14ournal ofNeurology, Neurosurgery, and Psychiatry 1994;57:134-143

NEUROLOGICAL MANAGEMENT

Headache

JM S Pearce

Current issues: general managementMORBIDITY: THE BURDEN OF HEADACHE TOPATIENT AND SOCIETYMorbidity due to headaches is a major prob-lem. Tension headaches outnumber migraineby 5:1. Table 1 compares the more commoncauses of chronic or recurrent headaches indifferent age groups.A recent study from Minnesota, United

States, confirms the overall incidence ofmigraine to be about 10% in the population,with a marked variation related to age and ahigher incidence in women in a defined popu-lation.' From 6400 patient records, 629 resi-dents fulfilled the International HeadacheSociety's (IHS) 1988 criteria2 for migrainebetween 1979 and 1981. The overall age-adjusted incidence was 137 and 294 per 100000 person years for males and femalesrespectively. The highest incidence was inwomen aged 20-24 years (689 per 100 000person years), whereas in boys aged 10-14years, the incidence was 246 per 100 000person years.

Based on a study of 1000 people using theInternational Headache Society's criteria, theoverall lifetime prevalence of classic migrainewas 5%, with a female to male ratio of 2:1.3The overall lifetime prevalence of commonmigraine was 8%, with a female to male ratioof 7: 1. Women were more likely to have com-mon than classic migraine. Neither classic norcommon varieties correlated with age, but inboth types the most conspicuous precipitatingfactor was stress and mental tension.The rate of consultation is a concern for

those planning medical resources. Amongsubjects with classic and common migraine,50% and 62%, respectively, consult theirgeneral practitioner because of migraine at

Table 1 Common headaches at different ages

3-16 (years) 17-60 (years) 60+ (years)

Migraine Migraine Referred from neckTension headache Tension headache Cranial arteritisPsychogenic/fatigue Cluster headache Paget's disease of the skullPost-traumatic Post-traumatic GlaucomaOccasionally Cranial and dural tumours Cranial and dural tumours.Tumours: posterior fossa,

intraventricular Cerebral tumours including Cerebral tumours includingabscess and subdural abscess and subduralhaematoma haematoma

Depression Rare cluster headacheReferred from neck Paget's disease of the skullPaget's disease of the skull Post-herpetic and cranial

neuralgiasPost-herpetic and cranial Post-traumatic

neuralgias*Continuing tension headacheContinuing migraine

*Uncommon presentation at this age.

some time in their lives. Patients frequentlyattend for medical help, and lose much timefrom work which is often unrecorded. In arandom sample of 740 subjects, aged 25-64years living in the Copenhagen County,Denmark, 119 had migraine and 578 hadtension headache (1 :5).4 Among subjects withmigraine 56% had consulted their generalpractitioner in the previous year for migraine;among subjects with tension headache 16%had had consultations. Specialists had beenconsulted by 16% of migraine sufferers andby 4% of subjects with tension-type head-ache. Less than 3% of all patients studied hadrequired hospital admission and laboratoryinvestigations for headache. Half the migrainesufferers and 83% of subjects with tension-type headache in the previous year had takendrug therapy. Thus migraine and tensionheadaches can be seen as potent sources ofdemand for medical attention and for theconsumption of drugs.

In the Danish study, 43% of employedmigraine sufferers and 12% of employed sub-jects with tension-type headache had lostworking time in the preceding year. The totalloss of work days due to migraine was esti-mated at 270 per 1000 persons per year; fortension headache the corresponding figurewas 820. Women consult a doctor more oftenthan men, but there is no sex difference inabsenteeism.

PRINCIPLE OBJECTIVES AND INSTRUMENTS OFMANAGEMENTIf a patient with headache has cranial arteritisor a brain tumour, then all sociological inter-vention must take second place to promptdiagnosis and treatment. Over 95% ofpatients seen in a general practitioner'ssurgery or hospital clinic, however, have ten-sion headache, migraine, or atypical headachewithout a structural lesion. For such patientswhose symptoms are strongly influenced bysocial, personal, and family problems, severaltherapeutic commodities are needed.

For effective therapy doctors need to beaccurate in diagnosis, clear in the direction oftreatment, honest when they do not knowhow or why a symptom has developed orchanged-yet reassuring. They should beunderstanding but not dismissive; and theyshould be prepared to see patients repeatedlyuntil the headaches are controlled, and espe-cially to continue to comfort and supportwhen they are not cured. Such paragons arenot ubiquitous. Painstaking efforts to com-municate with patients from the start savemany subsequent consultations.

304 Beverley Road,Anlaby, HullHU10 7BG, UKJM S PearceCorrespondence to:Dr Pearce

134

on 30 April 2019 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.2.134 on 1 F

ebruary 1994. Dow

nloaded from

Headache

Political and financial constraints havefailed to provide for the necessary numberof doctor hours to cater for this ideal. Inreality, this has been matched by the never-ending expectations of patients for the latesttechnical facilities, for a specialist's personalattentions (no longer is a consultant or evena neurologist adequate; it has to be a"headache expert"), and for clinical infallibil-ity-or else, litigation may result.

Such issues of time and expertise have, bydefault, produced the notion that nurses,social workers, osteopaths, homeopaths, andpsychologists, among others, might providefor these patients. Regrettably, they cannot.Although, under medical direction, they canhelp, they are no substitute for a doctor'straining and knowledge. Physicians areunwise to abrogate their traditional role bydelegating patients' management to let them-selves off the hook. Such fashionable trendsdenigrate the physicians' role, but much moreimportantly, diminish their therapeutic effec-tiveness.The armamentarium is, however, limited.

The frustration of implacable patients isreflected in their search for "alternative thera-pies". Medical science assesses its toolsby rigorous analysis and scientific trials.Medicine should not be daunted by practi-tioners of alternative medicine since, withrare exceptions, they have so far failed to sub-ject their unbounded claims of therapeutictriumphs to scientific scrutiny. When they doso, and when their methods produce vali-dated benefit, physicians should welcome anduse them.

General practitioners provide primary diag-nostic facilities which essentially serve to sep-arate those with major pathology (tumours,glaucoma, arteritis) from functional head-aches. Many of the latter group are dealt withadequately by history taking, examination,reassurance, and selected analgesics andantimigraine drugs. Patients posing diagnos-tic problems and those with refractory head-aches should be referred to a neurologist.

