The review question was formulated using the PICOS (Participants, Intervention, Comparison, Outcome, Study design) framework: “Do physical therapy and rehabilitation approaches improve the intensity of headache, frequency, duration, disability, and QoL in patients with chronic migraine?” (P: Patients with chronic migraine; I: Physical therapy and rehabilitation approaches; C: Comparison group (healthy or placebo); O: Intensity of headache, frequency, duration, disability, QoL; S: This systematic review and meta-analysis include randomized controlled studies that compare the effectiveness of physical therapy and rehabilitation approaches on intensity of headache, frequency, duration, disability, and QoL in chronic migraine).

A comprehensive literature search was conducted in the PubMed and Web of Science databases, covering the period from the inception of the databases to January 29, 2023. No language restrictions were applied. A meta-analysis was subsequently performed using the available quantitative data. The search strategy utilized specific keywords, which are listed in Appendix.

For inclusion in this systematic review and meta-analysis, we considered human randomized controlled trials (RCTs) conducted in adults aged 18 years and older who were diagnosed with chronic migraine, with or without aura. We excluded animal studies, studies involving children and adolescents below 18 years of age, abstracts, case reports or series, letters to the editor, retrospective-prospective cohort studies, case-control studies, review articles, conference proceedings, articles published in non-peer-reviewed journals, theses, dissertations, and studies focusing on headaches other than chronic migraine.

The process of study selection involved two steps. Initially, the articles retrieved from the literature search were screened based on their titles and abstracts, using the predefined inclusion and exclusion criteria. Subsequently, the full texts of the remaining articles were assessed to determine their eligibility for inclusion in the review. Two independent researchers conducted the searching and screening process, ensuring a thorough and unbiased selection of studies.

The methodological quality of the included RCTs was assessed using the Physiotherapy Evidence Database (PEDro) (https://pedro.org.au/) scale. The PEDro scale comprises 10 items, each with a binary response of YES (positive rating) or NO (negative rating). The total score on the scale determines the quality rating, with scores below 4 indicating poor quality, scores of 4–5 indicating fair quality, scores of 6–8 indicating good quality, and scores of 9–11 indicating excellent quality [20]. Two independent researchers (D.O and H.A) evaluated the methodological quality of the studies based on the PEDro scale.

Data extraction was performed to collect relevant information on trial and patient characteristics. This included details such as the country of recruitment, sample size in each arm, mean age of participants, and gender distribution. Additionally, information on the interventions used and outcome measures assessed in the studies were also collected. The primary outcomes of interest were defined as the alterations in intensity, frequency, and duration of headache attacks, as well as the impact on disability and QoL among individuals with chronic migraine. The data extraction process was conducted by the researchers to ensure accuracy and consistency in gathering the necessary information from the included studies.

The outcome measures used in the included trials were evaluated, and a frequentist random-effects meta-analysis was conducted to analyze the quantitative data of the common outcome measures. Before conducting meta-analysis, the mean change in each group was obtained by subtracting the post-intervention mean from the baseline mean. Standard deviations of the changes were available for only one study [21] and we used these values as a reference (SD${}_{\text{E,change}}$). To impute missing standard deviations of the changes, firstly, correlation coefficients in the experimental therapy group (Corr${}_{\text{E}}$) were calculated as

$${\mathrm{Corr}}_E=\frac{S{D}_{E,\text{baseline}}^2+S{D}_{E,\text{after}}^2-S{D}_{E,\text{change}}^2}{2xS{D}_{E,\text{baseline}}xS{D}_{E,\text{after}}}$$

and similarly for the control group (Corr${}_{\text{C}}$). Then, standard deviations of the change from baseline for the experimental therapy and comparator groups were estimated as

(x2)$$\begin{array}{c}\hfill \mathrm{S}{\widehat{\mathrm{D}}}_{\mathrm{E},\text{change}}=\hfill \\ \hfill \sqrt{{\mathrm{SD}}_{\mathrm{E},\text{baseline}}^2+{\mathrm{SD}}_{\mathrm{E},\mathrm{after}}^2-\left(2\mathrm{x}{\mathrm{Corr}}_{\mathrm{E}}{\mathrm{xSD}}_{\mathrm{E},\text{baseline}}{\mathrm{xSD}}_{\mathrm{E},\text{affer}}\right)}\hfill \end{array}$$

$$\begin{array}{c}\hfill \mathrm{S}{\widehat{\mathrm{D}}}_{\mathrm{C},\text{change}}=\hfill \\ \hfill \sqrt{{\mathrm{SD}}_{\mathrm{C},\text{baseline}}^2+{\mathrm{SD}}_{\mathrm{C},\mathrm{after}}^2-\left(2\mathrm{x}{\mathrm{Corr}}_{\mathrm{C}}{\mathrm{xSD}}_{\mathrm{C},\text{baseline}}{\mathrm{xSD}}_{\mathrm{C},\text{affer}}\right)}.\hfill \end{array}$$

