Postgraduate Medical Journal (December 1984) 60, 841-846

Headache

RUTH ATKINSON OTTO APPENZELLERM.D. M.D., Ph.D.

Department of Neurology, University of New Mexico Hospital/2 South, University of New Mexico School ofMedicine, Albuquerque, New Mexico 87131, U.S.A.

Introduction

Headache is one of the more common symptoms thatcauses a patient to seek medical attention. It mayaccompany serious organic disease or result from aneurogenic vascular disorder or a psychologicalproblem. Almost everyone has a severe headache atsome time, but chronic recurrent headaches mostoften bring the patient to a physician. The pathogen-esis of chronic recurrent headaches is incompletelyunderstood; their categorization can suggest, how-ever, therapies that may control the patient's symp-toms. Treatment has been empiric, and continues tobe so without full understanding of the mechanismsinvolved.

Mechanisms of headache

Pain perception depends on neural pathways.Information from peripheral pain receptors is pro-cessed in spinal cord, brainstem, thalamus, andcortex, and serotonin is probably an inhibitorytransmitter in pain modulation (Basbaum and Fields,1978). Pain may also result from injury to the centralor peripheral nervous system and can occur withoutactivation of peripheral receptors. It is not presentlyclear which types ofheadache originate centrally andwhich peripherally. In one study (Blau and Dexter,1981), the majority of patients had an intracranialpain component during a migraine headache andboth intra- and extra-cranial components were evi-dent in only approximately half of other patients.Headache may arise from dysfunction, displace-

ment or encroachment on pain-sensitive intracranialstructures that include the proximal portions of thecerebral arteries and the large veins and venoussinuses (Ray and Wolff, 1940). Other pain-sensitivestructures are parts of the meninges, the cranialnerves containing sensory fibres, the upper cervicalnerve roots, and scalp muscles. The brain itself, mostof the meninges, the ventricular linings and thechoroid plexuses are not pain sensitive.

Cranial blood vessels are innervated by branches

of the trigeminal, glossopharyngeal, and vagalnerves, and by the upper sensory roots of the spinalcord. Pulsation and stretching of arterial walls orcontraction of skeletal muscle may stimulate periph-eral receptors in these tissues to cause pain. However,people who habitually frown, clench their teeth, orare chronically anxious commonly contract head andneck muscles, but only a small number have head-ache. Similarly, vigorous physical exercise or a hotbath produces arterial vasodilatation but only rarelydoes headache occur in these circumstances. Duringa migraine attack, intracranial vasoconstriction andextracranial vasodilatation occur in sequence orconcurrently. The neurological symptoms precedingthe painful phase are caused by intracranial arterialconstriction and cerebral ischaemia, and pain isrelated to dilatation of the extracranial and someintracranial arteries. Alteration of arteriolar tone isassociated with liberation of vasoactive substancesincluding catecholamines, histamines, serotonin, kin-ins, neuropeptides, and prostaglandins. The pound-ing headache is presumed to be related to neurogenicvasodilatation and a sterile inflammatory responsegenerated by the interaction of these substances withblood vessels and nerves.

Migraine may be a genetic disorder of the brain-stem pain modulation system in which neuro-transmitter turnover, serotonin in particular, is defec-tive (Sicuteri, Anselmi, and Fanciullacci, 1974).Proposed abnormalities in endorphin turnover couldalso explain many of the signs and symptoms ofmigraine (Sicuteri, 1978).Headache classification permits an empiric plan of

therapy. Severe recurrent headaches are usually notcaused by organic disease. About 90% of all head-aches of this type result from aspects of musclecontraction and vascular dilatation. A patient maysuffer from more than one type of headache or havemixed muscle contraction-vascular headache. Forconvenience, headaches can be separated into non-organic (without a recognizable structural lesion) andorganic. The non-organic variety comprises muscle

copyright. on 10 M

ay 2018 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.60.710.841 on 1 D

ecember 1984. D

ownloaded from

R. Atkinson and 0. Appenzeller

contraction, vascular and mixed varieties that have ahumoral or neural basis. The organic headaches havea diagnosable underlying aetiology such as cerebro-vascular disease, sinusitis, dental problems, post-endarterectomy headache, phaeochromocytoma,acute febrile illness, or intracranial space-occupyinglesions.

