The Effectiveness of Acupuncture on Headache Intensity and Frequency in Patients With Tension-Type Headache: A Systematic Review and Meta-Analysis

Introduction Headache disorders are one of the most common health problems. Tension-type headache (TTH) is the most prevalent type of primary headache in adults. Several conservative treatments have been used for the management of TTH, such as analgesics, acupuncture, manual therapy, spinal mobilization. This study aims to examine the effectiveness of acupuncture in patients with TTH. Methods and materials PubMed, PEDro database, Cochrane Library, and Google Scholar were searched from January 2000 until February 2021, as well as the reference lists from identified articles. Studies of various acupuncture types were included, but only randomized controlled trials and clinical trials were selected. The studies were screened using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) question. Details about the type of acupuncture, sample size, outcome measures, results, and statistical significance were extracted from the selected studies. A short-term (after the last treatment) and long-term meta-analysis for pain intensity and frequency of headaches was conducted. The I2 index, as well as the x2 test, were used to determine the heterogeneity between studies. A random-effects meta-analysis was carried out. Results From all the studies found in the mentioned databases, only 15 studies with 1272 participants met the criteria. In the meta-analysis, four studies with 557 participants were included. The headaches' frequency after the last treatment was not significantly lower in the acupuncture group than in the placebo/sham group (mean difference: -1.53 [CI: -4.73, 1.67]). However, acupuncture seems to improve the frequency of headaches in the long term, although the results were not statistically significant p=0.06. Furthermore, there was a reduction of 1.55 days per month of headaches in the acupuncture group versus placebo (mean difference: -1.55 [CI: -3.19, 0.09]), but it was not statistically significant. The visual analog scale (VAS) score of the acupuncture group slightly reduced (-0.29) compared with the control group after the last treatment (mean difference: -0.29 [CI: -1.21, 0.62]), although the two groups were not statistically significant p=0.53. In the long term, acupuncture demonstrated a statistical (p=0.009) and clinical benefit compared with placebo/sham. Statistical analyses between the two groups showed a reduction of 0.41 in the VAS scale at the acupuncture group (mean difference: -0.41 [CI: -0.72, -0.10]). Conclusion Overall, after the meta-analysis of articles with high methodological quality, acupuncture's effectiveness compared to sham seems to be statistically non-significant on headache intensity and frequency in patients with TTH after the treatment. Both headache intensity and frequency were reduced in the long term, although only in the pain intensity, the results were statistically significant. Therefore, more studies on this topic should be conducted to examine its effectiveness in headache frequency and intensity.


Introduction
At present, headache disorders (HD) are one of the most common health problems, causing substantial pain and disability in the head, while the pain can also be spread and radiated to the neck [1,2]. Headaches are classified as primary or secondary, based on the clinical presentation. Primary headaches result from dysfunction or overactivity of the heads' structures, without a clear underlying cause. On the other hand, secondary HD are conditions such as neck trauma, vascular disorders, musculoskeletal problems, and arthritic changes [1]. Tension-type headache (TTH) is the most prevalent type of primary HD in adults, associated with a high economic burden for society, while patients' moods and general life activity are affected [1].
According to previous research, TTH pathogenesis remains unclear, although the etiological factors are possibly related to peripheral mechanisms (myofascial nociception) and central mechanisms (sensitization and inadequate endogenous pain control) [3]. The development of TTH is probably associated with myofascial trigger points (TrPs). The TrPs in the head and neck muscles are painful on compression, and they are related to refer to pain -tenderness, motor dysfunction, and autonomic phenomena [4]. Increased muscle stiffness in the trapezius muscle and rectus capitis posterior minor (suboccipital muscles) in TTH patients has been demonstrated. The muscle stiffness does not differ between headache outbreaks and nonheadache days [4].
Several conservative treatments have been proposed for the management of HD. The most common treatments include medication, physical therapy, and relaxation/cognitive therapy [1,5]. Various physical therapy interventions have been suggested to treat headaches. For example, manual therapy, massage, spinal mobilization or manipulation, soft tissue release, and acupuncture are widely utilized [1]. However, studies revealed that alternative medicine therapies, such as acupuncture, are the most used treatments requested by patients with headaches [5].
Acupuncture is defined as the needling of specific points of the body. It was initially developed as a part of Chinese medicine, wherein the purpose of treatment was to release the flow of the body's vital energy by stimulating points along energy pathways. However, acupuncture practitioners have abolished these concepts and approach medical acupuncture concerning neurophysiology [2].
There is a widely held belief, without proof, that using acupuncture as an intervention to treat TrPs in the muscles of the head and neck may reduce the frequency and intensity of headaches [1,3,6]. Notwithstanding the growing use of acupuncture in treating headaches, its clinical effectiveness is still debatable due to a lack of proper reviews [1]. So, it stands to reason that this meta-analysis aims to examine acupuncture's effectiveness in TTH patients. The aim of this study is to examine the effectiveness of acupuncture in comparison with sham or in a combination with interventions on headache frequency and pain intensity in patients with tension-type headaches.