The I${}^2$ statistic was utilized to assess the level of heterogeneity in the studies. Heterogeneity was classified as follows: 0% to 40%, not important; 30% to 60%, moderate; and 50% to 90%, substantial, with 75% to 100% indicating considerable heterogeneity [22]. The data synthesis and analysis allowed for a comprehensive evaluation of the collective findings from the included studies.

The included trials investigated various physical therapy and rehabilitation approaches for rehabilitation in chronic migraine patients. The specific interventions assessed in these studies were as follows: aerobic exercise [26], osteopathic manipulative treatment (OMT) [25], occipital transcutaneous electrical stimulation (OTES) [27], acupressure [23], hydrotherapy [21], instrument-assisted soft tissue mobilization (IASTM) [24], facial proprioceptive neuromuscular facilitation (FPNF), and connective tissue massage (CTM) [28]. The sample sizes in the different arms of the studies ranged from 15 to 49 patients, with a mean age ranging between 31 and 41.7 years. However, one study did not provide data on the age of the participants [21]. Most patients included in the studies were female, although two studies did not report data on the gender composition of their participants [21, 28]. A detailed summary of the study characteristics, including the interventions used, sample sizes, mean age, and sex distribution, can be seen in Table 1 (Ref. [21, 23, 24, 25, 26, 27, 28]).

These studies assessed different outcomes and utilized various scales to evaluate the efficacy of physical therapy and rehabilitation approaches. The intensity of the headache was measured using different scales across the studies. One study used a rating scale with categories of “1-mild, 2-moderate, 3-disabling” [26], while others used the Visual Analog Scale (VAS) [21, 27, 28], Numeric Rating Scale (NRS) [23], and Neck Pain Scale (NPS), which is essentially a VAS evaluating both headache and neck pain [24]. The duration of headache was evaluated in only one study, where patients were asked to select among options of 6, 12, 18, or 24 hours of duration [26]. The frequency of headache was assessed in four studies, but the definition varied among the studies. This was measured as headache days per month in the studies by Santiago et al. [26] and Cerritelli et al. [25], migraine attacks in the study by Xu et al. [23], and through a daily headache diary in the study by Sujan et al. [21].

Disability related to headache was evaluated in two studies, both of which utilized the Migraine Disability Assessment Scale (MIDAS) [25, 27]. QoL was assessed in four studies, with two studies using the Short Form 36 (SF-36) questionnaire [23, 24], and three studies employing the Headache Impact Test-6 (HIT-6) [21, 24, 25]. In the study by Santiago et al. [26], the efficacy of amitriptyline alone was compared with the combination of amitriptyline and aerobic exercise over a 12-week period. The findings showed significant reductions in headache frequency (days per month) (p = 0.001), duration of headache attacks (p = 0.017 for 6 hours, p = 0.001 for 12 hours, and p = 0.001 for 24 hours), and intensity of headache (p = 0.048). These results support the benefit of adding aerobic exercise to pharmacological therapy in the management of chronic migraine [26]. Cerritelli et al. [25] examined the efficacy of OMT as a modality of physical therapy in chronic migraine. The study aimed to evaluate the benefit of combining OMT with medical therapy compared with medical therapy alone over a 24-week period. The study design consisted of three treatment arms: OMT and medication therapy, sham manipulative therapy and medication therapy, and medication therapy alone. The results of the study indicated that the combination of OMT and medication therapy had a positive impact on QoL. There was a statistically significant decrease in HIT-6 scores (a measure of headache impact) in the OMT and medication therapy arm compared with both the sham manipulative therapy and medication therapy arm, as well as the medication therapy alone arm (p 0.05 for both comparisons). Additionally, the combined therapy with OMT led to significant reductions in disability scores, intensity of pain, and migraine days per month (p 0.05 for all parameters). It is worth noting that patients who received sham manipulative therapy along with medication therapy did not experience better outcomes than those who received medication therapy alone. These findings suggest that OMT when combined with medication therapy can provide additional benefits in terms of improving QoL, reducing disability, and alleviating pain and migraine frequency in chronic migraine patients [25].