Muscle contraction headaches

Emotional factors play a significant role in theaetiology of muscle contraction headaches. Theseheadaches often occur in persons who are tense andanxious and in whom somatization of anxiety causesskeletal muscle tension. Muscle contraction headacheis often associated with depression, and the depres-sive signs and symptoms suggest its diagnosis. Thistype of headache may also result from cervicalosteoarthritis, dental malocclusion, and temporo-mandibular joint dysfunction. They are also oftenassociated with so-called myofascial syndromes, trig-ger-point tenderness, and pain in neck and shouldermuscles. Thus, muscle contraction headaches mayoccur in tense and anxious individuals, in those withdepression, and in persons with structural abnormali-ties that trigger muscle contractions.

Vascular headaches

One classification of vascular headaches of themigrainous type includes common, hemiplegic,and ophthalmoplegic varieties. Cluster headache isusually considered vascular, and there are non-migrainous vascular headaches caused by factorssuch as acute febrile illnesses and hangover.Common migraine is not preceded by a well-

defined aura, though prodromal symptoms of severaldays duration may occur prior to an attack, such asfluid retention, or mood alterations including leth-argy, fatigue, irritability and exhilaration. About 80%oof all migraine patients have the common variety.

Classic migraine is preceded by a well-defined aurathat consists of disturbances of cerebral function.Visual disorders are the commonest of these symp-toms but other sensory or motor symptoms such asparaesthesiae, aphasia, and/or sensory hallucinationsmay occur. Fortification spectra or teichopsia or atransient visual sensation of bright, shimmeringcolours are pathognomonic of migraine. The visualsymptoms often include the presence of flashingwavy lines or zigzag patterns or may occur as a blindspot with luminous borders that slowly enlarge andpass across the visual field. This slow progression hasbeen termed 'build-up' and 'march' (Fisher, 1968),and is characteristic of migraine. It is not associatedwith the scintillations of calcarine region ischaemiain posterior cerebral artery thrombosis and embolism

or in transient hypo-perfusion for example after astoop or sudden rise to the upright posture.

Visual displays are likely to be stereotyped for eachperson. In most patients, the visual hallucinations arepresent in both visual fields and are of unformedobjects or shapes. Some patients have illusions ofaltered size, shape or position of structured objects,and in others, frank hallucinations of formed objectsor persons occur. Still others describe a dazzling,bright, white light that obscures vision. The neurolo-gical abnormalities may be present before, during, orafter the headache, and neurological symptoms mayoccur without the headache.

Diagnosis

The diagnosis of headache depends upon a care-fully taken history and neurological examination.Since most headache patients have normal generalphysical and neurological examinations, a detailedhistory is most important for correct diagnosis. Somepatients have more than one kind of headache.Necessary information includes duration of theproblem, description of the aura or other prodromalsymptoms, location of the pain at onset and its nature(throbbing, steady, piercing, knife-like). The lengthof time from onset to maximum severity ofsymptomsand the presence of associated symptoms, such asnausea and vomiting, are important in deciding theroute of administration of medication. Most oralpreparations require 25-30 min for absorption,whereas inhalation results in satisfactory blood levelsalmost immediately. Medication is most effective inmigraine when taken at the first sign ofan attack andis almost of no use if taken when the headache iswell-established. Frequency and duration of attacksand efficacy of medication for individual attacks areimportant in the decision to use prophylactic orabortive therapy only. The time of onset of theheadache may suggest the use of prophylactic medi-cation, for example, if always well-developed onawakening.A careful review of possible precipitating factors

helps to eliminate some. The patient should be asked,'Do you know of anything that always brings on aheadache?' and, similarly, 'Do you know of anythingthat makes the headache go away?'. If the answer tothe latter is alcohol, muscle contraction headache canbe strongly suspected since alcohol ingestion usuallyaggravates migraine.The sex of the headache patient provides a clue to

the headache type. Seventy percent of migrainesufferers are female and 90% of cluster headachepatients are male. Most patients with muscle contrac-tion headache are female.The age that headaches first began is important.