Materials And Methods
This study was conducted following the preferred reporting items for systematic reviews and meta-analyses guidelines (PRISMA). The great majority of the studies found during our initial search focused on dry needling rather than acupuncture. Linde et al. [7] conducted a detailed study on this subject, but they included only randomized controlled trials (RCTs). Therefore this systematic review and meta-analysis aimed to investigate both RCTs and non-randomized clinical trials as well as add the newest data regarding TTH and its management with the use of acupuncture.

Search strategy
The search was conducted through four electronic databases: MEDLINE via PubMed, CENTRAL via Cochrane Library, PEDro, and Google Scholar. Furthermore, the reference list of the articles found was examined to discover more studies on the same topic. Only articles between January 2000 and February 2021 were included. The fact that research on this subject has never been conducted before was the purpose that we choose to examine this 21 years. Also, both RCTs and clinical trials were included because of the lack of bibliography in articles about the management of tension-type headache using acupuncture. The search strategy was based on an algorithm made with the use of MeSH terms (Medical Subject Headings): "acupuncture" OR "acupuncture therapy" AND "tension-type headache."

Study selection
The Population Intervention Comparison Outcome question (PICO) was used to select studies ( Table 1). Three authors carried out the primary screening of the articles found in the four electronic databases and selected articles based on the title and abstract. Each one of them searched all databases independently, and then the outcomes were compared. In the case of different outcomes and disagreement between them, the issues were resolved with the two other authors' intervention. The full text of screened articles was studied if one of the treatment arms of RCTs or clinical trials was about tension-type headaches and the efficacy of acupuncture as a treatment. Although RCTs are considered for ranking second after systematic reviewswhich include only RCTs regarding internal validity -and their design is more suitable for research, in this study we included RCTs and clinical trials because of the limited bibliography found [8].

Exclusion criteria
Specific exclusion criteria were established from the beginning of the study. Trials published before 2000 were excluded, and those written in different languages other than English or Greek. Also, articles that could not be found in the full text were not included in the research.

Evaluation of methodological quality
The assessment of the methodological quality of included articles was performed by two authors independently and was based on the criteria of the Physiotherapy Evidence Database (PEDro) scale. The PEDro scale has 11 criteria, 10 of which evaluate the study's internal validity (criteria 2-9) and statistical information sufficiency (criteria 10, 11) [9]. Methodological quality was categorized as «low» for articles with a score of less than 3/10, «moderate» for articles with a score of 4-6/10, and «high» for articles with a score of 7-10/10. Differences between authors regarding the quality appraisal and extracted data were resolved through discussion with the other authors and team members.

Data extraction
Data were extracted from the selected studies by three authors regarding interventions, sample size, duration, follow-up, outcome measures, outcomes, and statistical significance independently. Especially, the primary outcomes that were extracted were the headache intensity evaluated with the VAS score and frequency (days/month).