In another study conducted by Sujan et al. [21], the effectiveness of hydrotherapy combined with medical therapy was investigated in chronic migraine patients over a period of 45 days. The results demonstrated significant decreases in HIT-6 scores for frequency of headache and intensity of headache when compared with medical therapy alone (p 0.001). These findings suggest that hydrotherapy, when used in conjunction with medical therapy, can lead to improvements in intensity, frequency of headache, and overall QoL [21].

In contrast, the study by Xu et al. [23] examined the impact of acupressure in addition to sodium valproate on chronic migraine. The treatment duration was 8 weeks, and the outcomes assessed included intensity of headache, frequency of migraine attacks, and QoL measured using SF-36. However, the results did not show a reduction in the intensity of headache, frequency of migraine attacks, or improvements in QoL compared with the use of sodium valproate alone. These findings suggest that acupressure as an adjunct to sodium valproate may not provide additional benefits in terms of intensity of headache reduction and overall QoL improvement [23].

In a recent study conducted by Torlak et al. [24], approaches involving IASTM and OTES were evaluated to test their effectiveness in chronic migraine patients. The treatment duration was 5 weeks. The results showed a significant decrease in the intensity of headache and neck pain in both the IASTM arm (p = 0.001, effect size = 0.78) and the OTES arm (p = 0.001, effect size = 0.82). Furthermore, the QoL was improved in both the IASTM arm (p = 0.01, effect size = 0.78) and the OTES arm (p = 0.01, effect size = 0.78) compared with the control arm, which did not receive any approach. However, there were no significant differences observed between the two interventions in terms of efficacy [24]. Bono et al. [27] also evaluated the effectiveness of OTES in improving chronic migraine and chronic tension-type headache. The intervention was administered for only 2 weeks and the study did not find statistically significant differences in disability scores and intensity of pain between real and sham stimulations [27].

Kurt et al. [28] evaluated the effectiveness of two interventions, FPNF and CTM, in patients with chronic migraine over a period of 6 weeks. Both interventions resulted in significant reductions in the intensity of headache (p = 0.001 for both), indicating their potential efficacy in managing chronic migraine. Importantly, no significant difference in effectiveness was observed between the FPNF and CTM interventions [28]. These findings suggest that both FPNF and CTM can be beneficial in reducing the intensity of headache in chronic migraine. Authors suggested that either intervention can be considered as a viable treatment option to manage chronic migraine [28].

Adverse events (AEs) associated with the physical therapy and rehabilitation approaches were reported in only one study [23]. Patients who underwent combined therapy with acupressure and sodium valproate experienced AEs such as nausea, diarrhea, hair loss, neck pain, flu-like syndrome, and pruritus. In contrast, patients who received only sodium valproate reported vomiting, worsening migraine, and back pain [23]. These findings highlight the importance of monitoring and managing potential AEs when implementing physical therapy and rehabilitation interventions in the management of chronic migraine. Healthcare providers should carefully evaluate the risks and benefits of such interventions and take appropriate measures to mitigate any potential side effects.

Data synthesis and analysis were performed on five out of the seven trials included in the systematic review. A total of 183 patients from the studies reporting changes in the intensity of headache [21, 23, 24] and 195 patients from the studies reporting changes in headache days [21, 25, 26] were included in the quantitative analysis. However, the data from the study that evaluated OTES [27] were not included in the analysis, along with the data from the studies that utilized manual physical therapies [21, 23, 24, 25]. Heterogeneity in terms of intensity of headache and headache days per month was assessed using the I${}^2$ statistics, which indicated the presence of heterogeneity and justified the use of a random-effects model for data analysis. This approach considers the variation across studies and provides a more conservative estimation of the treatment effects.

Three trials provided data on the improvement of the intensity of headache, assessed using a VAS or NRS with scores ranging from 0 to 10. Among these RCTs, the methodological quality of the studies conducted by Xu et al. [23] and Torlak et al. [24] was graded as excellent with scores of 10 and 9, respectively (Table 2, Ref. [21, 23, 24, 25, 26, 27, 28]). The study by Sujan et al. [21] was deemed to have good methodological quality with a score of 6 (Table 2). All three studies included in the meta-analysis had good to excellent methodological quality.

The random-effects model analysis showed that the standardized mean difference (SMD) of the intensity of headache was higher in patients who received manual physical therapy compared with those who did not receive it, although the difference did not reach statistical significance (SMD: 2.14, 95% confidence interval (CI): –0.25, 4.54; I${}^2$ = 96%, p = 0.0796). The substantial heterogeneity observed (I${}^2$ = 96%) suggests significant variability across the studies. Fig. 2 presents a graphical representation of the meta-analysis results for intensity of headache.

