Efficacy of a Novel App-based Migraine Treatment Platform - a Pilot Study

This study evaluates biofeedback delivered through a smartphone as a treatment for migraine in adolescents. Three out of four participants will receive treatment as smartphone delivered biofeedback training, and one out of four participants will receive treatment as a sham-biofeedback app.

Biofeedback is a behavioral treatment without known side effects. In biofeedback patients learn control over bodily reactions through feedback visualised on a screen. Typical bodily reactions that are displayed is heart rate and muscle tension. Biofeedback has for decades been considered as effective preventive treatment of migraine among children and adolescents. In spite of that, it is not available to most patients in need. This is most likely because it requires specialised stationary equipment and a trained therapist. In this project, the researchers have developed a new smartphone biofeedback app connected to wearable, wireless sensors measuring muscle tension, finger temperature and heart rate. The app and sensors will be used to deliver biofeedback treatment to adolescents suffering for migraine. The goal of the study is to investigate the effect of the biofeedback treatment and the sham-biofeedback treatment on migraine headaches.

The study will begin with a one-month baseline period, followed by a two-month treatment period. Participants will randomly be assigned to one of two groups by a computer-generated block-randomization list. In each block of four, participants will have a 75% chance of being allocated to the biofeedback group and 25% chance of being allocated to the sham group, to ensure 30 participants in the biofeedback group and 10 in the sham group. The randomization list will remain inaccessible to the investigators until end of follow-up of the last participant. Both participants and personnel will be blinded to the intervention given. It should be noted that this study is intended as a pilot study, and not a randomized controlled trial. Still, we wish to use randomization and blinding to ensure that the group receiving sham treatment is a random subset of the study population.

The randomization number will be drawn sequentially from a list of 40 numbers where one random in every four numbers results in downloading a sham-version of the app. Both versions of the app look alike and no pattern in the 5-digit number or the randomization list can reveal which version of the app is given. This will ensure blinding of both participants and study personnel. Breaking of the randomization will be made after follow-up of the last participant. The software developers will at this point reveal what 5-digit numbers corresponded to the biofeedback or sham version of the app.

The treatment comprises an app containing biofeedback training, instructions for self-delivery and a headache diary. The treatment utilizes these functionalities by a push-reminder in the app to complete a daily headache diary entry and a biofeedback session of 10 minutes duration daily. Prior to commencing treatment participants will be given basic information on the rationale behind biofeedback treatment, given instructions on how to use the equipment and software and instructions on how to complete a biofeedback session. The biofeedback is based on wireless sensors measuring muscle tension, finger temperature and heart rate.

Sham-biofeedback will be made by disrupting the connection between input of physiological parameters and the feedback. Thus, the feedback will simply be displayed as positive feedback occurring at random intervals throughout the sessions. The looks and contents of the normal app and the sham-app will be completely similar. The participants being allocated to the sham-goup will also use exactly the same sensor setup as the biofeedback group. The only difference between the two arms/interventions is the internal software algorithm, which will be inaccessible to the user. All participants in both groups will be given the same information and instructions.

Mean change in headache frequency from baseline to end of treatment. Evaluated as number of headache days.

Defined as positive if there has been more than 50% decrease in headache frequency from baseline to end of treatment

Mean change in maximal pain intensity from baseline to end of treatment. Rated on a 4 point scale (0=no headache; 3=severe headache).

Mean change in abortive drug consumption form baseline to end of treatment. Defined as days with use of abortive drugs.

Estimate of adherence by evaluating how many treatment sessions were completed during the treatment period.

Inclusion Criteria: diagnosis of migraine with or without aura according to ICHD-3 two to eight migraine attacks per month written informed consent signed by guardians, or by adolescent if age between 16-18 years. Exclusion Criteria: participant not speaking Norwegian reduced sensibility, hearing or vision to a degree that impairs proper use of the app serious psychiatric or neurologic disease and currently using migraine prophylaxis.

Nestoriuc Y, Martin A. Efficacy of biofeedback for migraine: a meta-analysis. Pain. 2007 Mar;128(1-2):111-27. doi: 10.1016/j.pain.2006.09.007. Epub 2006 Nov 2.

Palermo TM, Eccleston C, Lewandowski AS, de C Williams AC, Morley S. Randomized controlled trials of psychological therapies for management of chronic pain in children and adolescents: an updated meta-analytic review. Pain. 2010 Mar;148(3):387-397. doi: 10.1016/j.pain.2009.10.004. Epub 2009 Nov 11.

Stubberud A, Varkey E, McCrory DC, Pedersen SA, Linde M. Biofeedback as Prophylaxis for Pediatric Migraine: A Meta-analysis. Pediatrics. 2016 Aug;138(2):e20160675. doi: 10.1542/peds.2016-0675.