Assessment of risk of bias
To evaluate the risk of bias, the Cochrane reviews approach was used to determine the accuracy of the included studies' results. The assessment followed the criteria of the Grading of Recommendations Assessment, Development and Evaluation (GRADE) [10]: allocation concealment, adequate sequence generation (selection bias), blinding of participants and personnel (performance bias), incomplete outcome data addressed (attrition bias), free of selective reporting and free of other bias. Two of the authors graded the included articles, and if disagreement occurred, a fourth author would resolve the issue.

Meta-analysis
A meta-analysis was conducted to determine the effectiveness of acupuncture as monotherapy versus sham/placebo in patients with tension-type headaches. The frequency of headaches (days per month) and the pain intensity (VAS score 0-10) were the primary outcomes. For the trials that a meta-analysis was not possible (not relevant for the meta-analysis topic, not sufficient data), the results were presented as a narrative synthesis. Statistical analysis for the pain intensity and frequency of headaches were performed using the Review Manager 5.4 software (Cochrane Collaboration, Copenhagen). The odds ratio (OR) was calculated for continuous data, and the level of statistical significance was set at p<0.05. Analyses for the clinical and RCTs were conducted using a random-effects model of the Mantel-Haenszel method. The overall effects and the heterogeneity of continuous data across trials were examined by calculating Z values and χ 2 distributed Cochrane Q values, respectively. The I 2 -test of heterogeneity was also conducted to obtain I 2 values based on the formula (I 2 = 100%⁎ (Q − degree of freedom)/Q)) (Higgins et al. [11]). We considered I 2 less than 25% to indicate low heterogeneity and I 2 greater than 75% to indicate considerable heterogeneity.

Literature review results
The systematic review and meta-analysis process is depicted in the flow diagram displayed in (Figure 1). From the initial database search, 2134 articles were collected from four electronic databases (PubMed: 25; PEDro: 27; Cochrane Library: 71; Google scholar: 2010;). Studies were reduced to 15 following the removal of duplicates and applying the established exclusion criteria after their assessment. Based on the inclusion, exclusion criteria, and the design exhibited for the present systematic review, 15 articles (RCTs and clinical trials) were included for qualitative synthesis, and 1267 patients were examined. Subsequently, RCTs and clinical trials were identified and excluded with reasons for the scope of meta-analysis. The information of the selected articles displayed in Table 2. Ultimately, of the remaining 15 articles, 11 did not meet the inclusion criteria for the meta-analysis. A meta-analysis for pain intensity and frequency of headaches was conducted after the end of treatment and after the last follow-up (long term). For the meta-analysis of headache frequency, as well as for the meta-analysis of headache intensity, we used four studies. Although, in the second meta-analysis for the pain intensity after the last treatment, only three studies were eligible due to the absence of data.

Methodological quality of selected articles
Specifically, 15 studies were chosen based on our inclusion criteria (

Meta-analysis
Overall, there was "high" quality evidence of these four trials, including in total 557 participants (intervention and control group) (Karst et al. [12], White et al. [13], Karst et al. [14], Endres et al. [15]). Table  4 displays the main characteristics of these studies. It is well mentioned that the average number of sessions per patient was nine, and the average duration of treatment was 5.5 weeks. The main acupuncture points were selected were GB 20, LI 4, LR3, and the optional points in the head, neck, shoulders were chosen by practitioners according to the patient's symptoms and tenderness. The small number and oldness of studies about this topic show the necessity to conduct more new and well-organized clinical and randomized controlled trials.

4: Characteristics for the studies included in the meta-analysis
The risk of bias of the four studies included in the meta-analysis was estimated via the criteria of Grading of Recommendations Assessment, Development and Evaluation (GRADE) [10]. It was determined as low risk (Figure 2), so the true effect would probably be close to the estimated effect. For the allocation concealment, only the study of Karst et al. [14] was not in agreement with, while for the adequate sequence generation (selection bias), an earlier trial of the same author was not in compliance [12]. None of them blinded both participants and personnel (performance bias) because the techniques were hands-on, and it will be difficult for the physical therapists who provided the treatment not to distinguish the interventions performed in each group. The study of Karst et al. [12] was the only one with incomplete outcome data addressed (attrition bias). Selective reporting was not easily identified, so all of the studies were unclear about this specific criterion. Although, all of them were found free of other bias.