842

copyright. on 10 M

ay 2018 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.60.710.841 on 1 D

ecember 1984. D

ownloaded from

Headache 843

Those beginning in childhood and in the second orthird decades are often vascular. Those beginning inlater life suggest an organic aetiology or depression.If headaches appear for the first time after the age of40 years, alertness to structural causes should beheightened.

In distinguishing migraine from cluster headache,one should remember that migraine may, but usuallydoes not, awaken the patient from night-time sleep.Cluster headaches often awaken patients from sleepand may recur with a clock-like regularity during a24-hr period. Muscle contraction headaches tend tobegin late in the day and to increase in severity untilthe patient retires at night.

Migraine pain often begins unilaterally, but maybe bilateral or switch sides during and betweenattacks. The pain usually starts frontally or tempo-rally and generalizes. Recurrent pain localized in orbehind an eye, always on the same side, and ofexcruciating severity suggests cluster headache. Gen-eralized pain at onset or the 'hatband' distributionsuggests the diagnosis of muscle contraction head-ache.The frequency ofoccurrence varies widely with the

headache type. Migraine occurs at regular or spo-radic intervals throughout a patient's lifetime. Infemales, migraine may be strongly associated withthe menstrual cycle. Migrainous headaches may beabsent during vacations or occur in 'let down' periodssuch as weekends or vacations; it often disappearsduring pregnancy, particularly after the first tri-mester.

Cluster headache is occasionally chronic but mostoften occurs in bouts lasting from weeks to monthsduring which time the patient has one or severalheadaches per 24 hr. Muscle contraction headachesoften occur daily or persist for weeks, months oryears.

Migraine can occur daily. Attacks usually last forhours but may go on for 1 to 2 days. Cluster headachepain characteristically lasts from 15 to 90 min andindividual stabbing pains are of seconds duration.Muscle contraction headaches are present uponawakening only when the patient has gone to bedwith a headache the previous evening. The painusually intensifies as the day progresses and issomewhat relieved by sleep.Migraine pain often begins as a mild, dull ache

that intensifies to throbbing, incapacitating discom-fort. The pulsating quality is characteristic of vascu-lar headaches of both the migrainous and non-migrainous types. Pulsation of the pain is oftenaggravated by stooping. Nausea and vomiting, pho-tophobia, sonophobia, and peripheral vascularchanges causing cold hands and feet are frequentassociated symptoms of migraine. Cluster headachepain is deep, burning, or stabbing. It is often

excruciating; and the patient may be suicidal. Theassociated symptoms of cluster headache are mostoften ipsilateral to the pain and include lacrimation,facial flushing and nasal discharge. The eye is oftenred; conjunctival vessels are dilated, and ptosis andpupillary constriction often occur. Muscle contrac-tion headaches are dull, nagging, and persistent withperiods of heightened pain and severe discomfort.The patient may describe a tight, pressing or squeez-ing sensation or say that his head feels locked in avice. Intermittent piercing pain may also occur.Muscle contraction headaches may be episodic orchronic. The episodic variety occurs up to severaltimes a month and is commonly relieved by non-narcotic analgesics. In contrast, chronic muscle con-traction headache occurs daily and continuously formonths or years, and is not controlled by non-narcotics. Patients with muscle contraction headachemay also have neck and back pain and complain ofdifficulty in sleeping, depressed mood, fatigue andpoor appetite.

Migraine is familial and 70% of patients have anaffected parent. Among women with migraine, theincidence may be as high as 90%o and the affectedparent is usually the mother. If both parents areaffected, offspring have a 70% chance of also havingmigraine. The depressive symptoms associated withmuscle contraction headaches are thought to belearned behaviour and situational rather than gene-tic. Cluster headache is rarely familial.

Precipitating factors for migraine include physicaland emotional stress, fatigue, loss of sleep, missing ameal, and hormonal changes at the time of menstrua-tion. Ingestion of vasoactive substances such asalcohol or foods rich in tyramine (aged cheeses, etc)and phenylethylamine in chocolate may precipitatemigraine in certain subjects. Cluster headaches areoften precipitated in susceptible persons during acluster period by use of vasodilators such as nitrogly-cerine or alcohol. These substances may not initiate aheadache when the patient is not in a cluster period.Fatigue, stress and emotional crises exacerbate painin muscle contraction and vascular headache pa-tients.