"Headache clinics" are convenient, andwell suited to trials and research; but, to be ofbenefit to patients they should, ideally, beavailable on a daily basis, staffed by one ormore experienced physicians to provide conti-nuity of service: these ideals are not generallypracticable.

AuditThis concept, woolly in definition and ofteninadequate in application, is not easy toimpose on the management of patients withheadaches. Patients and their general practi-tioners can be questioned about their satis-faction with the service; improvement inheadache scores would need several applica-tions per patient over a long period of time todemonstrate efficacy above the substantialplacebo effect of seeing a new doctor, receiv-ing the investigations that patients perceive asreassuring, and being offered new treatments.Without implementation of results of audit("closing the loop") and further appraisal, the

exercise is meaningless. To instigate stan-dards of good practice, yardsticks need to beestablished and agreed.

In a district or regional service, the assess-ment of wrong diagnoses-for example,missed tumours, subarachnoid haemorrhage,or missed cranial arteritis, is feasible.Similarly errors of diagnosis, choice of first-line treatment and patterns of secondaryreferral can be appraised in casualty depart-ments and in general practitioner referrals.

The headache patientThis account is restricted to headache andexcludes the related but separate facial painsand neuralgias. The clinical approach restsheavily on a detailed analysis of symptomsand a thorough general and neurologicalexamination. Selective investigations areapplicable to only a small number of patientswho require neurological referral (table 2).Headaches may be acute, usually signifyingmeningeal irritation or, rarely, raised intracra-nial pressure (ICP), recurrent as in migraine,or chronic as in tension headaches.

HEADACHE AS A SYMPTOM OF INTRACRANIALDISEASEFew headaches fail to evoke some anxiety,which can distort, disguise, or magnify theprimary clinical features. Confronted by apatient with headaches, the first clinicalresponsibility is to exclude a structural ordynamic cause. Any expanding mass-tumour, abscess, or haematoma-can causeraised pressure.

Headaches of abrupt onset may signifytrauma, spontaneous intracranial haemor-rhage, hydrocephalus, or acute meningealirritation at any age; the elderly are notimmune. The most common cause is acutemeningeal irritation due to subarachnoidhaemorrhage or to meningitis-bacterial orviral, (rarely HIV, fungal, or malignant). Anabrupt onset, fever, neck stiffness, andKernig's sign accompany the obvious severepain, vomiting, and photophobia. Hospitalreferral for CSF examination is mandatory ifhaemorrhage or infection is considered. CT(or MRI) should be the first investigation toexclude a mass lesion, haematoma, or hydro-cephalus, each of which contraindicates lum-bar puncture. Subarachnoid haemorrhagemay be shown, but can be missed by CT inup to 20% of patients on admission.When a mass lesion is excluded, lumbar

puncture may be necessary to examine the

Table 2 Indications for specialist referral andinvestigations

* Sudden onset, new headache* Atypical symptoms or signs suggesting organic pathology* Abnormal signs-for example, papilloedema to suggest

raised intracranial pressure; a red, tender scalp vessel toindicate cranial arteritis

* An unremitting course, unresponsive to conventionaltreatment

* The advent of progressive physical signs

135 on 30 A

pril 2019 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.2.134 on 1 February 1994. D

ownloaded from

Pearce

CSF for subarachnoid bleeding and menin-gitis. Acute migraine and tension headacheuncommonly produce a meningitic picture,diagnosable only by exclusion. Acuteheadache should be distinguished from thecommon and totally benign "exploding headsyndrome" in which patients are alarmed by asudden, momentary, very loud noise in thetwilight stage of sleep.5

Raised-pressure headacheThis is an aching, throbbing headache aggra-vated by alcohol, exertion, and by coughingor straining. These features suggest a vascularor hydrodynamic mechanism which maybe caused by a tumour or hydrocephalus.Similar symptoms, however, are also com-mon in migraine and other vascular pains.The location is non-specific, although when aprogressive headache radiates to the neck,tonsillar coning is imminent. The headacheis: (a) worse in the morning and may wakenthe patient from sleep; (b) aggravated by sit-ting up or standing and relieved by lyingdown; (c) aggravated by coughing, straining,and vomiting; (d) relieved by aspirin or parac-etamol in the early stages (in contrast topsychogenic headache); (e) associated withvomiting and eventually by papilloedema andprogressing focal signs. By the stage of stuporand hemiplegia with a dilated (Hutchinson's)pupil, diagnosis has been delayed too long.

Chronic and recurrent headachesREFERRED PAINThe orbits, paranasal sinuses, cervical spine,mediastinum, and teeth can cause painreferred via branches of the trigeminal nerveto the forehead and temple. Sinusitis andtoothache are commonplace examples, butotitis media, glaucoma, orbital cellulitis, andcavernous sinus thrombosis may producereferred frontotemporal pain. Secondaryinvoluntary contraction of scalp and facialmuscles further complicate the picture, caus-ing a secondary generalised "tension head-ache" which may obscure the primary source.Primary care physicians can treat acute infec-tion by antibiotics but, when cranial spread issuspected, patients shouldbe immediatelyreferred to the neurological/neurosurgicalcentre.

TENSION HEADACHETension headache is the most common ofhuman complaints, constituting 70% of refer-rals to a "headache clinic". It is often a short-lived complaint with an obvious precedingcause: overwork, lack of sleep, or an emo-tional crisis. This is benign and often desig-nated by patients as "my normal headaches".

Current systems classify the most commonrecurring headaches as either migraine or ten-sion type. This traditional approach can bequestioned. Some suggest that these twoheadache patterns are but different expres-sions of the same pathophysiological process,having overlapping symptomatic presenta-

tions with certain features emphasised to agreater or lesser extent. Additionally, thesame therapies have been shown to be effec-tive for patients in either headache group.6 Analternative continuum classification modelhas been suggested, as there is an undoubtedoverlap between common migraine and ten-sion headache, although their common co-existence has added to the confusion.By contrast, Solomon and Lipton propose

