The prognosis and methods of coping with headaches were studied by telephone interview in 98 children followed for 10 years after diagnosis at the Department of Pediatrics, Dalhousie University and Children’s Hospital, Halifax, Canada. Of 77 contacted, 18 (23%) were diagnosed initially with tension-type headaches, 49 (64%) with common migraine, and 5 (6%) had classical migraine with aura. Headaches had persisted in 73% but were improved in 81%. Prognosis was related to the type of headache at initial diagnosis: only 18% of those with migraine were symptom free after 10 years whereas 50% of those with tension headaches had obtained relief. Headache types changed with time: 11% of children with initial tension headaches had migraine at follow-up, whereas 22% with common migraine had developed headaches of tension-type after 10 years. At diagnosis the mean headache frequency was 11/month compared to 2/month at 10 year follow-up. Treatment at follow-up consisted of 1) a period of sleep in 17%; 2) relaxation therapy in 10%; and 3) avoidance of precipitants in 5%. Food was a precipitant in 5 patients. One-third required no treatment for headaches; 41% were using nonprescription medications: and only 3% had recently seen a physician for prescription medication. Associated symptoms included motion sickness in 26% somnambulism in 14%, “Alice in Wonderland” hallucinations of size in 10%, and hallucinations of motion and time (“The Rushes”) in 17%. 
COMMENT. In this group of children, no patient was prescribed medication at diagnosis and only two required prescription drugs at follow-up. Nonprescription medications had been employed by 30 - 40%, but rest and relaxation techniques were encouraged as alternatives to drugs when possible.
Treatment of childhood migraine with autogenic relaxation and skin temperature biofeedback was analysed in 30 patients, ages 7 to 18 years, using a controlled group outcome design, at the Department of Psychology, University of South Alabama, Mobile, AL . Headache frequency and duration, but not headache intensity, improved in the treatment groups as compared to the waiting list controls; 80% of the biofeedback group, 50% of the autogenics group, and none of the controls were symptom-free. In this study, biofeedback and relaxation therapies were practical and effective in the management of childhood migraine and were free from adverse side-effects.
In a previous report from the University of Ottawa, Canada, McGrath et al found that relaxation training was no more effective than brief reassurance and self-control suggestion techniques in treating pediatric migraine (see Progress in Pediatric Neurology I, Chicago, PNB Publ, 1991, pp 144-5). Perhaps certain foods may have been observed to play a greater role in precipitating headaches in the Halifax study if the influence of diet had been stressed at the time of initial diagnosis.