Observations of patients with chronic, recurrentheadaches have resulted in categorizing the personal-ities of those with various headache types. Migrainesufferers have been described as compulsive, rigid,and perfectionistic. Though seemingly true in anec-dote, large studies have not substantiated this impres-sion. Cluster patients often have peptic ulcers, areaggressive, frustrated and consume too much alcoholand tobacco. Muscle contraction headache patientshave been said to be dependent with repressedhostility and many take large quantities of differentdrugs which they find difficult to relinquish.

Effort migraine can occur after exhaustive muscu-

copyright. on 10 M

ay 2018 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.60.710.841 on 1 D

ecember 1984. D

ownloaded from

R. Atkinson and 0. Appenzeller

lar activities of any kind, but is rare in well-trainedathletes. It is usually associated with dehydration,excessive heat load and hypoglycaemia, and is morefrequent at unaccustomed altitudes. Serious intra-cranial disease is often feared in this situation, but thecondition is almost always benign and does not recureven with repeated exertion if physical fitness im-proves and dehydration, excessive heat load and highaltitude are avoided. Headache is a prominentsymptom of all forms of mountain sickness that isalso associated with loss of appetite, nausea andvomiting. Migraine can be precipitated by relativelymild head trauma sustained by soccer players and'goggle headache' has been described in swimmerswhose goggles compress the supraorbital nerves andcause a neuropathy. So-called 'turtle headache'occurs in runners who cover their heads with theirshirts after exercise and in others who sleep with theirheads covered presumably from rebreathing expiredair with consequent hypercapnic cerebral vasodilata-tion.

Management of the patient with chronic recurrentheadache

The patient's decision to seek medical help and thephysician's support, reassurance and concern usuallydiminish the frequency and severity of the head-aches. Appropriate laboratory tests may establish thepresence of symptomatic headaches that requiretreatment for an underlying condition. An explana-tion of the process causing the headaches should begiven and the probability of reducing the frequencyand severity of the attacks, but not 'curing' them,should be discussed. Precipitating factors should beidentified and eliminated when possible. Most pa-tients with long headache histories usually know theirheadaches' triggers. Patients with less than twicemonthly attacks are unlikely to have unsuspecteddietary triggers. On the other hand, patients withweekly attacks may need an elimination diet in orderto identify the offending food.Food allergy was studied (Egger et al., 1983) in

children, ages 3-16 years, who had had at least oneheadache per week for the previous 6 months. Theywere placed on an oligoantigenic diet typicallyconsisting of one meat (lamb or chicken), onecarbohydrate (rice or potato), one fruit (banana orapple), one vegetable (brassica), water, and vitaminsupplements for 3 or 4 weeks, depending on thefrequency of headaches. Other foods were reintro-duced to those patients who had no headaches oronly 1 during the past 2 weeks. Those who did notimprove were offered a second oligoantigenic dietwith no foods in common with the first diet. Normaldaily helpings of excluded foods were reintroduced

singly, one a week. If no reaction occurred, thepatient was advised to eat the food regularly. Foodsthat provoked symptoms were withdrawn at the endof the week. Foods were systematically reintroducedunitl the diet was nutritionally and socially accept-able. Of 88 children who completed the study, six didnot improve, 78 recovered completely on the first orsecond oligoantigenic diet, and four improvedgreatly. An interesting associated finding was thefailure of non-specific trigger factors to provokemigraine in responders when the patients were onappropriate diets.

It has long been contended that dietary precipi-tants of migraine are the result of idiosyncrasy orof metabolic rather than allergic factors, contraryto the implications of this study. Special diets maybe dangerous and are socially disruptive. Theyshould be tried only when symptoms are severeand only under experienced medical and dieteticsupervision. If milk and milk products are omittedfor any length of time, calcium supplements shouldbe used, especially in children and postmenopausalwomen.

