Meditation for Migraines: A Pilot Randomized Controlled Trial

Rebecca Erwin Wells MD, MPH; Rebecca Burch MD; Randall H. Paulsen MD; Peter M. Wayne PhD; Timothy T. Houle PhD; Elizabeth Loder MD, MPH


Headache. 2014;54(9):1484-1495. 

In This Article


Study Population

See the Figure for research design. In this single-site study, we recruited migraine patients via flyers, referrals, and medical records from January-March 2012 from Brigham and Women's Hospital, primarily through the John R. Graham Headache Center, a tertiary care academic headache center in Boston, MA. Potential participants were evaluated via an initial telephone screen and, if deemed potentially eligible, were subsequently evaluated at an in-person visit by a neurology headache specialist (REW or RB) who assessed whether they met criteria for inclusion in the study. Participants then deemed eligible to participate in the study maintained a paper headache log for a 28-day run-in period. Both the Guidelines for Controlled Trials of Drugs in Migraine[29] and the Guidelines for Trials of Behavioral Treatments for Recurrent Headache[30] recommend a 1-month baseline period. This log was then reviewed by the headache experts to confirm eligibility for the study. The study was approved by the Brigham and Women's human subjects' research review board. Each participant signed a written informed consent document. This study was prospectively registered ( identifier NCT01545466).


CONSORT Flow Diagram: Flow of participants through trial; *Mindfulness Based Stress Reduction (MBSR), 8-weekly group classes plus retreat day taught by trained instructor.

Inclusion criteria included: diagnosis of migraine with or without aura (according to the International Classification of Headache Disorders-II);[31] 4–14 migraine days/month; ≥one year history of migraines; ≥18 years old; able and willing to attend weekly sessions and willing to participate in daily mindfulness assignments of up to 30–45 minutes/day; agreeable to participate and to be randomized to either group; fluent in English; and in good general health with no additional diseases expected to interfere with the study. Exclusion criteria included: current regular meditation/yoga practice; major systemic illness or unstable medical/psychiatric condition (eg, suicide risk) requiring immediate treatment or that could compromise protocol adherence; medication overuse headache, (according to the International Classification of Headache Disorders-II);[31] current/planned pregnancy or breastfeeding; new prophylactic migraine medicine started within 4 weeks of the screening visit; unwilling to maintain stable migraine medication dosages; and failure to complete baseline headache logs.

Study Design

This study was a prospective, randomized (1:1) clinical trial to evaluate if an 8-week MBSR program is superior to usual care. Participants were allowed to continue taking their prophylactic and abortive medications as usual, and were asked to not change dosages for the duration of the trial. As seen in Table 1's baseline characteristics of the study participants, 89% of those in the control group were being treated with daily prophylactic medications, which were allowed to be continued for the duration of the trial. All participants (100% of intervention and 100% of the control group) were already taking abortive headache medications, and they were allowed to continue these for the duration of the trial. Patients randomized to maintain usual care were offered MBSR at the conclusion of the study and were asked not to start a yoga or meditation class on their own in the meantime. Thus, participants in the control group were able to continue in their usual care during the duration of the study but were also offered the intervention as a courtesy at the conclusion of the study. In an attempt to blind the control group, participants were told there were two start times for the MBSR course, with randomization to either date. The group randomized to the later date was the control group, continuing usual care during the interim. Treatment assignments were generated using permuted block randomization with randomly varying block size and sealed in numbered, opaque envelopes. Given the low risk potential for this study, a full data and safety monitoring board was not required.

MBSR Intervention

The standardized MBSR class met for 8-weekly 2-hour sessions, plus one "mindfulness retreat day" (6 hours) led by a trained instructor (RHP) who followed the structured MBSR protocol created by Dr. Jon Kabat-Zinn.[15] The protocol for this entire course (the 8-weekly classes plus the retreat day) was identical to the Kabat-Zinn protocol, without modifications for migraineurs. The original MBSR protocol was created to be provided in a group setting, as was done with this course. The instructor has been fully trained in MBSR by the Center for Mindfulness in Medicine, Health Care, and Society at the University of Massachusetts Medical School, where Dr. Jon Kabat Zinn created and developed this intervention. In addition, the instructor has had 5 years of supervised MBSR group leadership training under the Center for Mindfulness' staff. The intervention is based on systematic and intensive training in mindfulness meditation and mindful hatha yoga in the context of mind/body medicine. The theoretical underpinnings of mind/body medicine were taught throughout the course with interactive discussions and experiential practice. Mindfulness, defined as non-judgmental moment-to-moment awareness, was cultivated through mindful eating, mindful breathing with sitting and walking meditation, body scan (sequential mindful attention to different body parts) and mindful movement (yoga). Weekly course content is described extensively in Kabat-Zinn's book,[15] but briefly the first class begins with mindfulness of breathing, mindful eating, and the body scan, and subsequent classes build on these practices and slowly add in the other meditative practices. The all-day retreat includes elements of all the mindfulness practices. The instructor also gives information about stress and stress relief during the fourth class. During each class, participants can share their experiences of the practice of mindfulness with other students. A central theme of the course involves teaching participants to use the MBSR skills as a means to reduce the negative effects of stress reactivity and to develop more effective ways of responding positively and proactively in stressful situations and experiences. Specifically, by repeatedly bringing attention back to the natural rhythm of the breath, participants are encouraged to build their capacity to attend to physical and mental percepts. In addition to learning and practicing the formal practices of mindfulness meditation, participants are advised to incorporate mindfulness into their daily lives so that routine activities (brushing teeth, taking a shower, washing dishes, etc) can become a meditative practice. The ultimate goal is for patients to build their mindfulness practice and MBSR skills to develop a more flexible capacity to utilize mindfulness in a variety of everyday situations. During all classes, chairs were provided for seated exercises and mats were provided for the yoga. In addition to practicing during class, each participant was given the same standard guided audio recordings and encouraged to practice at home to build their daily mindfulness practice for 45 minutes per day, at least 5 additional days per week. Compliance was monitored through class attendance and by daily logs of home practice.