that the diagnosis of migraine without aura(common migraine) is warranted if any two ofthe following symptoms are present: unilat-eral site, throbbing quality, nausea, photo-phobia, or phonophobia.7 These criteria arederived from a study comparing the featuresof 100 patients with migraine without auraand 100 patients with chronic daily ("ten-sion") headache. The authors' proposed cri-teria for the diagnosis of migraine withoutaura were highly sensitive and adequatelyspecific in discriminating the two groups.There are also differences in laboratorydata. Binding of 3H-labelled imipramine toplatelets was measured and a significantreduction was found in migraine comparedwith controls but not in tension headache. Inmigraine, there was no significant relationshipbetween imipramine binding and depressionor anxiety score suggesting that the reductionin platelet imipramine binding is a concomi-tant of migraine itself.8 A significant reduc-tion in peripheral blood mononuclear cellfl-endorphin concentrations was observed inmigraine patients with and without aura, butnot in patients with tension-type headache.9Adult and childhood migraineurs withoutaura have an increased amplitude of the con-tingent negative variation between attacks.'0Likewise vascular phenomena, well describedin migraine, contrast with transcranialDoppler ultrasound of blood flow velocitiesin chronic tension headache which show nosignificant differences from controls.The case for classic migraine as a separate

entity is even more persuasive. The teenagerwith infrequent but prostrating attacks of photo-psiae, dysphasia, vomiting, and hemicrania, isat least clinically distinct from the angst-ridden 40-year-old patient with continuous,daily, vertex, pressure headache for 20 yearswhich does not prevent remunerative work.

Acute tension headacheThis type of headache rarely presents as anemergency. When it does, the headache hasincreased over a few hours, but has becomevery severe, simulating subarachnoid bleed-ing. Lumbar puncture may be necessary toexclude meningeal irritation by blood orinfection. More often the emotional basis isobvious and recovery ensues quickly afterreassurance, analgesia, and sedation. Liaisonwith the general practitioner should lead tocontinued support and prompt return to work.

Chronic tension headache (chronic dailyheadache)More common than the acute syndrome, painis diffusely felt all over the head, often located

136 on 30 A

pril 2019 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.2.134 on 1 February 1994. D

ownloaded from

Headache

on the vertex, or may start in the forehead orin the neck. Primary tension headache is psy-chogenic; its mechanisms are not whollyunderstood.'1 It is commonly bilateral, butmay be unilateral. Patients characteristicallycomplain of pressure, a feeling of tightness, ora heavy weight pressing on the crown. "Atight band like a skullcap" or "as if a clamp orvice was squeezing my head" are commondescriptions. Many say "it is not really a pain,but a pressure". The symptoms may alsoseem to derive from inside the cranium: "as ifmy head is bursting" or "about to explode".A "creeping" sensation (formication) may befelt under the scalp, or a sense of sharp knivesor burning hot needles driven in, may berelated.

Tension headache occurs daily, worse inthe evenings. Visual disturbance, photopho-bia, and vomiting seldom occur. Mostpatients continue their normal work.Symptoms continue for years without evidentdeterioration of general health. Symptoms areworse when the patient is tired or under pres-sure of work, or domestic stresses.'2 Most suf-ferers have insight, so that a carefully takenhistory clarifies both the diagnosis and theaggravating circumstances; many are emo-tional and anxious with fears of braintumours, hypertension or 'clots in the brain'.I enquire specifically about fears of seriousbrain disease, whether the fears are voiced bythe patient or not. Illnesses seen on televisionprogrammes, like maladies of relatives andacquaintances, are often "contagious"; thesetoo need careful appraisal. Thorough clinicalexamination is of the utmost therapeuticvalue and provides a rational basis for effec-tive reassurance that is denied to the psychol-ogist or counsellor.

Treatment is most effective when the his-tory is short. To cure such headaches aftermany years is a daunting and often unsuc-cessful task.'3 An important step is to enquireabout the events that determined the onset.These are often forgotten, or perhaps sup-pressed, yet repeated enquiry at subsequentconsultations will often unravel the apparentmystery, the consequent knowledge beingsufficient to explain the cause to the patient.Sensitive patients with fragile personalitiesmay be unable to cope with life's stresses andunconsciously use headaches to escaperesponsibilities with which they can't cope.Sedatives, tranquilizers, and tension-relievingdrugs are of limited value unless the psycho-logical issues are adequately handled. Glibreassurance will not eradicate headache iffundamental psychological problems areunresolved. When the history is short and if acause is exposed, explanation and reassurancemay suffice.

Patients often abuse analgesics whichaggravate the situation, but the abuse, withpersuasion, can be reversed with dramaticbenefit (see below). More often, when dailypain has persisted for years, the prognosis ispoor, but short courses of benzodiazepines oramitriptyline may be helpful. Supportive psy-chotherapy in liaison with a practice nurse or

psychotherapist may help: much depends onthe quality, experience, and good sense of theindividual available.

Latent depression presenting as tensionheadache is easily overlooked. Early morninginsomnia, negativism, anhedonia, guilt, anddiurnal mood swings are suggestive. Theheadache is worse in the morning (resemblingthat of raised pressure), and a cause for themisery is not always apparent. Full doses oftricyclic antidepressants or fluoxetine areneeded.46 The prognosis for depression isoften good.

MIGRAiNEClassic migraineClassic migraine is synonymous withmigraine with aura and occurs in 20% ofpatients. It is a paroxysmal disorder withheadaches, often unilateral at the onset, asso-ciated with nausea, anorexia, and often vom-iting; it is preceded or accompanied by visual,sensory, motor, and mood disturbances andis often familial.

Common migraineThis is synonymous with migraine withoutaura and occurs in 75% of patients. The termrefers to similar paroxysmal headaches with-out the aura. Both types of attack may occurat various times in the same patient. It iscommon for migraineurs to have tensionheadaches between their migraines: theseshould be identified to prevent misdirectedtreatment. Daily headaches are nevermigrainous.

Migraine variantsThese are hemiplegic, basilar, and ophthal-moplegic, migraine sine cephalgia, etc, andoccur in less than 5% of patients, usuallyrequiring a neurologist's appraisal and some-times brain imaging.