Anxiety and stress are probably the most frequentprecipitating factors for migraine, cluster, and musclecontraction headaches. Relaxation techniquesthrough biofeedback training, yoga, and/or physicalendurance training are important adjunctive mea-sures to help the patient relax.The use of certain medications may cause exacer-

bation of vascular headaches of the migrainoustype. The most common nowadays are oral contra-ceptives, or other exogenous oestrogens given forpostmenopausal symptoms. Oestrogen use may in-crease the frequency and severity of existing vascularheadaches or may produce daily non-vascular head-aches that are refractory to treatment. In mostpatients with headaches of sufficient severity towarrant medical help, exogenous oestrogens shouldbe discontinued as an initial step in therapy. Patientswho are on oral contraceptives and also have aheadache problem are at increased risk for stroke(Shafey and Scheinberg, 1966), and it may be thataddition of antimigraine medication containing avasoconstrictor increases this risk. After oestrogentherapy has been stopped, a marked decrease infrequency and severity of headaches occurs and thepatient becomes more responsive to usual therapy. Invery rare cases, migraine improves with oestrogentherapy.

Thyroid replacement therapy may aggravate vas-cular headaches, and reserpine (Curzon, Barrye andWilkinson, 1969) and nitroglycerine (Schnitker andSchnitker, 1947) have been reported to triggermigraine. These drugs are important biologicalmarkers for migraine since they do not produceheadaches in non-migrainous persons.

844

copyright. on 10 M

ay 2018 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.60.710.841 on 1 D

ecember 1984. D

ownloaded from

Headache 845

Pharmacotherapy of vascular and mixed headache

Abortive therapy

The ergot alkaloids used alone or combined withcaffeine are the most effective agents for the earlytreatment of acute migraine attacks. The dramaticresponse of a headache to ergotamine therapy isalmost conclusive evidence that it is vascular inorigin. Ergotamine can be given orally, sublingually,by oral inhalation, parenterally or by rectal supposi-tory. Parenteral and oral inhalation routes produceearly effective blood levels. Ergotamine may causenausea, vomiting and epigastric discomfort regard-less of the route of administration, and other adverseside effects including rebound headaches if a patientuses the drug daily. If suppositories are prescribed, asmall dose (1/2 suppository) may be used initially,followed by another half in 10 min and a whole onein 20 min if necessary to control the pain. The patientshould cut the suppository lengthwise (not across)and take no more than 2 per day or 4 per week. Onemechanism of action of ergotamine in migraine isextracranial vasoconstriction. Isometheptene mucateis another vasoconstrictor; it is less effective thanergotamine, but is usually better tolerated.The combination of vasoconstrictors and anti-

emetic agents such as meclizine hydrochloride, 25 or50 mg, metoclopramide 10 or 20 mg, promethazinehydrochloride 12-5-50 mg orally or rectally are oftenmore efficient than either drug used singly. Non-narcotic analgesics such as paracetamol (acetamino-phen) and, particularly, effervescent aspirin plusmetoclopramide may be very effective when takenearly.

Prophylactic therapy of migraineProphylactic daily medication to reduce the fre-

quency and severity of migraine should be consi-dered in patients with more than weekly attacks orwhen abortive therapy is ineffective. Methysergide,now rarely used, produced sustained improvement inover 50% of patients (Curran, Hinterberger andLance, 1967). It is a peripheral serotonin antagonistbut may be helpful in migraine because of itsvasoconstrictor properties affecting the external caro-tid circulation. The dose is from 2 to 6 mg per day. Itis usually not effective for treatment of an acutemigraine attack. Strictly nocturnal headache may becontrolled with bedtime medication only (Curran etal., 1967). Continuous therapy with methysergideshould not exceed 3 months. A drug-free interval of 3to 4 weeks after each 3-month course of treatmentmust be used and the dose should be reducedgradually during the last 2 to 3 weeks of eachtreatment course to avoid 'headache rebound'. Sideeffects are frequent and its long-term, uninterrupted

use may result in retroperitoneal, pleuropulmonaryand cardiac valve fibrosis. It should not be used inpregnancy, in patients with arterial or venous vascu-lar problems, hypertension, infection or liver or renalfailure.

