Introduction

Cervicogenic headache is a well-known secondary headache originating from cervical spine dysfunction. Due to the musculoskeletal dysfunctions underlying the pathology, physiotherapy, exercise, and manual therapy are prescribed to improve strength, endurance, and mobility deficits contributing to the onset of headaches. Both exercise and manual therapy were previously found effective for the treatment of cervicogenic headache. Exercise is generally regarded as beneficial for more sustained outcomes in the longer term, while manual therapy is mainly thought to be useful for short-term pain reduction. However the long-term benefits of manual therapy are unclear from the literature, many people tend to prefer a manual treatment approach. Mulligan manual therapy is a form of manual therapy where pain-free sustained mobilization forces are applied to the joints of the cervical spine. Although all of the named options have shown beneficial effects in the treatment of cervicogenic headache, no one has yet made the comparison between these different treatment options. Until this study came along!

Methods

This study was designed as a three-armed, parallel-group, randomized controlled trial. Individuals between 18 and 60 years old were eligible for inclusion and diagnosed with cervicogenic headaches based on the International Classification of Headache Disorders (ICHD-3) criteria. This included the following:

Other inclusion criteria were the following:

• Headache intensity > 6/10 on a visual analogue scale at enrollment
• A history of headaches of more than 1 year with a mean frequency of at least 1 headache per week
• Hypomobility of the upper cervical spine (C0-C3) on manual examination
• Reproduction of the headache with palpation of the upper cervical spine (C0-C3)

The exclusion criteria were: History of cervical spine instability, cervical arterial insufficiency, vertigo, dizziness, rheumatoid arthritis, cervical fractures, pregnancy, cognitive impairment, or other contraindications to manual therapy.

Included participants were randomly assigned to one of three groups:

1. Mulligan manual therapy (MMT) plus exercise,
2. sham MMT plus exercise,
3. Exercise alone.

Exercises

The exercises were the same for all 3 groups. The exercises were performed over approximately 20 minutes per session and included:

• Cervical flexion loading exercises as described by the paper of O’Leary et al. (2007). In brief, the study by O’Leary compared more specific craniocervical flexion exercises with a Stabilizer for pressure biofeedback (CCFEx) to a more general controlled head lift exercise with the upper cervical spine in neutral in the supine position (CFEx). See the picture below. The CCFEx progressed using advancing levels of pressure on the stabilizer, while the CFEx progressed using position and additional loads of 0.5kg increments on the forehead. They found that both exercises achieved similar results regarding isometric craniocervical flexor muscle performance. I assume the current paper we review used the loading CFEx-exercises using weights.
• Prone scapular retractions
• Passive self-stretching exercises
• Active mobility exercises of the neck

MMT plus exercises

The MMT plus exercise group participated in the same 20-minute exercise program and received 10 minutes of MMT techniques pragmatically chosen. Four different techniques were trialed to determine which technique immediately reduced headache intensity:

1. Sustained natural apophyseal glides (SNAGs) targeting the upper cervical spine: A posteroanterior mobilization of C2, sustained for 10 to 30s with a maximum of 6 repetitions
2. Reverse headache SNAGs: an anteroposterior mobilization of C2, sustained for 10 to 30s for a maximum of 6 repetitions
3. Modified headache SNAGs: a posteroanterior mobilization of C3, sustained for 10 to 30
4. Upper cervical traction: with the participant in a supine position. The therapist pronates the forearm against the subject’s occiput while fixing the subject’s chin. The resultant traction force will be sustained for 10–30 s for a maximum of 10 repetitions

If one of the above techniques successfully reduced headache intensity, it was used as treatment. The same MMT technique was applied to the subsequent treatments in case the subject presented with a headache. If the participant had no more headaches at the time of treatment, a C1-C2 rotation SNAG was used to improve restricted cervical rotation ROM.

Sham MMT plus exercises

Apart from doing the same exercises for 20 minutes, the sham intervention mimicked MMT without applying manual force using the headache SNAG technique, holding positions for 10 to 30s and 6 repetitions applied. The sham MMT also lasted for 10 minutes

Each participant received six treatment sessions over four weeks, followed by a home exercise program. The first two weeks included two sessions per week, and the last two weeks included one session per week. Each treatment session lasted a maximum of 30 minutes.

The primary outcome was measured by headache frequency (days/month). Secondary outcomes included:

• Headache intensity, from 0-10 recorded through a headache diary
• Headache duration, measured in hours per week and recorded through a headache diary
• Headache medication intake, recorded through a headache diary
• Headache disability was measured using the Headache Activities of Daily Living Index which ranges from 0-45 with higher scores indicating more disability
• Upper cervical rotation range of motion (ROM), measured using the Flexion Rotation Test
• Pressure pain thresholds,
• Patient satisfaction from 0-100 on a numerical rating scale

The smallest worthwhile differences were 1 point for headache intensity, 7 hours per week for headache duration, 2.5 tablets per week for headache medication intake, 4.5 points for headache disability, 40% for patient satisfaction, and 5 degrees for upper cervical rotation ROM. No smallest worthwhile difference was defined for the pressure pain thresholds.

Follow-up assessments were conducted at 4, 13, and 26 weeks.

Results

A total of 99 participants were recruited and randomized to one of the three groups. The groups were balanced at baseline. The baseline headache frequency was 6 days per month. As such, the authors set the threshold for the smallest worthwhile effect using the requirement of a 50% reduction in headache frequency at 3 days per month.

The primary outcome headache frequency was reduced to 3 days per week in the MMT plus exercise group compared to 5 days per week in the sham MMT plus exercise group and exercise alone group at 4 weeks. In the MMT plus exercise group, this frequency was further reduced to 2 days per week and 1 day per week at weeks 13 and 26 respectively. In the other 2 groups, the headache frequency at 13 weeks remained unchanged and decreased to 4 days per week at 26 weeks.

