Introduction

Tension-type headache (TTH) is identified as the most prevalent primary headache, affecting approximately 26% of the general population. Thereby, it significantly impacts daily activities, for which it is listed among the top 10 disabling disorders globally. TTH is categorized into infrequent episodic, frequent episodic, and chronic types based on headache frequency. Chronic daily headaches adversely affect the quality of life. While pharmacological treatment is commonly prescribed, scientific evidence supports the effectiveness of physical therapy, specifically manual therapy, in managing TTH. Current literature suggests combining manual therapy with therapeutic exercise for improved symptom relief. However, there is a lack of consensus on the gold standard treatment. Therapeutic exercise has demonstrated efficacy in reducing headache frequency, duration, and intensity up to 3-6 months post-treatment. Therefore, this study aims to evaluate the effectiveness of 12 weeks of strength training in chronic TTH.

Methods

For this purpose, a parallel superiority randomized controlled trial was conducted including participants from 18-65 years suffering from chronic TTH. Chronic headache was defined as being present for more than 6 months.

The intervention group received a home program with exercises to strengthen the craniocervical, shoulder, and shoulder girdle muscles. These exercises were performed 2x per week in the first 6 weeks and 3x per week in the remaining 6 weeks. Three sets of 8-10 repetitions were performed at a Borg score of 7/10, meaning strong effort was needed to complete the exercises. Every session started with a mobility warm-up where the arms and cervical spine were targeted. Using an elastic band, the following exercises were performed:

• shoulder flexion,
• shoulder abduction,
• horizontal row,
• bird exercise for shoulder,
• horizontal row and
• vertical row not overpassing 60 degrees internal rotation.

Hereafter, the participants performed the following isometric strengthening exercises of the cervical musculature: left and right flexion, extension, and lateral inclination. These movements were resisted by the patient’s hand for 6sec. The control group participants were asked to continue their normal daily activities.

The primary outcomes were the headache characteristics duration (measured in hours per day), pain intensity (VAS scale 0-100), and headache frequency (days per month). Secondary outcome measures were muscle thickness at rest and with resistance, measured by ultrasound. The Craniocervical flexion test was assessed to analyze the endurance of the deep neck flexors. ROM was measured using the CROM device and pain pressure thresholds were registered with the use of an algometer. The locations measured for their pain pressure thresholds were:

• Upper trapezius
• Masseter
• Temporalis
• Anterior tibialis
• The median nerve in the elbow crease

Results

Forty participants were equally randomized to the intervention or control groups. The majority of the included participants were female; 85% in the intervention group and 75% in the control group. They were on average between 34 and 40 years of age and had a normal BMI with on average 24 kg/m2. Both groups were equal at baseline.

The trial showed a reduction of the headache intensity by 1.8 points on the VAS scale and a reduction in duration by 7 days per month at 12 weeks. The authors found that the effect size was large. No difference was found in the headache frequency.

The secondary outcomes revealed that the strength of the deep cervical flexor muscles was improved significantly in the intervention group with a large effect size. The muscle thickness was improved in the right multifidus, bilateral deep flexors, and also with a large effect size. Regarding cervical ROM, lateral flexion was improved in the intervention group. Pain pressure thresholds were improved in the temporalis muscles and trapezius muscles bilaterally, and in the left masseter and left median nerve, also with a large effect size.

Questions and thoughts

The trial measured the time x group interactions, but the between-group differences (except from the resulting p-value and the effect size) were not displayed. Instead, only the within-group differences were noted and tabulated. This is weird since it is a superiority trial and we want to know what the between-group difference in the primary outcomes would be. In one of our previous research reviews, the same issue came forward. However, in that particular study, the primary between-group analysis showed no differences and thus they interpreted the within-group differences to highlight the benefit of the intervention. This was highly misleading. Yet in the study by Martín-Vera you are reading about now, the within-group differences are shown, but the authors state that the findings are based on the between-group differences, and that is as it should be. It is however a pity that we only get the p-value. We don’t have the real difference between the intervention and control group, nor do we have the confidence interval.