Amitriptyline is useful in the prophylaxis ofmigraine, independent of its antidepressant action(Gomersall and Stuart, 1973; Couch, Ziegler andHassanein, 1976). The effective dose varies from 10to 150 mg given before bedtime. Amitriptyline is bestintroduced gradually to avoid excessive sedation.Doxepin in doses of 25 to 150 mg at night is often alsouseful and may cause fewer side effects than amitrip-tyline.Cyproheptadine is a central serotonin-agonist

(Haigler and Aghajanian, 1977) with platelet anti-aggregating properties (Ambrus, Ambrus and Thur-ber, 1977) that is sometimes useful in migraineprophylaxis. The dose varies from 8 to 24 mg a day.This is given once daily in the smaller doses and twicedaily in the larger ones. Sedation and weight gain arethe commonest side effects.

Propranolol 40 to 320 mg per day in three divideddoses is of established value (Forssman et al., 1976) inthe prophylaxis of migraine. Its mechanism of actionis uncertain, but may be related to central effectsrather than peripheral beta-blockade (Stensrud andSjaastad, 1976). The average effective dose is 120 mgdaily. Congestive heart failure, asthma and renalinsufficiency are absolute contraindications to pro-pranolol therapy.The slow calcium channel blockers have been

shown to be effective prophylactic treatment for theprodromes and headache of classic migraine and forcontrol of common migraine and cluster. There is noeffect on muscle contraction headaches or on themuscle contraction component of combined head-aches. Nimodipine (60 and 120 mg/day) was mosteffective with the fewest side effects. Beneficial effectswere apparent 2-4 weeks after starting therapy incommon and classic migraine. Relief from clusterheadaches began after 4 weeks of treatment withnimodipine. Nifedipine (mean dose 60 mg/day) andverapamil (mean dose 240 mg/day) also reduced themonthly frequency of all three types of vascularheadaches, but were less effective than nimodipine incontrolling migraine. Comparable delays after initia-tion of therapy before onset of headache reliefoccurred with all three drugs. The most frequent sideeffects were peripheral vascular (oedema, flushing,dizziness) and skin rash. None were incapacitatingand all promptly reversed when the medication wasdiscontinued (Meyer and Hardenberg, 1983).The treatment of cluster headache is essentially the

same as the prophylactic treatment of migraine witha few additions. The daily use of ergotamine may beeffective and warranted in this condition, keeping in

copyright. on 10 M

ay 2018 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.60.710.841 on 1 D

ecember 1984. D

ownloaded from

846 R. Atkinson and 0. Appenzeller

mind that the exposure should be relatively short. Itis sometimes used 1-3 times daily, depending onheadache frequency, for 5 days per week with 2 daysoff medication during which time the patient assesseswhether or not he still needs the medication.Oxygen inhaled at 7-10 litres per min is sometimes

effective in aborting an acute cluster headache andmay reduce frequency of headaches during a clusterperiod when used for 5 to 10 min 3 to 4 times a day.Vigorous physical exertion, such as running in place,at the first sign of a headache may decrease severity,or abort a cluster headache.

Prednisone in doses from 20 mg every other day to40 mg/day suppressed cluster cycles in 75% ofpatients (Kudrow, 1978). The drug is usually givenfor 2 weeks, and then tapered and discontinued.

Lithium carbonate is often effective in treatingcluster headaches, especially chronic non-cyclic at-tacks (Kudrow, 1977). The effective dose has rangedfrom 600 to 900 mg/day. Favourable responsesoccurred within the first week of therapy in themajority of patients. Its mechanism of action incluster headache is not known, but the site issuspected to be central. Nausea, vomiting, diarrhoea,tremor, blurred vision and unsteady gait are thecommonest side effects and these are dose related.The drug is often effective at plasma levels wellbelow those required for psychiatric purposes.The physician's attitude and communications of

support and concern are most important in treatingthe patient with muscle contraction headache. In-struction in relaxation techniques and stress manage-ment, combined with tricyclic antidepressant drugtherapy, are often useful to these patients.

