by gender (female, male, diverse) and age group (18-30, 31-45, ≥46 years). Randomization will be performed by an independent researcher not otherwise involved in the study.

Sample size: for the effectiveness piece, a total of 60 participants (30 per group) will be recruited. For the primary aim of assessing the feasibility of this study, a sample size of 60 participants is considered appropriate based on guidelines from pilot projects. For the prediction piece, a total of 195 patients will be targeted. The sample size was calculated considering a linear regression model, and the number of potential independent variables according to suggested guidelines. When using multiple linear regression, considering a power of 80%, α=0.05, and ten independent variables included in the model, 155 patients should be included; considering a 20% dropout rate, 195 patients will be targeted.

All patients will be evaluated by a physician, who will make diagnoses and be responsible for referring patients to physiotherapy treatment.

Interventions: The therapeutic strategies described below will be considered as part of the treatment. All participants will have been diagnosed with migraine by a physician at the pain therapy centre or by an external medical doctor. Most patients receiving treatment at the centres are likely taking medication prescribed by their physician. Changes to the medication prescribed by the treating physician will be documented during the study.

The modalities described below will be administered over 18 sessions in line with physiotherapy standards, making the treatment both realistic and practical for implementation in a clinical setting. In accordance with the prescribed physiotherapy guidelines, the Schmerztherapiezentrum Dr. Brau und Dr. Michel can issue three prescriptions with the code 'chronic pain syndrome (CS)' before a six-month break must be taken. The 18 sessions can be carried out as physiotherapy (Krankengymnastik in Germany) or physiotherapy on equipment (Gerätegestützte Krankengymnastik (KGG)). One Krankengymnastik (KG) session comprises 20 minutes of individual therapy and can be passive, active, or educational. One option is to carry them out as a double treatment. This means that two treatments are carried out one after the other. This would then be 40 minutes per appointment, with three treatments instead of six per prescription. KGG comprises 60 minutes with a maximum of three people at a time and involves independent training with the support and observation of physiotherapists.

Group 1: Multimodal intervention: the multimodal physiotherapy treatment will be delivered by trained physiotherapists. The treatment has different components as described below:

1. Education: The subjects will receive a presentation on migraine pain via YouTube video. They are encouraged to watch it before starting physiotherapy. They can then ask questions during therapy. The presentation, spanning approximately 80 minutes, introduces a patient from the physiotherapist's practice to discuss migraine, as well as the associated treatment. The patient's symptoms and treatment journey are used to provide both theoretical and practical insights about their condition. During the presentation, the patient is repeatedly referenced to convey the topic not only theoretically but also practically. This is because migraine is a widespread condition.

2. Aerobic exercise: Patients are encouraged to regularly engage in endurance sports on their own. Endurance sports are a form of training characterised by 'repeated, structured physical activity in which the body's metabolism requires oxygen to produce energy. Large muscle groups are repeatedly exercised, and the heart rate is kept elevated for a longer period of time (between 20 and 60 minutes). Examples include cycling, running, hiking, swimming and walking. Patients will be instructed to use their heart rate and the Borg scale to target the training. High-Intensity Interval Training (HIIT) will be targeted per literature endorsements and positive effects. The training will depend on each subject's own assessment and preferences.

3. Relaxation therapy: In addition, participants are encouraged to practise relaxation techniques daily, as these have been shown to have positive results in some studies. Relaxation therapy appears to reduce the frequency and intensity of migraine attacks, alleviate chronic pain symptoms (including neck pain), decrease the need for medication, and improve psychological parameters such as self-efficacy, anxiety, and depression. This is due to stress reduction and positive effects on the autonomic nervous system. Patients are shown a video on YouTube about the breathing technique 4-7-8, that they are asked to follow.

   The 4-7-8 breathing technique has a relaxing effect, relieves stress, and helps people to calm down. The name already reveals how it works: the patient is asked to breathe in for 4 seconds, hold the breath for 7 seconds, and then breathe out for 8 seconds. Then this is repeated at least 8 times. The exercise can be performed in sitting, standing, or lying down.

4. Specific exercises and physiotherapy techniques: The following plan for specific exercises and techniques of the multimodal treatment is suggested:

   1. Session 1+ 2 (40 minutes): Detailed findings with a physical and objective examination of the structures that induce and influence headaches, the physical examination will be conducted according to the physiotherapist training and reasoning to deliver the manual therapy and exercise programs (e.g., passive physiological intervertebral movements, additional movements of the upper cervical spine, examination of myofascial trigger points in the head and neck area, among others).
   2. Session 3+4 (40 minutes): Repetition and discussion of pain education and the home exercise programme, passive techniques: Manual techniques (especially passive physiological intervertebral movements (PPIVMs) C1-C3), soft tissue techniques of the affected structures (trigger point therapy of the painful muscle), instructions for self-massage of the painful muscles
   3. Session 5+6 (40 minutes): Revision of therapy: Revision of manual techniques, soft tissue techniques, and self-massage, learning how to activate the deep neck flexors and extensors, strengthening the shoulder girdle muscles, answering individual questions
   4. Session 7 (60 minutes with a maximum of 3 participants): At least 30 minutes on an ergometer with a heart rate monitor and training at the appropriate heart rate.
   5. Sessions 8-12 (60 minutes with a maximum of 3 participants): Independent training: Activation of deep neck flexors and extensors, strengthening of the shoulder girdle muscles, exercises for the jaw, stretching and self-massage of hypertonic muscles, exercises for balance and the vestibular system.

