Abstract
Aims: The objective of this systematic review was to assess the effectiveness of spinal manipulations as a treatment for migraine headaches.
Method: Seven databases were searched from inception to November 2010. All randomized clinical trials (RCTs) investigating spinal manipulations performed by any type of healthcare professional for treating migraine headaches in human subjects were considered. The selection of studies, data extraction and validation were performed independently by two reviewers.
Results: Three RCTs met the inclusion criteria. Their methodological quality was mostly poor and ranged between 1 and 3 on the Jadad scale. Two RCTs suggested no effect of spinal manipulations in terms of Headache Index or migraine duration and disability compared with drug therapy, spinal manipulation plus drug therapy, or mobilization. One RCT showed significant improvements in migraine frequency, intensity, duration and disability associated with migraine compared with detuned interferential therapy. The most rigorous RCT demonstrated no effect of chiropractic spinal manipulation compared with mobilization or spinal manipulation by medical practitioner or physiotherapist on migraine duration or disability.
Conclusions: Current evidence does not support the use of spinal manipulations for the treatment for migraine headaches.
Introduction
Migraine headache (MH) is a common episodic neurobiological disorder, typically presenting with recurrent attacks of severe headache and autonomic dysfunction (1). The pathophysiology of MH is complex and may involve activation of sub-cortical structures such as the midbrain and the pons (2,3) as well as the trigeminovascular system (4). Recent research has also established the first genetic risk factor for MH (1). The prevalence of MH is high, affecting more than 10% of the general population (4). MH is associated with a high burden of personal suffering and considerable socioeconomic costs (5,6). The World Health Organization (WHO) ranked MH as number 19 among all diseases causing disability (7).
Most chiropractors feel that spinal manipulative therapy (SMT) is an effective treatment for MH. For instance, a 2004 survey of the UK General Chiropractic Council showed that 91% of all UK chiropractors believe that migraine ‘can be treated or managed by chiropractors’ (8). However, SMT has repeatedly been found to be a concept of debatable plausibility (9,10), which renders it an unlikely candidate for an effective therapy of MH. In addition, SMT is neither inexpensive nor entirely free of risks (11–13). Several hundred severe complications after upper spinal manipulations have been reported (e.g. (14,15)). However, the estimates as to the incidence of these complications vary hugely (16). Thus it seems important to be certain whether SMT is a safe and effective therapy for MH.
There are numerous reviews of SMT for headaches of all types but none exists of SMT specifically for MH. Therefore, this systematic review is aimed at critically evaluating the trial data for or against the notion that SMT is effective in treating MH.
Methods
Electronic searches were carried out in the following databases: Amed, Embase, Medline, Cinahl, Mantis, ICL and Cochrane Central Register of Controlled Trials (from their inception to November 2010). The search terms were constructed over two concepts: spinal manipulation and migraine headaches. Our own extensive department files were hand-searched. The abstracts of the articles thus located were screened in EndNote to remove duplicates and irrelevant studies. No language limitations were imposed.
To be included, a clinical trial had to be randomized or quasi-randomized, test the effectiveness of SMT and focus on the treatment of MH in human subjects. Any type of control intervention and any clinical outcome were admissible. Trials of cervicogenic headaches (e.g. (17–24)) or tension type headaches (e.g. (25–30)) were excluded.
Key data of the included trials were extracted according to pre-specified criteria. Data extraction was performed by two independent reviewers. The methodological quality of all reviewed studies was estimated using the Jadad score (31). Again, this was done by two independent reviewers (EE,PP).
For the purpose of this review, we defined SMT as a manual technique commonly used by chiropractors, osteopaths, physiotherapists or bone setters to correct misalignments of the spinal joints (32).
Results
The search strategy generated a total of 457 ‘hits’. After initial screening of abstracts, 112 references were considered to be potentially relevant. A total of 23 papers were retrieved for further evaluation of which 3 RCTs involving 430 patients with MH were eligible for inclusion (see Figure 1) (33–35). Their key data are summarized in Table 1.
Flowchart of eligibility assessment and inclusion. Controlled studies of spinal manipulations for the treatment of migraine headaches SMT: spinal manipulative therapy.
Details of the spinal manipulation intervention
Tuchin et al. (33) aimed to assess the efficacy of chiropractic SMT in the treatment of migraine. They reported statistically significant improvement in migraine frequency, duration, disability and medication use in SMT group. This trial lacked allocation concealment, appropriate randomization, blinding and intention to treat analysis. Other sources of bias included problems with selective outcome reporting and handling of incomplete data. We scored this study as 1.
Parker et al. (34) aimed to evaluate the efficacy of cervical manipulation performed by chiropractors, physiotherapists and medical practitioners for migraines versus mobilization performed by physiotherapists or medical practitioners. They reported no between group differences in migraine symptoms, but chiropractic patients did report a greater reduction in pain associated with their attacks. This study lacked appropriate randomization, blinding, allocation concealment and intention to treat analysis. We scored it as 1.
Nelson et al. (35) aimed to measure the relative efficacy of amitriptyline, spinal manipulation and the combination of both therapies for the prophylaxis of migraine headache. They reported clinically important improvement in both primary and secondary outcomes in all three study groups over time. This was a relatively well designed trial, but lack of blinding and control for placebo effects limit its conclusiveness. We scored this study as 3.
Discussion
Quality assessment of the included studies (Jadad score) a
1 is the worst and 5 is the best outcome on the jadad scale.
Only one RCT was of relatively high quality and it arrived at negative conclusions (35). One study of low quality arrived at negative conclusions (34) and the third RCT, also of low quality, favoured SMT (33). The findings of this trial (33) might not be due to specific therapeutic effects of SMT but could be due to a placebo response.
Adverse effects (AEs) reported in RCTs
Additional sources of bias in the study by Parker et al. (34) were related to lack of power calculations and ambiguous eligibility criteria. In the study by Tuchin et al. (33) there were serious threats to internal validity that included lack of means and standard deviations; statistical procedures incapable of detecting intergroup differences (Student's t-test and one-way ANOVA); and poor reporting quality. In the study by Nelson et al. (35), additional sources of bias pertain to the length of treatment (14 visits in the SMT group versus 3 visits in drug therapy) and the high drop-out rate. Finally, none of the RCTs adhered to International Headache Society guidelines for migraine prevention trials (37).
Two studies (out of three) reported adverse effects (33,34); and one RCT failed to provide such information (Table 4) (35). In the study by Parker et al. (34) the likelihood of adverse effects was statistically significant in the SMT group (p < 0.005).
Claims made by chiropractors on the internet
The first ten ‘hits’ obtained by Google search ‘chiropractic AND migraine’ are shown.
Our review has several limitations. Even though our searches were extensive, we cannot be entirely sure that all relevant articles were located. Publication bias may have resulted in negative studies remaining unpublished. The number of trials included in our review, their total sample size and their methodological quality were too low to allow definitive judgments. Also, although all included RCTs were considered to have homogenous MH populations, statistical pooling was not feasible owing to lack of reporting of sufficient raw data. However, this review has several strengths, which include the comprehensive search strategy, the inclusion of only the highest quality trial design and use of suggested methods for systematic reviews of interventions for MH.
Future studies of SMT should be in line with accepted standards of trial design and reporting (e.g. CONSORT guidelines). In particular, studies should be of adequate sample size based on power calculations, use validated outcome measures, control for non-specific effects and minimize other sources of bias. Reporting of these studies should be such that results can be independently replicated.
To conclude, based on the scarce scientific evidence and the poor quality of the randomized trials, SMT cannot be recommended for the treatment of MH.