Frequency of headaches
In the meta-analysis for the frequency of headaches, all four studies were eligible. The number of days per month with a headache was evaluated after the last intervention and after the last follow-up (long term). After the last intervention, the pooled estimate for the acupuncture versus placebo/sham was not statistically significant p=0.35. This was due to the small number of participants and the small number of studies conducted. Statistical analysis suggests that headaches' frequency was not significantly lower in the acupuncture group than the placebo/sham group (mean difference -1.53 [CI: -4.73, 1.67]) ( Figure 3). The heterogeneity was not high (I 2 =61%, p=0.05). Studies of Georgoudis et al. [26] support that acupuncture has beneficial effects not as monotherapy but as a part of combination therapy. Moreover, dry needling is a very effective treatment for patients with tension-type headache (Gildir et al. [20], Karakurum et al. [19]), but acupuncture, as proven in this meta-analysis, has no significant results.

CI, confidence interval; df, degrees of freedom
In the long term, acupuncture seems to improve the frequency of headaches in patients with TTH, although the results were not statistically significant p=0.06. The mean last follow-up at these four studies was two months. There was a reduction of 1.55 days per month of headaches in the acupuncture group versus placebo (mean difference: -1.55 [CI: -3.19, 0.09]) ( Figure 4). The heterogeneity was low (I 2 =11%, P=0.34).

FIGURE 4: Frequency of headaches after the last follow up
CI, confidence interval; df, degrees of freedom

Pain intensity (VAS)
In the meta-analysis for the pain intensity after the last intervention, only three of the four trials were eligible. The study of Endres et al. [15] did not have sufficient data for the VAS score. The score was evaluated on a scale between 0 to 10 after the last intervention, both in the experimental and control group. Statistical analysis between the two groups suggested that the VAS score of the acupuncture group slightly reduced (-0.29) compared with the control group after the last treatment (mean difference: -0.29 [CI: -1.21, 0.62]). However, the pooled estimate between the two groups was not statistically significant p=0.53 because of the small number of studies available. The heterogeneity was not high (55%, p=0.11) ( Figure 5).  Figure 6).The heterogeneity was zero (I 2 =0%, p=0.65). As a result, acupuncture seems to be an effective method for the long-term reduction of the intensity of headache pain in patients with TTH.

FIGURE 6: Pain intensity after the last follow-up
Note: CI, confidence interval; df, degrees of freedom

