Biofeedback, Relaxation Techniques and Attitudinal Changes in Adolescents with Migraines: A Feasibility Study


From 3.2% to 9% of school-age children suffer from migraine headaches. Many physicians are concerned that pharmacological treatment of migraines can have undesirable side-effects, as well as lead to drug dependence in adolescents. A number of review articles have shown that biofeedback, behaviour modification and relaxation exercises can significantly help migraine sufferers. This article describes a feasibility study undertaken at a West Toronto high school to see if non-pharmacological treatment of migraines was effective. (Can Fam Physician 937; 33:417-421.)

Key words: migraine headaches, children, biofeedback

Dr. Borins, a certificant of the College, is an instructor in the Department of Family and Community Medicine at the University of Toronto. He is also an active staff member of the Department of Family & Community Medicine at St. Joseph’s Health Centre, Toronto. Crystal Hawk is a health-care educator and psychotherapist in private practice in Toronto. 

Childhood Migraine is much more common than most people realize. Researchers Bille and Sparks have found that from 3.2% to 9% of children suffer migraine. Family physicians do not usually see a large patient population of child migraine. Bille and others noted that children seldom consult a doctor because usually one of the parents has had migraine for many years, and in these families well-known symptoms are taken relatively lightly. Parents also become resigned, believing that little can be done about migraine attacks except to give medication and so do not bother to take the child to their physician.

Migraine has always been viewed as a model for a psychosomatic disorder because the mind and the emotions have a profound effect on the nature of the physical disease, and vice versa. Techniques which alter physiology could well be adapted to other so-called "psychosomatic diseases" such as peptic ulcers, Raynaud’s disease, temporomandibular joint dysfunction and asthma.

The pain of migraine may be caused by dilated, hyperpermeable, extracranial and intracrainial arteries. The dysfunction of vascular behaviour in the head could be related to intense sympathetic activation followed by over-compensation and the subsequent vasodilation. This accounts for the brief phase of vasoconstriction thought to be associated with the pre-headache aura and the subsequent painful vasodilation. Theoretically, biofeedback and relaxation interferes with this initial sympathetic activation and teaches the autonomic system to be less reactive.

Two review articles have suggested that biofeedback relaxation-treatment procedures and behaviour-modification techniques can significantly help both adults and children suffering from headaches.

A feasibility study was undertaken in a high school setting to see if a program on biofeedback, relaxation and medical information could be introduced.


Announcements were made in a West Toronto high school of 1,300 children that a pilot project teaching students how to prevent migraine headaches would be undertaken, and volunteers were requested. Of the 17 students with headaches identified, three students with migraines agreed to participate. Two were girls (aged 14 and 16), and one was a boy (aged 15).

Many investigators have used the diagnostic criteria of Vahlquist, which requires paroxysmal headaches separated by pain-free intervals and at least two of the following: nausea, aura, unilateral pain and positive family history. For the identification of migraine, we decided to rely on the standards in the community: that is to say we accepted as confirmation the diagnosis of the patient’s doctor. Our questionnaire, too, helped to confirm the diagnosis.

A pre-program questionnaire was circulated to each student and the parents, detailing the frequency, duration and severity of the student’s migraines. The amount of medication taken and the number of hours missed from school and other activities were documented. Family history and precipitating factors were elicited.

The volunteers then collected data on their migraine headaches. They were seen after school for two  hours twice a week for three weeks and once each week for the next two weeks. They were educated about all aspects of migraine headaches. They were taught thermo-biofeedback, using a hand-held thermometer; an active progressive relaxation technique; and an abdominal breathing exercise; they were instructed to practice these exercises daily. Parents, teachers and a school nurse were also involved in supportive and educational functions.

Family dynamics play an important part in the experience of children with migraine. Therefore, as a part of the study, meetings with parents were held at the beginning and end of the program. At these meetings details of the study were discussed and their support and answers to questions were elicited. Because interactions among family members may affect a child’s headache patterns, a whole family might be affected as changes took place in the children. The parents were taught that support and new kinds of attention should be provided when the children were headache free.