Compliance Adherence to the on-site session will be recorded by the therapist. Compliance with the home exercise program will be monitored via Research Electronic Data Capture (REDCap). Participants will be prompted to record their daily activity in a diary, documenting exactly how often and consistently they perform the prescribed exercises.

Co-interventions Participants will be asked to refrain from undergoing new treatments during the study to avoid co-interventions. Any changes to the trial protocol will be recorded in REDCap by the researchers and accounted for in the final analyses.

Patients will be assessed regarding outcomes (see outcomes section) and baseline information. The following baseline information will be collected in addition to the outcomes

1. Patient characteristics (e.g., age, sex, pain characteristics, treatment history).
2. Comorbidities: Subjects will be requested to draw on a body map any additional pain locations, as well as any additional diagnosed conditions.
3. Diagnostic criteria for Temporomandibular disorders: A complete examination of the craniomandibular system using the Diagnostic Criteria for Temporomandibular Disorders (DC/TMD) will be used.

It is important to highlight that the psychometric properties of the outcomes used in this study have been extensively supported in the literature to be used in patients with migraine. These variables will be collected at the baseline and at the end of the treatment period, with the exception of the patients' characteristics and DC/TMD questionnaire and physical assessment, which will be collected only at the baseline. In addition, questionnaires on self-perceived improvement and the therapeutic alliance will be collected only at the end of treatment.

Feasibility outcomes Additionally, this study will focus on the feasibility and clinical applicability of the intervention. Three criteria to determine the success of a trial have been established: 1) recruitment rate: at least 30-40% of eligible subjects are recruited and randomized; 2) retention: drop out during the intervention period cannot exceed 33%; 3) adherence: At least 66% of participants tolerate the intervention and ≥70% adhere to the protocol. Based on these recommendations and the feedback from patients and therapists, the investigators will modify our protocol to maximize feasibility in a subsequent trial. This will be assessed using a questionnaire completed by the patients and physiotherapists in the intervention group. The questionnaire will explore key aspects of implementing the intervention in routine clinical practice.

Analyses Feasibility outcomes (i.e., recruitment rate, retention, adherence) will be descriptively analyzed.

The descriptive parameters, such as the mean and standard deviation, will be used to describe the characteristics of the sample, such as demographic and clinical characteristics, and changes within the groups. Of interest are frequency distributions for categorical variables (e.g., gender, type of treatment, and migraine frequency, and neck pain history among others). Subsequently, inferential statistical tests are used to examine the differences between the intervention and comparison groups.

The analysis of effectiveness will follow the intention-to-treat principles. Efficacy analysis will follow the per-protocol (compliers only; >70% of intervention) and as-treated (intervention received) principles. Due to the pilot design of the study and the small sample size used, statistical analyses will be exploratory, and effect sizes will be calculated to enlighten the clinical significance of these findings. The investigators will use Cohen's d with pooled standard deviation to display the effect sizes. To calculate Cohen's d, the Cambridge effect size calculator Excel tool will be used (https://www.cem.org/effect-size-calculator) complemented with RevMan 5.4 calculations to display forest plots To investigate the objectives of this study, two analyses will be performed. The analyses include 1) within-group and 2) between-group differences.

Paired t-test will be conducted to test the within-group differences, and the Wilcoxon-Rank Test to confirm p-values. For testing differences between two independent groups (between-groups difference), one-way ANOVA with Welch's t-test adjustment will be conducted. Two necessary assumptions for conducting ANOVA are normally distributed data as well as the homogeneity of variances for both groups. ANOVA test and t-tests are very robust to violations of normality when the sample sizes between groups are similar. However, the results of these tests can be affected by violations of homogeneity of variance. According to Rasch et al., pre-testing leads to an unknown risk of type-I and type-II errors, so they recommended the Welch's t-test without pre-testing over other statistical tests. This is supported by Ruxton et al., recommending the use of Welch's t-test over other statistical tests, e.g. the Mann-Whitney-U-Test.