Natural history and managementIn childhood, attacks begin before the age of10 years in a third of patients. They may beoverlooked if the child is unable to describeheadaches and strange visual or sensory expe-riences clearly. They are also concealed bylabels of "bilious attacks" or "periodic syn-drome". Diagnosis can be difficult as affectedchildren may simply appear pale, ill, limp,and inert, complaining of poorly localisedabdominal pain. Headache is usually present,vomiting is common, and there may be afever of up to 38-5°C so that the suspicion ofappendicitis or mesenteric adenitis oftenarises. It should be remembered that over80% of migraineurs have their first attacksbefore the age of 30 years, and the diagnosisshould therefore be viewed with suspicion ifonset is after the age of 40, although anincreased frequency of attacks at themenopause is common. Some subjects haveonly a few attacks in a lifetime but most haveseveral attacks each year. Promises of remis-sion at the menopause are often ill-founded,though attacks tend to lessen after the age of50 years. Remission occurs in 70% of

137 on 30 A

pril 2019 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.2.134 on 1 February 1994. D

ownloaded from

Pearce

pregnancies. It is well recognised that exacer-bation, or complicated migraine with infarc-tion, may result from oestrogen-containingcontraceptives.'4Many attacks in adults end gradually after

24 to 48 hours but attacks in children oftenlast only two to six hours. In later life head-aches may disappear completely, attacks pre-senting with teichopsiae and no headache(migraine sine cephalgia). The mechanism isuncertain. Indeed, the apparent onset ofmigraine in the elderly implies atheroscle-rotic, thromboembolic disease. Food idiosyn-crasies, food allergies, red wines, specificdietary amines, and omission of food areoccasional precipitants. Emotional stress andfatigue is the principal aggravating factor butis not causal.

Migraine is determined by a primarily cere-bral (neural) mechanism with a fluctuatingthreshold which determines the timing andpattern of attacks.'5 A neural trigger activatesthe trigeminovascular reflex, releasing vaso-genic amines from blood vessel walls accom-panied by their painful, pulsatile distension.The cerebral mechanism responds to mood,emotions, tiredness, relaxation, hormonalchanges, and to bright lights, and noise. Itsthreshold is susceptible to hypothalamic func-tion which, in tum, is modulated by seasonalpattems, diurnal and biological clocks, andby hormonal factors and coitus.'6 Personalityand variations of mood and behaviour alsoinfluence the pattem of attacks, remissions,and treatment.

SYMPTOMATIC TREATMENTAssessment of the patient's habits, work, per-sonality, and stresses is important. Suitablytrained clinic nurses can assist in the time-consuming elucidation of these factors inselected cases. Known precipitants should besought, with the aid of a diary and should,when possible, be eliminated (table 3). Thepatient's recognition of stressful pattems andthe acceptance of a benign disorder willachieve some benefit. Good rapport enhancesreassurance and facilitates the markedplacebo effect of all therapy which often con-founds the analysis of drug trials.The aims of treatment are the control of

symptoms, and the prevention or reduction ofattacks. Many prophylactic drugs act by cen-tral serotonin (5HT,) antagonism,'7 whereascontrol of an attack relies on constriction ofcranial vessels mediated by a-adrenergic or5HT, receptors.'8

Table 3 Common precipitants of migraine

* Fatigue, overwork, travel* Relaxation after stress-holiday and Saturday morning

headache* Bright lights, discos* Sleep excess or shortage-Sunday morning headache* Missing meals* Rare dietary sensitivity* Alcohol, red wines* Menstruation* Exercise-related vascular headaches: footballer's migraine,

coital cephalgia

AnalgesicsRest, dark, and quiet, where practicable, aresupplemented by simple analgesics (paraceta-mol 1 g or soluble aspirin 0 6 g) or non-steroidal anti-inflammatory drugs such asnaproxen 500 mg. The addition of caffeineand spasmolytics add to expense but not tobenefit. Codeine 15-30 mg may enhance painrelief. Analgesics should be taken immedi-ately the attack begins, then repeated four tosix hourly as needed. Absorption is improvedby ingestion with metoclopramide 10 mg ordomperidone 10-20 mg. If vomiting is severe,suppositories of domperidone, prochlorper-azine, or chlorpromazine are valuable.Analgesic abuse can cause headaches.

ErgotamineErgotamine (or dihydroergotamine'9) is aneffective remedy for acute attacks in about50% of cases and should be tried before moreexpensive agents. Ergotamine has no place asa prophylactic, however, and, when overused,can lead to habituation with ergotamine-dependent headache, similar to chronic anal-gesic-dependent headaches which mimicmigraine status. When suspected, the drugsshould be withdrawn, often in hospital undershort-term sedation.

Ergotamine is an a-adrenergic agonistwith potent 5HT1 receptor affinity, a potentvasoconstrictor. Absorption is erratic, oraldoses often producing sub-therapeutic bloodlevels. Suppositories (1-2 mg), inhalation(0-36 mg), or sublingual (1-2 mg) are themost effective acceptable routes, but injec-tions are not generally tolerated. If two dosesat intervals of two to six hours are ineffective,no more should be given for that attack. Ifa patient is taking ergotamine more thantwice each week, there is a major riskof habituation. Many patients experiencevague malaise, nausea, and cramps withergotamine, but coronary, cerebral, andlimb (St Anthony's fire) ischaemia are rare,but well-proven, hazards. Vascular claudi-cation, angina, and pregnancy are contra-indications.

SumamiptanThis is a specific and selective agonist of5HT1 receptors on cranial blood vessels caus-ing vasoconstriction.Y It has negligible effectson other receptors. Sumatriptan does notpenetrate the blood-brain barrier and hasno CNS effects. Intravenous sumatriptandoes not change regional cerebral blood flow;it constricts the carotid vascular bed, but hasno effect on pial vessels in cats. It has aplasma half-life of two hours.The subcutaneous preparation (6 mg)

gives relief of headache in 77% patients at 60minutes, and in 83% at two hours with corre-sponding improvement in nausea, vomiting,and photophobia.2' In cluster headache too, itis effective with relief of symptoms at 15 min-utes in 74% compared with 26% givenplacebo.22

Oral medication (100 mg) provides reliefin about 70% of attacks within two hours.23

138 on 30 A

pril 2019 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.2.134 on 1 February 1994. D

ownloaded from

Headache

Second and third attacks respond as well asthe first. Comparative trials have shown slightbut significant superiority to aspirin 900 mgplus metoclopramide 10 mg. Recurrence ofheadache occurs within 48 hours in 42% ofpatients.24 In comparisons of sumatriptan andCafergot, oral sumatriptan 100 mg relievedheadache after two hours in 66% comparedwith 48% of those given oral Cafergot.2' Thepoor oral absorption of ergotamine, however,was a serious limitation of this study.

Another study showed a good response tooral sumatriptan in 51% of patients at twohours compared with 9% given placebo; res-cue medication was needed in 41% of thesumatriptan group, but in 88% of the placebogroup.26 A total of 39%, however, had recur-rent headache within 24 hours-a significant"rebound effect" which may relate to the nat-ural history of migraine, or to a pharmacolog-ical effect.