Since our purposes were to duplicate this program in other schools, and to have the project continue after the researchers left, two sessions were spent with the teachers and school nurse to familiarize them with the program. These staff members also provided another support system for the student subjects. Teachers or the school nurse could provide follow-up for students at a later time to reinforce techniques and discover problem areas.


All three students reported a reduction in the number and severity of their headaches, as well as in the amount of medication taken and time lost from activities. Since the sample size was so small, there was no control group, and since no diaries were kept to substantiate the numbers reported, no claims of therapeutic efficacy can be made. For these reasons, only anecdotal information will be given.

KN, a 14-year-old female grade nine student, had her first headache at age 13. She had been having about three headaches a month, each lasting from four to eight hours. She had been taking either two Aspirin or one Tylenol for pain, but never more than three tablets per day. However, the medication was often ineffective and seemed to make her drowsy, She diligently practiced the program, and during its six weeks duration she was migraine free. Five months after the program she was doing the finger-warming exercise nightly at bedtime and had experienced only two headaches. When she felt the start of a headache, she was able to interfere quickly with the cycle and with the pattern as well. In both instances the headaches were started by food triggers, and she did not allow them to proceed to the pre-treatment level. Fifteen months after the program began, KN had been migraine free for the past five months.

TG was a 16-year-old female grade ten student whose migraines had begun when she was eight years old. She was having one headache every two weeks, lasting from 12 hours to five days. She had missed 22 days of school during the previous six months, mostly as a result of these headaches. To control the pain, she had been using 222s, Aspirin or Bufferin, taking two tablets three or four times daily. These medications were causing drowsiness and dyspepsia. Until the program began, she had been also seeing a chiropractor weekly, with no improvement or change in the pattern of her pain.

During the program and for four months afterward TG was migraine free. She practiced the finger-warming exercise daily, and she twice aborted an attack when she noticed a "tightness in her head", by doing the finder-warming exercise and head massage.

At 15 months follow-up she had had no absences from school because of migraines. She still experienced the occasional migraine once every two to three months, but they were less intense and lasted only a few hours. 

KC was a 15-and-a-half-year-old male grade ten student who experienced only about one headache per month. He was taking three Cafergot or one to two Tylenol tablets two to three times daily, as needed, but said that the medication offered little relief. Because his migraine pattern was monthly, it was difficult for him to be motivated to practice a daily self-help program which provided immediate results.  He had two migraines within four months from the start of the program. One he described as his usual "headcracker", but the second he was able to abort by practicing some relaxation techniques.

At 15 months follow-up he had had no headache for five months and had been successful in using the finger-warming exercise to "calm down" his last migraine.


Subjectively, all three students benefited from this project. However, the small sample size, the lack of control group and the absence of diaries to substantiate the results that were reported make it impossible to draw any definite conclusions. A number of difficulties were encountered in this pilot project that would need attention in a larger study.


If 3.2% to 9% of school children have migraines and 30% to 50% of adults with migraine had such headaches before the age of 20, why where there only three students who agreed to participate out of a school population of 1,300?

Announcements made to high school students do not often attract their attention. Some thought that in coming forward and identifying themselves they might be ostracized by their peers or considered "sick" or "different". Others said they had busy school schedules and could not sacrifice the time needed to participate.

More care would be needed to contact each student individually and confidentially. Written communication, returned to a confidential depot, might allay students’ fears of being singled out. Highly respected recruiters who are influential with students could be identified and used to encourage students to take part in such a survey.

For none of the three students did their physicians report migraines on their school physical forms.  Surprisingly, some parents had difficulty contacting their doctors to get their confirmation of a diagnosis of migraine. Others were reluctant to phone their doctors, and at least two doctors considered that the project would involve them in too much paperwork and advised their patients not to take part. The doctor of one of the participants refused to co-operate, and the child’s mother had to phone the doctor back a second time to get him to agree to sign the paper confirming the migraine diagnosis.