Editor's note: A newer anti-serotonin agent, pizotifen, is availablefor anti-migraine prophylaxis in the U.K.

ReferencesAMBRUS, J.L., AMBRUS, C.M. & THURBER, L. (1977) Study of

platelet aggregation in vivo. Effect of the antiserotonin agentcyproheptadine. Journal of Medicine, 8, 317.

BASBAUM, A.I. & FIELDS, H.L. (1978) Endogenus pain controlmechanisms: review and hypothesis. Annals of Neurology, 4,451.

BLAU, J.N. & DEXTER, S.L. (1981) The site of pain origin duringmigraine headaches. In: Progress in Migraine Research 1. (Eds.R.F. Clifford and K. J. Zilkha), Ch. 3, p. 13. The Pitman Press,Bath, Great Britain.

COUCH, J.R., ZIEGLER, D.K. & HASSANEIN, R.S. (1976) Amitrip-yline in the prophylaxis of migraine. Neurology, 26, 121.

CURRAN, D.A., HINTERBERGER, H. & LANCE, J.W. (1967) Methys-ergide. Research and Clinical Studies in Headache, 1, 74.

CURZON, G., BARRYE, M. & WILKINSON, M.I.P. (1969) Relation-ships between headache and amine changes after administrationof reserpine to migrainous patients. Journal of Neurology, Neuro-surgery and Psychiatry, 32, 555.

EGGER, J., CARTER, C.M., WILSON, J., TURNER, M.W. & SOOTHILL,J.F. (1983) Is migraine food allergy? A double-blind controlledtrial of oligoantigenic diet treatment. Lancet, ii, 865.

FISHER, C.M. (1968) Migraine accompaniments versus arteriosclero-tic ischemia. Transactions of the American Neurological Associa-tion, 93, 21 1.

FORSSMAN, B., HENRICKSSON, K.G., JOHANNSON, V., LINDVALL, L.& LUNDIN, H. (1976) Propranolol for migraine prophylaxis.Headache, 16, 238.

GOMERSALL, J.D. & STUART, A. (1973) Amitriptyline in migraineprophylaxis. Journal of Neurology, Neurosurgery and Psychiatry,36, 684.

HAIGLER, H.J. & AGHAJANIAN, G.K. (1977) Serotonin receptors inthe brain. Federation Proceedings, 36, 2159.

KUDROW, L. (1977) Lithium prophylaxis for chronic cluster head-ache. Headache, 17, 15.

KUDROW, L. (1978) Comparative results of prednisone, methyser-gide, and lithium therapy in cluster headache. In: CurrentConcepts in Migraine Research. (Ed. R. Green), p. 159. RavenPress, New York.

MEYER, J.S. & HARDENBERG, J. (1983) Clinical effectiveness ofcalcium entry blockers in prophylactic treatment of migraine andcluster headaches. Headache, 23, 266.

RAY, B.S. & WOLFF, H.G. (1940) Experimental studies on headache.Pain sensitive structures of the head and their significance inheadache. Archives of Surgery, 41, 813.

SCHNITKER, M.T. & SCHNITKER, M.A. (1947) A clinical test formigraine. Journal of the American Medical Association, 135, 89.

SHAFEY, G. & SCHEINBERG, P. (1966) Neurological syndromeoccurring in patients receiving steroids (oral contraceptives).Neurology, 16, 205.

SHAND, D.G. (1974) Individualization of propranolol therapy.Medical Clinics of North America, 58, 1063.

SICUTERI, F. (1978) Endorphins, opiate receptors and migraineheadache. Headache, 17, 253.

SICUTERI, F., ANSELMI, B. & FANCIULLACCI, M. (1974) Theserotonin theory of migraine. In: Advances in Neurology, Vol. 4.(Ed. J. J. Bonica), p. 384. Raven Press, New York.

STENSRUD, P. & SJAASTAD, 0. (1976) Short-term clinical trial ofpropranolol in racemic form (Inderal), d-propranolol and placeboin migraine. Acta Neurologica Scandinavica, 53, 229.

copyright. on 10 M

ay 2018 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.60.710.841 on 1 D

ecember 1984. D

ownloaded from