In over 100 000 treated attacks, toxicityhas not been a problem.27 There are nowproven cardiovascular ischaemic sequelaewith angina, infarction28 and ventriculararrhythmias, but these are rare if establishedischaemic heart disease, prinzmetal angina,and arrhythmias are respected as contraindi-cations. Vague non-ischaemic chest discom-fort occurs in some patients. Nausea,vomiting, and tingling, deemed mild andtransient, occur in 38% patients within fourhours. Rebound in a third of patientsresponds well to further doses, but com-pounds the pressing issue of cost.Sumatriptan is an effective, safe, and promptremedy, suppressing all the symptoms, notheadache alone.29 It works in 70% of suffer-ers, however-not in every patient. Thepresent high cost limits its use.Good communication with family doctors,

advice by clinic or practice nurses on the useof self-injectors and inhalers and the fre-quency of tablets are important. Audit of theresults of treatment by completing headachecharts can be invaluable, but patients shouldnot overuse diary cards which can engenderexcessive introspection and neurosis.

PROPHYIAXISProphylaxis should be considered if attacksoccur more often than twice each month.Non-pharmacological techniques are success-ful in certain subjects, but claims for theirgeneral application should be viewed withscepticism. Current data support no signifi-cant difference in the efficacy of hypnosis,biofeedback, and relaxation training.Prophylactics aim to reduce the frequency,but none is a panacea. They should be givenfor three to six months, then reassessed. Inpatients with exacerbations related to stress,amitriptyline 100-150 mg, at night, intro-duced gradually, is often effective. Which ofits actions-sedative, antidepressant, anti-serotoninergic, or calcium channel block-ade-is implicated, is not known. fl-blockerswithout intrinsic sympathomimetic activitypropranolol,'30 atenolol, and metoprolol-reduce the frequency in about 60% of cases,

and are most effective in the tense or hyper-tensive subject with tachycardia and overtphysical signs of autonomic "overdrive".Their action is probably central.

Serotonin inhibitors are valuable in60-70% of patients. Cyproheptadine 4 mgthree times daily inhibits calcium channels,and serotonin and histaminergic activity.Pizotifen 0 5 mg three times daily or 1-5 mgat night is moderately effective and free ofhazards other than sedation and weightgain.3' Methysergide 1-2 mg three times dailyis the most effective drug in this group butshould be used under hospital supervision incourses not exceeding three to four months.Pleural, pericardial, and retroperitoneal fibro-sis are rare but serious, side effects that resultfrom prolonged use; they usually regresswhen the drug is stopped. Myocardial andperipheral vascular ischaemia are uncommoncomplications. Although calcium antagonistscan cause a vasodilator-type headache, some-for example, flunarizine and verapamil, havebeen established as useful prophylacticdrugs.32

Menstrual migraineMenstrual migraine, defined as occurringexclusively within 48 hours of menstruation,is uncommon but may be relieved by suma-triptan, or oestrogen patches or implants.Migraine between periods but worse withmenstruation, is common (35%) but resistantto diuretics and hormonal manipulation.

CLUSTER HEADACHESynonyms are migrainous neuralgia, Harris'ssyndrome, and Horton's syndrome.

Often misdiagnosed, this is a distinct syn-drome separable from migraine; it predomi-nantly affects men (male:female ratio 10:1).It begins at any age, most often 20 to 50years, and is manifest as daily bouts of unilat-eral headache of great severity lasting 30-120minutes. The brevity, severity, lack of aura,and vomiting occurring daily in clusters, last-ing usually for 4-16 weeks, clearly separate itfrom migraine. The pain is boring, aching, orstabbing and is centred on one orbit withradiation to the forehead, temple, or cheekand jaw ("lower-half headache"). It charac-teristically strikes at night, an hour or so aftersleep, and may recur during the day, oftenat the same time ("alarm-clock headache").In many cases the ipsilateral eye becomesred and bloodshot, watering profusely. Thenostril may be blocked or run. A transientHomer's syndrome is seen in 25% of casesand occasionally persists.

Restlessness, crying, and head bangingbetray the frightening severity of the painand, in contrast to migraine, most patientsget out of bed and pace the floor, even takingnocturnal walks. Alcohol and other vasodila-tors precipitate attacks: nitroglycerine is use-ful as a provocative test-a typical attackfollowing within an hour of a sublingual0.5 mg tablet. "Acute episodic cluster" lastsfor one to four months although occasionallythey continue for a year or more, when it is

139

on 30 April 2019 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.2.134 on 1 F

ebruary 1994. Dow

nloaded from

Pearce

Table 4 Diagnostic citeria for cluster headaches

Episodic(a) At least five attacks fulfilling (b)-(d)(b) Severe unilateral orbital or supraorbital pain, with or

without temporal pain, lasting 15-180 minutes untreated(c) Headache is associated with at least one of the following

signs which have to be present on the painful side* Conjunctival injection* Lacrimation* Nasal congestion* Rhinorrhoea* Forehead and facial sweating* Miosis* Ptosis* Eyelid oedema

(d) Frequency of attacks: from one every other day to eightper day.

(e) History and/or physical and neurological examinations donot suggest other disorders associated with head trauma

ChronicAttacks occur for more than one year without remission, orwith remission lasting less than 14 days. The attacks areclinically indistinguishable from episodic cluster headache.

known as "chronic migrainous neuralgia".Remissions are complete but the clustersrecur every year or two. The quality, timing,duration, and distribution of pain separate itfrom trigeminal neuralgia, migraine, andother cephalgias. There is no family history,and it is uncommon for a patient to have bothmigraine and cluster headache. There is anunexplained high incidence of peptic ulcer."The diagnostic criteria (IHS) are sum-

marised in table 4.

TreatmentThe aim is prevention of attacks.'4 Duringclusters, alcohol is prohibited. Ergotamine isgiven one hour in anticipation of daytimeattacks, and at bedtime for nocturnal attacks.Suppositories are the most useful preparation.Control is good in 75% patients and the drugis stopped each Sunday to see if the clusterre-emerges; if so, ergotamine is continued for

a further week, until the cluster ends. If ergotis unsuccessful, sumatriptan 6 mg subcuta-neously or 100 mg orally,22 methysergide1-2 mg three times daily or verapamil40-80 mg three.times daily are useful alterna-tives. Oxygen 5-10 litres per minute for 10minutes at the onset is often effective29 but f/-blockers and pizotifen are not. Lithium isuseful in the chronic variant if other methodsfail. In intractable cases, a short course ofsteroids often provides relief. Surgery is sel-dom indicated, but trigeminal lesions meetwith occasional success in refractory patients.