Why was the left multifidus not thicker at follow-up despite the increased thickness observed in the right multifidus? Maybe there are differences in muscle atrophy, like was the case in Yun et al., 2019 who found asymmetric atrophy of cervical multifidus muscles in patients with chronic unilateral cervical radiculopathy. Peng et al., in 2022 found a smaller cross-sectional area of the longus colli but not in the multifidus muscles in patients with chronic nonspecific neck pain. As this population of people suffering from chronic TTH, the chronicity may have led to muscle changes. However, this was not examined by the present study.

What was a bit astonishing to me was the fact that this program was completed unsupervised and that despite this, the compliance with the intervention was 87%. It is not specified how this compliance was measured and the trial did not use an exercise diary.

Talk nerdy to me

Apart from the standard assessment, for the first time, bilateral changes in sensory sensitivity in structures resulting from neuropathic changes are examined. The pain pressure threshold improved in the temporalis muscle, trapezius muscle, left masseter, and left median nerve but not in the tibialis muscle. This may mean that this population studied had no central sensitization as the pain pressure thresholds at the remote location (tibialis anterior) at baseline were already high. They had a pain pressure threshold of on average between 6.3 and 6.7 kg/m2. This is quite high, as the mean reference values for pain threshold in females without pain were reported to be 4kg/m2 in the study by Waller et al., 2016. They found on average that hyposensitive people had a pain pressure threshold ranging from 5.2 to 7.9 kg/m2 for the 75th and 95th percentiles respectively.

This program of strength training in chronic TTH gave no information about progressions or regressions. Only the intensity of the exercises was defined to be 7/10 on the Borg scale. I assume that whenever needed, harder resistance bands were dispersed to progressively increase the resistance. This study did not measure whether the resistance exercises led to improvements in shoulder and neck muscle force.

A good point was the establishment of the diagnosis by a neurologist following the International Headache Society (IHS) classification of Headaches. There was no missing data and not a single subject dropped out from the study.

The reduction of the headache intensity by 1.8 points on the VAS scale is just below the normally accepted minimal important difference of 2 points.

Take home messages

This study examined strength training in chronic TTH. It found that a 12-week program focusing on strengthening the muscles in the neck and shoulder area led to positive improvements in pain intensity and duration. Other secondary outcomes were improved as well. Individuals experiencing chronic tension-type headaches are recommended to engage in targeted strength training for the cervical and shoulder muscles as it has been shown to reduce both the intensity and duration of headaches. This study further showed that with a cheap self-management program, important improvements in headache characteristics can be made.

Useful tools

The presence of neuropathic changes can be evaluated with the Douleur Neuropathique-4 Questionnaire (DN4).

Reference

Martín-Vera D, Sánchez-Sierra A, González-de-la-Flor Á, García-Pérez-de-Sevilla G, Domínguez-Balmaseda D, Del-Blanco-Muñiz JÁ. Efficacy of a strength-based exercise program in patients with chronic tension type headache: a randomized controlled trial. Front Neurol. 2023 Sep 18;14:1256303. doi: 10.3389/fneur.2023.1256303. PMID: 37789886; PMCID: PMC10543698. 

Additional references

Yun Y, Lee EJ, Kim Y, Kim JC, Lee SA, Chon J. Asymmetric atrophy of cervical multifidus muscles in patients with chronic unilateral cervical radiculopathy. Medicine (Baltimore). 2019 Aug;98(32):e16041. doi: 10.1097/MD.0000000000016041. PMID: 31393343; PMCID: PMC6709268. 

Peng Q, Zhang Y, Yang S, Meng B, Chen H, Liu X, Zhao W, Hu M, Zhang L, Tao Y. Morphologic Changes of Cervical Musculature in Relation to Chronic Nonspecific Neck Pain: A Systematic Review and Meta-Analysis. World Neurosurg. 2022 Dec;168:79-88. doi: 10.1016/j.wneu.2022.09.057. Epub 2022 Sep 17. PMID: 36126892.

Waller R, Smith AJ, O’Sullivan PB, Slater H, Sterling M, Alexandra McVeigh J, Straker LM. Pressure and cold pain threshold reference values in a large, young adult, pain-free population. Scand J Pain. 2016 Oct;13:114-122. doi: 10.1016/j.sjpain.2016.08.003. Epub 2016 Aug 24. PMID: 28850507.