Fig. 2.

Forest plot analysis of changes from baseline in the intensity of headache. SMD, standardized mean difference; CI, confidence interval.

Three trials examined headache frequency as the number of headache days per month. The study by Cerritelli et al. [25] was rated as excellent in terms of methodological quality (score: 9), while the studies by Santiago et al. [26] and Sujan et al. [21] were considered to have good methodological quality (scores: 6 for both) (Table 2). All three studies included in the meta-analysis of headache days had good to excellent methodological quality. The random-effects model analysis revealed a significantly greater SMD in patients who received manual physical therapy as an add-on to medical therapy, compared with those who received medical therapy alone (SMD: 3.01, 95% CI: 0.03, 5.98; I${}^2$ = 96%, p = 0.047). The substantial heterogeneity observed (I${}^2$ = 96%) indicates significant variability across the studies. Fig. 3 shows a visual representation of the meta-analysis results for headache days per month.

Fig. 3.

Forest plot analysis of changes from baseline in headache days per month.

In the study by Santiago et al. [26], the methodological quality score was good (score: 6). However, this study was not included in the quantitative analysis because the outcome measure (intensity of headache) was evaluated using a different method compared with the studies included in the meta-quantitative analysis. The study conducted by Kurt et al. [28] also had a good methodological quality score (score: 8). This study was not included in the analysis because both groups in the study received active manual physical therapy approaches, and there was no control group. The study by Bono et al. [27] had an excellent methodological quality score (score: 11). However, data from the OTES were not included in the analysis of manual therapy data. Although these studies were not included in the quantitative analysis, their methodological quality was assessed, and the reasons for exclusion are provided. The results of the methodological quality assessments are shown in Table 2.

Limited number of studies: The systematic review identified only seven RCTs that met the inclusion criteria, indicating a limited number of studies available on the topic. This could potentially limit the generalizability and robustness of the findings.

Heterogeneity among studies: The included studies varied in terms of study design, sample size, interventions, treatment duration, and outcome measures. This heterogeneity may introduce variability in the results and make it challenging to draw definitive conclusions.

Small sample sizes: Three out of seven trials randomized 20 or fewer patients in intervention arms. The largest sample included 54 patients in the intervention arm.

Lack of long-term follow-up: The duration of the interventions and follow-up periods in the included studies were relatively short, ranging from 2 weeks to 24 weeks. This short-term follow-up may not capture the long-term effectiveness and sustainability of physical therapy and rehabilitation approaches for chronic migraine.

Potential publication bias: The systematic review searched for studies in only two databases, which might have introduced publication bias. Relevant studies published in other languages or unpublished studies might have been missed, potentially leading to an incomplete representation of the evidence.

Methodological quality of included studies: The methodological quality of the included RCTs was assessed using the PEDro scale. While this scale is commonly used, it is important to note that it has some limitations and may not capture all aspects of study quality. Therefore, the overall quality of the included studies should be interpreted with caution.

Multidisciplinary approach: Our review highlights the need for a multidisciplinary approach in the management of chronic migraine, and indicates that healthcare professionals from different disciplines, such as neurology, physical therapy, and rehabilitation, should collaborate to provide comprehensive care for individuals with chronic migraine.

Non-drug interventions as adjunct treatments: Our findings emphasize the potential benefits of non-drug interventions, such as physical therapy and rehabilitation approaches, as adjunct treatments for chronic migraine when drug therapy is insufficient or associated with intolerable side effects. A combination of pharmacological and non-pharmacological interventions may offer a more comprehensive approach to managing chronic migraine.

Individualized treatment plans: Our study highlights the importance of tailoring treatment plans to individual patients, considering factors such as intensity, frequency, duration of headache, disability, and QoL. Healthcare providers should assess and consider the specific needs and characteristics of each patient when determining the most appropriate physical therapy and rehabilitation interventions.

Adverse event monitoring: We have discussed AEs associated with the physical therapy and rehabilitation interventions, indicating the importance of monitoring, and managing potential side effects. Healthcare providers should closely monitor patients undergoing these interventions and take appropriate measures to minimize AEs.

Further research: The limitations identified in our systematic review highlight the need for further research in this field. Future studies should aim to address the identified limitations, including conducting larger and more rigorous RCTs with longer follow-up periods, considering diverse populations, and exploring the comparative effectiveness of different physical therapy and rehabilitation approaches for chronic migraine.

Furthermore, it is worth noting that the methodological quality of the studies included in the analysis was generally good to excellent. This indicates that the results concerning the intensity of headache and frequency of headache days were supported by articles of high methodological quality.