Chronic paroxysmal hemicraniaThis is a rare variant35 of cluster headache,occurring predominantly in females, withidentical attacks, often five to 20 per day,which last from three to 15 minutes; theyrespond almost invariably to indomethacin,75-150 mg daily.

CERVICOGENIC HEADACHE

Head pain referred from cervical spondylosisis undoubtedly common, with pain on one or

both sides of the neck radiating not only tothe occiput but also to the temples andfrontal region. It may be a dull "toothache"pain, worse in the morning when the neck has

beeAn kinked on high pillows during sleep; itcan initiate migraine. It can last throughoutthe day, aggravated by neck movement andtension and is a nondescript pain, withoutaccompanying vomiting or physical signsother than restriction of lateral flexion androtation of the neck. Such signs are common,however, in those without headache. Vagueand intermittent symptoms of tinnitus, dizzi-ness, and visual disturbance are sometimesattributed to compression- of the vertebralarteries, but this is unproven. Pain arises fromthe posterior zygapophyseal joints and relatedligaments as the result of osteophytes withirritation of the C2 root or greater occipitalnerves.36 Manipulation endangers the verte-bral arteries and is contraindicated. Collarsare comforting but of little value. Injections ofthe facet joint region anatomically related topain will induce useful temporary remissionin about 70% of subjects.37 Hydrocortisone25 mg or methylprednisolone are used, with 1 mlof 2% lignocaine on three occasions at weeklyintervals. The benefit may wane in a fewmonths, when the injections can be repeated.

GIANT CELL ARTERITISSynonyms are cranial or temporal arteritis.

This condition is crucially important as atreatable cause of headache, it is also a pre-ventable cause of blindness and strokes. Thediagnostic features are shown in table 5.

Pain is generalised or may be sited over theclassic, but rare, reddened, tender, superficialtemporal or occipital artery. The history is ofa few weeks' duration. The patient is unwellwith aches and pains in the shoulder andpelvic girdle muscles (polymyalgia rheumat-ica) and there may be fever, sweats, and mas-seter claudication. Visual involvement (50%patients) is due to an ischaemic optic neu-ropathy which presents with unilateral blind-ness or due to a branch retinal or ciliaryocclusion and is irreversible. Posterior cere-bral artery lesions cause a hemianopia.

Ophthalmoplegia and diplopia may be thepresenting sign, based on ischaemic lesions inthe third, fourth or sixth cranial nerves,before the onset of headache and malaise:hence the importance of early diagnosis. Itaffects the vertebral, and less often thecarotid, arteries and may present as a strokeor transient ischaemic attack.

Every elderly subject with recent headachesor unilateral visual loss should be suspectedof harbouring this condition. Serial erythro-cyte sedimentation rates (ESR; usually

Table S Diagnostic criteria for cranial arteritis38

* Age > 50 years at onset* New onset of localised headache* Temporal artery tenderness or decreased temporal artery

pulse* Elevated ESR > 50 mm/h* Biopsy showing necrotising arteritis

The presence of three or more of these five criteria provides asensitivity of 93-5% and a specificity of 91-2%. Scalp tender-ness and claudication of the jaw or tongue or on deglutitionincreases sensitivity to 95 3%

ESR = erythrocyte sedimentation rate.

140

on 30 April 2019 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.2.134 on 1 F

ebruary 1994. Dow

nloaded from

Headache

70-120 mm/h), sometimes supplemented byan adequate biopsy of the clinically affectedscalp vessel, which should be serially sec-tioned, will prove the diagnosis. Cranialarteritis may spontaneously remit and theESR then falls; thus a single, normal ESRdoes not exclude arteritis. Biopsy is indicatedin clinically borderline cases.

Atypical presentations that should promptassiduous investigation include: (a) patientswith minimal headache; (b) headache in anappropriately aged patient with, initially, anormal ESR; (c) fever of unknown origin; (d)psychiatric symptoms of hallucinations,depression, and "confusional state"; (e) iso-lated third or sixth nerve palsy; rarely inter-nuclear ophthalmoplegia.

Steroids will avert blindness in almostevery case, and should be started immediatelythe patient is seen and the blood samplefor an ESR taken. They do not affect thebiopsy changes for at least 48 hours.Headaches often abate within 24 hours oftreatment.The initial dose of-60 mg prednisolone

daily is quickly reduced as symptoms abateand the ESR falls. The maintenance dose of5-10 mg daily is usually reached within amonth or two and is governed by clinicalprogress and ESR measurements. Laterelapses are common39 and treatment may berequired for many years with gradual reduc-tions by 1 mg per month, only when thepatient and the ESR have'been normal forover two years. In many, small doses are nec-essary for life, as proved by serious relapsesmonths after the dose of prednisolone isreduced.39

HYPNIC HEADACHE (SOLOMON'S SYNDROME)This is a curious headache, seen mainly inthose over 60 years. Patients are woken bypulsating headache, sometimes accompaniedby nausea, at the same time, one to threetimes each night.40 This occurs most nights,lasts about 30 minutes and may coincide withrapid eye movement sleep. It is uncommonand differs from chronic cluster headache inage, generalised location, and absence ofautonomic features. There are no physicalsigns and the disorder is- benign. Theresponse to lithium carbonate 300 mg atnight is often spectacular.

POST-TRAUMATIC HEADACHE AND ITSMANAGEMENTHeadache after head injury is a commoncomplaint. In most circumstances a knock onthe head will cause local bruising and abra-sions no different from those resulting from akick on the shin; local pain subsides in threeto 10 days without sequelae. The emotionalvulnerability of the head and the easyrecourse to medicolegal compensation com-plicate both symptoms and mechanisms.Many victims of severe head injury, withpost-traumatic amnesia of 24 hours or more,waken with no headache. Similarly, theheadache of patients after major craniotomyseldom lasts more than three to seven days.