Outcome Measures

The primary outcome was change in migraine frequency (number of migraines/month [28 days]) from baseline to initial follow-up (immediately after MBSR ended). Secondary outcomes included: change in headache severity, duration, self-efficacy, perceived stress, migraine-related disability/impact, anxiety, depression, mindfulness, and quality of life from baseline to initial follow-up. We also explored results from baseline to final follow-up (a month after MBSR ended).

We tracked recruitment and enrollment rates, class attendance, daily logs of home practice and adverse events. Participants maintained daily paper headache logs from the initial baseline screening period to the final follow-up to record the number of headache days/month, severity of each headache (0–10 scale, 0 being no pain and 10 being severe pain), duration (in minutes), and medicines taken. At all 3 study visits, all participants completed a battery of standardized, validated instruments. Headache-related disability was measured with the Headache Impact Test-6 (HIT-6)[32] and the 1-month (rather than 3 months) Migraine Disability Assessment (MIDAS).[33] Quality of life was measured with the Migraine Specific Quality of Life Questionnaire, version 2.1.[34] Depression was assessed with the PRIME-MD Patient Health Questionnaire-depression module (PHQ-9).[35] Anxiety was measured with the State Trait Anxiety Inventory.[36] Stress was measured with the perceived stress scale-10.[37] Mindfulness was measured with the Five Facet Mindfulness Questionnaire.[38] Self-efficacy was measured with the Headache Management Self-Efficacy Scale.[39]

Study data were collected and managed using the Research Electronic Data Capture system (REDCap) hosted at Brigham and Women's Hospital.[40] REDCap is a secure, web-based application designed to support data capture for research studies, providing: (1) an intuitive interface for validated data entry; (2) audit trails for tracking data manipulation and export procedures; (3) automated export procedures for seamless data downloads to common statistical packages; and (4) procedures for importing data from external sources.

Sample Size

We estimated standard deviations for these analyses based on a previous study of yoga for migraines.[41] Allowing for 10% loss due to dropouts, a sample size of 34 would provide 80% power to detect between-group differences of about 3 fewer migraines/month, approximately half the effect size in the yoga study. Time constraints unrelated to the feasibility of the study (eg, REW's relocation) limited recruitment to 3 months and decreased our ability to reach our target sample size.

Statistical Analyses

Descriptive statistics were used to analyze adherence and baseline characteristics. For the purposes of determining whether potential subjects were eligible for the trial, once subjects were confirmed to have migraine, baseline headaches were considered to be migraines. For all final analyses, migraines were more precisely defined as those headaches that were >4 hours long with a severity of 6–10, based on patient diary information. Mann-Whitney U-tests were performed to compare changes from baseline to follow-up in the intervention vs control group and for any change scores that violated parametric assumptions, the 95% CI was produced with the Hodges-Lehman estimate. To estimate effect sizes for variables that satisfied parametric assumptions, we ran independent t-tests on the change scores and report the differential change and 95% confidence intervals (CI). Given that headache characteristics were analyzed over 28 days, "baseline" reflects the 28-day period before the intervention, "initial follow-up" reflects the last 28-day period of the intervention, and "final follow-up" reflects the 28-day period after the intervention ended. For the standardized instruments, "baseline," "initial follow-up," and "final follow-up" reflect the time points of each assessment. All analyses were blinded and performed on an intention-to-treat basis using IBM SPSS Statistics 21 (IBM Corporation, Armonk, NY, USA). Although statistical inferences are conducted, this effort is a pilot study, and efforts are made to report and interpret effect sizes throughout. Where appropriate, all testing was 2-tailed with P < .05.