This led to a between-group difference in the primary outcome headache frequency of -3 at four weeks to -3 at 13 weeks and -4 at 26 weeks. The upper limit of the confidence intervals indicated that at 13 and 26 weeks, the smallest worthwhile effect threshold was exceeded. At 4 weeks, the effect was thus uncertain, while a statistically significant and clinically relevant difference was obtained at 13 and 26 weeks.

Most of the secondary outcomes followed the course of the primary outcome over the weeks.

Headache intensity showed close to no effect at 4 weeks when looking at the 95% confidence interval since that interval spanned the value of the smallest worthwhile effect. At 13 and 26 weeks, a significant and clinically relevant effect was noted.

Headache duration showed no improvement at all time points when looking at the confidence intervals. This was also the case regarding medication intake.

Headache disability improved to an uncertain level at 4 weeks but to a clinically relevant difference beyond the smallest worthwhile thresholds at 13 and 26 weeks.

Satisfaction outcomes indicated no effect at 4 weeks since the confidence interval was not significant. At 13 and 26 weeks, the satisfaction was clinically significant and relevant in the MMT plus exercise group.

Upper cervical rotation ROM improved to a statistically significant and clinically relevant difference at all time points in the MMT plus exercise group. The pain pressure thresholds showed the largest sustained effect at the tibialis anterior. At the zygapophyseal joints and the suboccipital region, the effects were small at 4 weeks and increased to moderate improvements at 26 weeks.

Questions and thoughts

Manual therapy is often blamed for being unspecific and passive. This study proves that you don’t need 30 minutes of passive treatment to obtain clinically relevant results. Exercise alone in this study also reduced headache frequency but not to a clinically relevant level, since no 50% headache frequency reduction was obtained (instead a reduction from 6 days per week to 4 was observed). This was in contrast to exercise combined with manual therapy for cervicogenic headache, where more than 80% reduction in headache frequency was found (from 6 days per week to 1). This study therefore shows that combining manual therapy for cervicogenic headaches with exercise yields the best improvement.

The current study also demonstrated that these beneficial effects of exercise combined with manual therapy for cervicogenic headache are not due to placebo effects, since the sham manual therapy plus exercise group had the same improvement as the exercise alone group and the real manual therapy plus exercise group outperformed both on the primary outcome.

In this study, only one technique that immediately reduced headache intensity was used. Instead of “delivering” all kinds of passive techniques over the whole duration of a treatment session, the current study only used 1 technique for 6 to 10 repetitions of 10-30 seconds, depending on the technique used. The goal was to find one technique that immediately reduces headache intensity; otherwise, another technique was chosen. Using this simple approach, you can make a large difference over a very short period of treatment time. The other 20 minutes of the treatment session were spent on exercises. This way you get the best of both worlds to maximize your treatment effect.

Apart from the significant improvements in the primary outcome headache frequency, the secondary outcome patient satisfaction showed a large difference in satisfaction outcomes across the groups. While a 40% difference in satisfaction was set as the threshold for defining the smallest worthwhile effect, this study showed satisfactory results for manual therapy plus exercise since at 13 weeks the difference was more than 50 points higher at 13 and 26 weeks compared to the other groups. (see also table 5).

Patients are greatly satisfied when receiving manual therapy for cervicogenic headache combined with exercise and have a clinically relevant large reduction in headache frequency at 13 and 26 weeks after only 4 weeks of treatment. So why wouldn’t you use manual therapy?

Talk nerdy to me

A lot of criticism of manual therapy is often expressed. Often the commentaries point to placebo and expectation effects. In trials, one hurdle to tackle is blinding the participants. When a participant is not blinded and receives a treatment, they might think that this treatment will be successful. In this study, the authors created a sham manual therapy group to assess possible placebo effects. The assessment of blinding success indicated that blinding was successful in this study since 21 and 20 out of 33 in the manual therapy and sham manual therapy groups respectively thought they did not receive manual therapy. This way we know that the effects are not due to placebo.

The expectations of the participants can also play a huge role in treatment outcomes. But if you expect manual therapy to give good outcomes and you think you did not receive manual therapy yet you still had very good outcomes, I think that possible patient expectations at baseline had little to no influence on the outcomes.

We should however take into account the high baseline pain intensity levels. A headache intensity of at least 6 out of 10 is high and not always seen in first-line or secondary care. Generally speaking, pain intensity in cervicogenic headache is mild which means 4-7 out of 10. Thus, this study potentially included participants with more severe cervicogenic headache features. As the requirement for inclusion was the headache being present for more than 1 year with such a high pain intensity level, we can assume that this study included people with more severe characteristics than the one you will potentially encounter in clinical practice. This means that these participants, due to their high baseline pain intensity could also have regressed to the mean during the study period.

Take-home messages

Six sessions over four weeks of manual therapy for cervicogenic headache combined with exercise is more effective in reducing headache frequency than sham manual therapy plus exercise and exercise alone at 13 and 26 weeks. Sham manual therapy plus exercise and exercise alone did not achieve significant, nor relevant results. The effective combination of exercise and manual therapy for cervicogenic headache included 10 minutes of manual therapy using only 1 mobilization technique and 20 minutes of exercises.

Reference

Satpute K, Bedekar N, Hall T. Mulligan manual therapy added to exercise improves headache frequency, intensity and disability more than exercise alone in people with cervicogenic headache: a randomised trial. J Physiother. 2024 Jul;70(3):224-233. doi: 10.1016/j.jphys.2024.06.002. Epub 2024 Jun 19. PMID: 38902195.