The commonest complaints are heard fromthose with minor injury (loss of consciousnessless than 20 minutes; Glasgow coma score13-18; stay in hospital less than 48 hours).Despite minor cognitive deficits, most casesleave hospital within a few days, have noorganic signs, recover quickly, and return towork without further complaints. This is par-ticularly true of those suffering injury duringcontact sports. Attributable, post-traumaticheadaches persist for a few weeks, and sel-dom more than six months, unless compli-cated by other issues.The main concern of physicians is the

assessment of symptoms in those with head-aches without accompanying neurologicalsigns, but often with a collection of intrinsi-cally subjective symptoms often called the"post-traumatic syndrome". The complaintsare: forgetfulness, irritability, slowness, poorconcentration, fatigue, dizziness (usually notvertigo), somnolence, intolerance of alcohol,light, and noise, loss of initiative, depression,anxiety, loss of interests, and impaired libido.The number of complaints is often inverselyrelated to the severity of the injury.

In 2493 individuals examined as part of anationwide general population survey, post-traumatic stress disorder was found in 1% ofthe total population, about 3-5% in civiliansexposed to physical attack and in Vietnamveterans who were not wounded, and 20% inveterans wounded in Vietnam.42 Althoughhyperalertness and sleep disturbances,occurred commonly in the general popula-tion, the full syndrome (DSP&-III-R) wascommon only among veterans wounded inVietnam. This argues against the validity ofthe widespread use of this label in those sub-jected to minor injuries.Trauma probably never causes migraine,

but pre-existing migraine may be temporarilyworse, usually for up to three to six months,probably as a non-specific reaction to stressor to disabilities.43The failure of doctors to provide complete

reassurance soon after injury is important indetermining patients' fears of brain damageor subdural haematoma; it also delays returnto work and induces iatrogenic morbid anxi-ety. Headaches are at the site of trauma, oftenwith scalp tenderness, or more often-liketension headaches-are diffuse, aching, tight,or heavy. They resist analgesics, and investi-gations in most patients are both unrevealingand unwarranted. Headaches sometimesimprove when the patient is counselled andreturns to work, but do not invariably disap-pear even when satisfactory settlement isattained. Anxieties, phobias, loss of selfesteem, resentment, and depression are gen-uine accompanying features in some cases,and serve to induce or to aggravate headache.Deliberate exaggeration, or malingering inoccasional cases, are motivated by quest forattention and financial gain. In litigants thereis great pressure from trades union officialsand lawyers which, together with the com-mon delay in settlement, serve to prolong andexaggerate the symptoms.

141 on 30 A

pril 2019 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.2.134 on 1 February 1994. D

ownloaded from

Pearce

Table 6 Simplified plan of managementfor headache

Patient presents with headaches of abrupt onset* Exclude trauma. If signs of meningeal irritation, suspect intracranial haemorrhage or

meningitis and admit to hospital. Perform CT or MRI and, if no tumour, haematoma, orhydrocephalus found, perform lumbar punctureNB: lumbar puncture within six hours of the ictus may miss an early bleed

* If no signs of meningeal irritation, consider a mass lesion, arrange CT or MRI and refer toneurosurgery if necessary

Patient presents with increasing headache, short history* Exclude local cranial pathology, such as glaucoma, sinusitis, dental disease; look for signs of

raised intracranial pressure, suspect mass or hydrocephalus. Arrange CT or MRI beforelumbar puncture

Patient presents with chronic or intermittent headache with no neurological signs* Exclude local cranial pathology: glaucoma, sinusitis, dental disease. If continuous, suspect

tension headache; if paroxysmal, suspect migraine with or without aura. Look for featuresof cluster headache, other headache syndromes, and cranial neuralgias

POST-HERPETIC NEURALGIAThis dreaded complication of trigeminal her-pes zoster, particularly common in the elderly(50% of over 70s) and rare under the age of60, is defined by pain persisting for morethan one month after the eruption. Two-thirds spontaneously recover in one year,leaving a minority with protracted andintractable pain. Errors in diagnosis areavoided by insisting on the presence ofhealed post-herpetic hyperpigmented ordepigmented areas, or both, in a segmentaldistribution. An underlying neoplasm of theaffected roots and systemic immunosuppres-sion should be excluded by judicious investi-gation.

MThe mechanism of pain is uncertain. Thepreferential loss of large diameter neurons,according to the "gate theory" of Melzackand Wall, permits increased transmission ofnociceptive information through the dorsalhorn, thus evoking pain. There can be nogeneral recommendation for acyclovir orsteroids in the acute stages, but in the frailand elderly who are highly vulnerable, it maybe justifiable to employ both drugs44 at theonset of herpes zoster infection in an attemptto avert this potentially grave affliction, inwhich suicide is well-known. In the estab-lished case, symptomatic improvement is dis-appointing. It is worth trying tricyclicsbuilding slowly up to high doses (such as

imipramine or amitryptiline 125 to 200 mgdaily). Additional regular non-opiate anal-gesics by day and, in severe cases, oral mor-

phine or heroin at night will often bejustifiable. Local counterirritants, now called"neuroaugmentation", in the form of freezingsprays or topical capsaicin 0-025% cream canbe helpful. Transcutaneous electrical nervestimulators find occasional success.

ConclusionMost headaches presenting to the generalpractitioner and hospital physician have noominous intracranial cause, but are a sourceof suffering and loss of working time. Manyheadaches are infrequent and self-limiting. Apainstaking clinical approach by a good lis-tener will resolve many problems and willprevent the refractory course in some of thosein whom a cursory initial examination fails to

secure reassurance. Patients presenting prob-lems in diagnosis, and those unresponsive toappropriate treatment will benefit by referralto a neurologist.

Table 6 shows a simplified plan of manage-ment.

1 Stang PE, Yanagihara T, Swanson JW, et al. Incidence ofmigraine headache: A population based study inOlmsted County, Minnesota. Neurology 1992;42:1657-62.

2 Intemational Headache Society. Classification. Diagnosticcriteria. Cephalalgia 1988;8(suppl 7):19-45.

3 Rasmussen BK. Migraine with aura and migraine withoutaura: An epidemiological study. Cephalagia 1992;12:221-8.

4 Rasmussen BK, Jensen R, Olesen J. Impact of headacheon sickness absence and utilisation of medical services:A Danish population study. J Epidemiol CommunityHealth 1992;46:443-6.

5 Pearce JMS. Clinical features of the exploding headsyndrome. J Neurol Neurosurg Psychiatry 1989;52:907-10.

6 Marcus DA. Migraine and tension-type headaches: Thequestionable validity of current classification systems.Clin J Pain 1992;8:28-36.