Perhaps many physicians resented their patients being involved in treatment of a medical condition at school. Sometimes doctors can feel threatened that someone else is "stealing" their patients. An alternative approach to treatment might be seen as a "put down" to their approach. Doctors are continually being bombarded with forms to sign and have a tendency to want to decrease paperwork.

A letter written by a physician to deflect these concerns might help to obtain physicians’ co-operation. A follow-up phone call by the physician researcher might ensure that communication to family physicians would be optimal. Since it was the intention of this project to apply the techniques tested in a family physician’s office, it would be helpful to have patients’ doctors involved in a positive way.

We have often observed among migraine sufferers a lack of willingness to take part in self-help programs. Perhaps there is some secondary gain inducing people to hold onto their migraine pattern. Perhaps some have sought treatment so often that they mistrust all approaches which claim success. Perhaps there are some qualities of a psychosomatic personality which preclude these people from gaining mastery over their autonomic nervous system.

Assessing the Efficacy of the Intervention

Spontaneous remissions are common with migraine, and Brensky found that irrespective of the form of treatment, about one-half of all patients had more than a 50% reduction headache frequency in the six months following their initial visit to a neurologist. The reason may be that any positive energy introduced into a stable pathological condition upsets the balance and has some placebo effect. Perhaps referral to a neurologist may co-exist with an acute exacerbation of the disease which might have resolved spontaneously if a neurologist had not been consulted.

Because of the strong placebo effect, children in a future study would need to be allocated at random to different intervention groups. Data could be collected from children in all groups by means of diaries.

A control group, for example, might use diaries to collect data about the frequency and severity of their headaches, the amount of medication used, and the hours missed from activities.

A second group might keep diaries and also take part in a program involving medical information and support of peers and parents.  A third group might be exposed to the whole package of interventions including thermal biofeedback, relaxation and abdominal breathing. Comparisons might be made across all three groups and the findings compared to previous baseline data.

Many authors agree that children are excellent subjects for biofeedback, autogenic training and relaxation exercises. However, we found that the most difficult task in working with adolescents is to attract these young people to participate in their own self-help program. Teenagers do not like to be singled out as "different". They are involved in busy school schedules, lack the time to participate, and may not receive support and encouragement from their families and peers.


The authors wish to thank Dr. Earl Dunn, Rita Shaughnessey and Maggie Likavec for their valuable assistance and the Department of Family and Community Medicine of the University of Toronto for funding the project.


1.  Bille B. Migraine in school children. Acta Paediat 1962: 51:14-151.
2.  Bille B. Migraine in childhood and its prognosis. Cephalagia 1981: 1:71-5.
3.  Sparks JP. The incidence of migraine in school children. Practitioner 1978; 221:407-11.
4.  Dalessio D. Wolff’s headache and other head pain. London, U.K. : Oxford University Press, 1972.
5.  Sarjeant J. Walters E, Green E. Psychosomatic self-regulation of migraine headache. Semin Psychiatry 1973; 5:415-28.
6.  Shinnair S. D’Souza BJ. The diagnosis and management of headaches in childhood. Pediatr Clin North Am 1981; 29:79-94.
7.  Brown JK. Migraine and migraine equivalents in children. Dev Med Child Neurol 1977; 19:683-92.
8.  Jessup BA, Neufeld RWJ, Merskey H. Biofeedback therapy for headache and other pain: an evaluative review. Pain 1979; 7:225-70.
9.  Hoelscher TJ, Lichstein KL. Behaviour assessment and treatment of child migraine: implications for clincal research and practice. Headache 1984; 24:94-103.
10.  Vahlquist B. Migraine in children. Int Arch Allergy Appl Immun 1955; 7:348-55.
11.  Werder DS, Sargent JD. A study of childhood headache using biofeedback as a treatment alternative. Headache 1984; 24:122-6.
12.  Certain foods provoke migraine. Nutrition Reviews 1984; 42(2):41-2
13.  Prensky AL, Sonmer D. Diagnosis and treatment of migraine in children. Neurology 1979; 506-10.
14.  Libo LM, Arnold GE. Does training to criterior influence improvement? A follow-up study of EMG and thermal biofeedback. J behav Med 1983; 6:4.