The primary analysis will be conducted using linear mixed-effects models to evaluate the effectiveness of the intervention on migraine-related disability as measured by the Headache Impact Test (HIT-6™). Given the longitudinal structure of the data, with repeated measurements (e.g., baseline and post-intervention) nested within individuals, a multilevel modeling approach will be applied to account for within-subject correlations. Fixed effects will include treatment group (multimodal intervention vs usual care), time, and the group × time interaction, which represents the primary effect of interest. Additional covariates of interest, such as baseline pain intensity and treatment expectancy, will be included as fixed effects to explore their influence on treatment outcomes. A random intercept for participants will be included to account for individual variability. If appropriate, additional random effects (e.g., therapist or treatment site) will be considered. Linear mixed-effects models are robust to missing data under the assumption of missing at random (MAR). If necessary, sensitivity analyses using multiple imputations will be conducted to assess the impact of missing data on the results.

To identify factors related to treatment success of different exercise modalities, the investigators will perform two different analyses: 1) considering headache disability (Headache Impact Test Short Form (HIT-6TM)) (continuous variable) (continuous variable) as an outcome, and 2) using the response to treatment as an outcome using the GRS (dichotomous outcome). Participants will be classified as responders and non-responders according to the 15-point GRS for each treatment group at the end of the treatment. To classify patients as responders and non-responders, a cut-off value of ±4 points of change (for improvement or worsening) will be chosen; this would assure a moderate change. The main time point will be the end of the treatment. Both models will be built based on the following procedures: 1) univariate single linear/logistic regression analysis will be performed to determine the association between the independent variables (possible predictor factors) and the dependent variables of interest (headache disability (Headache Impact Test Short Form (HIT-6TM)), GRS). The independent variables that are significant with a p≤0.2 will be considered in the first multivariate linear/logistic regression model. Models will be constructed with and without the inclusion of baseline headache frequency to explore its contribution. The investigators then proceed to a backward elimination process to select the most significant variables. Variables will be entered into or hierarchically removed from the multivariate model, based on the Akaike Information Criterion. As a second option, the Least Absolute Shrinkage and Selection Operator (LASSO) algorithm will be applied to automatically select the relevant characteristics and regularize the model. The selection of variables for the multivariate model will be based on both the level of significance and the magnitude of the coefficients obtained through LASSO. This type of analysis is used in data with a large number of variables and small sample sizes, thus avoiding overfitting and selecting the best predictive model. Once the ideal model has been selected, a nomogram will be created and validated to provide an intuitive visual tool for predicting treatment response based on the analyses. The receiver-operating characteristic (ROC) curve will be used to assess each variable's ability to discriminate between responders and non-responders, following established guidelines (logistic regression).

Research team Principal investigator: Prof. Dr. Susan Armijo-Olivo

Other scientists who will conduct the research:

Marie Marks (B.Sc. M.Sc. cand.) Dr. Ester Castro-Carletti, PhD MSc. PT Cristian Justribo Manion MSc. PT Sabiha Irem Sancaktar Short-term students from Bachelor's, Master's, and Ph.D.

Inclusion Criteria

• Aged 18-70 years.
• Have a diagnosis of migraine established by a headache specialist according to the International Classification of Headache Disorders, 3rd edition (ICHD-3).
• Have experienced migraine symptoms for more than 3 months, with headaches occurring on at least 2 days per month.
• Provide written informed consent before participation.
• Complete a study-specific self-assessment questionnaire based on the ICHD-3 criteria to confirm the migraine diagnosis and classify participants as having episodic or chronic migraine.

Exclusion Criteria

• Other primary or secondary headache disorders.
• Acute neurological diseases or abnormal neurological examination findings.
• Rheumatic diseases, vascular diseases, cancer, or other serious medical conditions.
• Comorbid chronic pain disorders (e.g., fibromyalgia).
• Diagnosed psychiatric disorders (e.g., major depression or schizophrenia).
• Substance abuse.
• Physiotherapy for headache or neck pain within 3 months before study enrolment.
• Previous surgery involving the jaw, cervical spine, or head.
• Pregnancy.
• Pain not clearly attributable to migraine.
• Inability or unwillingness to provide informed consent.
• There are no restrictions regarding sex, ethnicity, or country of residence.

Participants will not be informed of the study's specific hypotheses, and intervention descriptions will be kept neutral to avoid performance bias. Subjects will be told that both interventions are effective for managing migraine. Outcome assessors (measuring clinician-assessed outcomes) will be blinded to group assignments (to avoid detection bias), and the statistician will be blinded throughout the analysis, in accordance with established guidelines. Furthermore, therapists are assigned exclusively to a single study arm; those delivering the multimodal intervention will not provide standard care, and vice versa (to decrease performance/contamination bias). The allocation sequence will be uploaded to a secure online system, and access will be restricted to the study coordinator responsible for assigning participants to therapists. This procedure ensured allocation concealment, as all other study personnel remained unaware of the randomization sequence and upcoming group assignments.