7 Solomon S, Lipton RB. Criteria for the diagnosis ofmigraine in clinical practice. Headache 1991;31:384-7.

8 Jarman J, Davies PTG, Fernandez M, et al. Platelet['H]-imipramine binding in migraine and tension head-ache in relation to depression. J Psychiatr Res 1991;25:205-11.

9 Leone M, Sacerdote P, D'Amico D, Panerai AE, BussoneG. Beta-endorphin concentrations in the peripheralblood mononuclear cells of migraine and tension-typeheadache patients. Cephalalgia 1992;12: 155-7.

10 Besken E, Pothmann R, Sartory G. Contingent negativevariation in childhood migraine. Cephalalgia 1993;13:42-3.

11 Pearce JMS. Tension headaches: clinical features andmechanisms. In: AV Holden, W Winlow, eds. Theneurobiology of pain. Manchester University Press 1984:235-43.

12 Eh'de DM, Holm JE. Stress and headache: comparisons ofmigraine, tension, and headache-free subjects. HeadacheQ 1992;3:54-60.

13 Kunkel RS. Muscle contraction (tension) headache. Clin JPain 1989;5:39-44.

14 Silberstein SD, Merriam GR. Estrogens, progestins andheadache. Neurology 1991;41:786-93.

15' Pearce JMS. Is migraine explained by Leao's spreadingdepression? Lancet 1985;ii:76.

16 Silbert PL, Edjs RH, Stewart Wynne EG, Gubbay SS.Benign vascular sexual headache and exertionalheadache: interrelationships and long term prognosis.J Neurol Neurosurg Psychiatry 1991;54:417-21.

17 Lance JW. A concept of migraine and the search for theideal headache drugs. Headache 1990;Jan:17-23.

18 Lance JW. 5-Hydroxytryptamine and its role in migraineEurNeurol 1991;31:279-81.

19 Scott AK. Dihydroergotamine: A review of its use in thetreatment of migraine and other headaches. ClinNeuropharmnacol 1992;15:289-96.

20 Humphrey PPA, Feniuk W. Mode of action of the anti-migraine drug sumatriptan. Trends Pharmacol Sci 1991;12:444-6.

21 The Sumatriptan Auto-injector Study Group. Self-treatment of acute migraine with subcutaneous suma-triptan using an auto-injector device. Eur Neurol 1991;31: 323-31.

22 Ekbom K, Waldenlind E, Levi R, et al. Treatment ofacute cluster headache with sumatriptan. N Engl J Med1991;325:322-6.

23 Pearce JMS. Sumatriptan: efficacy and contribution to

migraine mechanisms. J Neurol Neurosurg Psychiatry1992;55:1103-6.

24 The Oral Sumatriptan and Aspirin plus MetoclopramideComparative Study Group. A study to compare oralsumatriptan with aspirin plus metoclopramide in theacute treatment of migraine. Eur Neurol 1992;32:177-84.

25 The Multinational Oral Sumatriptan and CafergotComparative Study Group. A randomized double-blindcomparison of sumatriptan and Cafergot in the acutetreatment of migraine. Eur Neurol 1991;31:314-22.

26 Goadsby PJ, Zagami AS, Donnan GA, et al. Oral suma-triptan in acute migraine. Lancet 1991;338:782-3.

27 Glaxo Holdings plc cited in: Sumatriptan in clinicalpractice. Ninth Migraine Trust International Sympo-sium, London, 8 September 1992.

28 Ottervanger JP, Paalman HJA, Boxma GL, Stricker BHC.Transmural myocardial infarction with Sumatriptan.Lancet 1993;341:861-2.

29 Pearce JMS. Sumatriptan in migraine. BMJ 1991;303:1491.

30 Ludin HP. Flunarizine and propranolol in the treatmentof migraine. Headache 1989;29:219-24.

31 Saxena PR, Den-Boer MO. Pharmacology of antimigrainedrugs. J Neurol 1991;238:(suppl)S28-35.

142 on 30 A

pril 2019 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.57.2.134 on 1 February 1994. D

ownloaded from

143

32 Igarashi M, May WN, Golden GS. Pharmacologictreatment of childhood migraine. Pediatr 1992;120:653-7.

33 Kudrow L. Diagnosis and treatment of cluster headache.Med Clin North Am 1991;75:579-94.

34 Pearce JMS. Cluster headache and its variants. Festschriftfor Lord Walton. Postgrad Medj 1992;68: 517-21.

35 Sjaastad 0, Dale I. A new (?) headache entity "chronicparoxysmal hemicrania". Acta Neurol Scand 1976;54:140-59.

36 Sjaastad 0. Cervicogenic headache: the controversialheadache. Clin Nevol Neurosurg 1992;94(suppl):S147-9.

37 Bovim G, Sand T. Cervicogenic headache, migraine with-out aura and tension-type headache. Diagnostic block-ade of greater occipital and supra-orbital nerves. Pain1992;51:43-8.

38 Hunder GG, Bloch DA, Michel BA, et al. TheAmerican College of Rheumatology 1990 criteriafor the classification of giant cell arteritis. ArthritisRheum 1990;33: 1122-8.

39 Bengtsson BA, Malmvall BE. Prognosis of giant cellarteritis including temporal arteritis and polymyalgiarheumatica. A follow-up study on ninety patientstreated with corticosteroids. Acta Med Scand 1981;209:337-45.

40 Newman LC, Uipton RB, Solomon S. The hypnicheadache syndrome: a benign headache disorder of theelderly. Neurology 1990;40:1904-5.

41 Pearce JMS. The post-traumatic syndrome and whiplashinjury. In: Recent advances in clinical neurology, Vol 8.Ed.C. Kennard, 1994 (in press).

42 Helzer JE, Robins LN, McEvoy L. Post-traumatic stressdisorder in the general population: Findings of theepidemiologic catchment area survey. N Engl Med1987;317: 1630-4.

43 Weiss HD, Stem BJ, Goldberg J. Post-traumatic migraine:chronic migraine precipitated by minor head or necktrauma. Headache 1991;31:451-6.

44 Watson CPN. Postherpetic neuralgia. Neurol Clin 1989;7:231-48.

Headache

on 30 April 2019 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.57.2.134 on 1 F

ebruary 1994. Dow

nloaded from