Discussion
Tension-type headaches (TTH) are the most common type of primary headaches in adults and have a significant impact on people's quality of life [1]. Though the underlying causes behind the formation of this pathology are still debatable, many theories have been formed to examine and explain TTH. Various methods have been used to treat patients with TTH, but it seems that acupuncture has become a popular intervention among them. In the 15 included studies in this review, many different acupuncture methods and techniques were described, such as acupuncture in general, sham acupuncture, laser acupuncture, electroacupuncture, and dry needling. Though only five of them examined the comparison between acupuncture and sham, and four of these were suitable for use in the meta-analysis. This research's outcomes were found to agree with the results of a previous study on the same subject conducted by Linde et al. [7] and a systematic review of systematic reviews by Huang et al. [4].
Overall, compared to sham acupuncture, acupuncture was found to have no statistically significant shortterm results on the headache intensity (p=0.53) and frequency (p=0.35) in patients with TTH. An important reduction was discovered in between-group comparison after the last follow-up (1.55 days/month), but it was not significant (p=0.06). All five of the trials comparing acupuncture to placebo was characterized of high methodological quality by PEDro (Karst et al. [12], White et al. [13], Karst et al. [14], Endres et al. [15], Söderberg et al. [17]) but only four of them were used in meta-analysis (Karst et al. [12], White et al. [13], Karst et al. [14], Endres et al. [15]). The small number of the included articles in the meta-analysis and their small sample sizes could imply that heterogeneity could be high, but this is not the case as heterogeneity was not found important. Although pain intensity and headache frequency seem to decrease after the use of needles after the treatment, the results were not statistically significant. Meanwhile, after the last follow-up, both values were found to lessen, though statistically, only the pain intensity was important.
Furthermore, electroacupuncture collated to placebo was studied in three trials of high methodological quality (Xue et al. [22], Wang et al. [24], Chassot et al. [23]). Although they were found to have similar results on TTH, electroacupuncture showed a statistically significant effect on the reduction of pain intensity and headache frequency in patients with TTH except for the trial of Wang et al. [24]. Another type of acupuncture performed in one of the included researches was laser acupuncture, and the study was categorized as a moderate methodological value (Ebneshahidi et al. [25]). This intervention, when compared to placebo laser acupuncture, was found to have an important statistical effect on TTH (p<0.001).
Four trials were included in the systematic review that compared acupuncture to various methods and techniques. Two of them that the same author conducted were categorized of high methodological quality and tested the effectiveness of acupuncture in contrast with relaxation training or physical training (Söderberg et al. 2006 [16], Söderberg et al. 2011 [17]). None of the interventions showed a statistically significant effect on TTH through relaxation training, which seems to have a higher impact on reducing headache frequency and intensity after the end of treatment compared with the other interventions. Also, the physical training group had a more significant effect at the three-month follow-up than acupuncture or relaxation training. A recent study conducted by Schiller et al. compared acupuncture to usual care or medical training and combined with medical training and was found of high methodological quality [18]. When performed additional to acupuncture and collated usual care, the results showed that medical training had a statistically significant effect (p=0.012) while both of the other treatment groups (acupuncture, medical training) differentiate significantly from usual care. Acupuncture in conjunction with stretching and physical therapy techniques (microwave diathermy and myofascial release) was found to improve further (p<0.001) TTH when compared to acupuncture and stretching without physical therapy techniques while in both groups, a statistical significance (p<0.005) was noted (Georgoudis et al. [26]).
Lastly, dry needling that was performed in three studies which were categorized of moderate methodological quality by PEDro seems to have a statistically significant effect (p<0.005) in headache frequency and intensity compared to sham dry needling (Karakurum et al. [19], Gildir et al. [20], Kamali et al. [21]). This view is in agreement with a previous systematic review conducted by France et al. [27]. In conclusion, acupuncture seems to have limited efficacy as monotherapy for reducing days of headache in patients with tension-type headache. Although, after long-term follow-up, acupuncture seems to have a statistically significant effect on pain intensity in patients with TTH. Acupuncture was found to be more efficient when combined with other treatments. Furthermore, dry needling has beneficial effects in the treatment of tension-type headaches and electroacupuncture seems to have a controversial effect on tension-type headache. Nevertheless, more studies on this topic should be conducted in order to examine thoroughly the effectiveness of acupuncture on TTH.

Limitations
This systematic review and meta-analysis certainly has some limitations. Firstly, the selection of studies was conducted in four databases with a limitation of language (English, Greek), so it is possible that some articles with the same subject could be left out. A significant number of trials were excluded because they were written in Korean, Chinese, or other Asian languages. Also, the assessment of methodological quality and evaluating the potential risk of bias of included articles was subjective as each author individually extracted the needed details. Furthermore, the number of the relevant studies found was small, and the methodological and statistical quality of a great majority was low, so consequently, the appropriate trials with complete data included in the meta-analysis were few. Another factor that should be considered is the small population of participants in most of the included studies. All of the above could respectively affect the accuracy and be the reason for the variation of the meta-analysis outcomes.

Conclusions
Overall, after the meta-analysis of articles with high methodological quality, acupuncture's effectiveness compared to sham seems to be statistically non-significant on headache intensity and frequency in patients with TTH after the treatment. Both headache intensity and frequency were reduced in the long term, although only in the pain intensity, and the results were statistically significant. Therefore, more studies on this topic should be conducted to examine its effectiveness in headache frequency and intensity.

Additional Information Disclosures
Human subjects: All authors have confirmed that this study did not involve human participants or tissue. Animal subjects: All authors have confirmed that this study did not involve animal subjects or tissue.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.