Neck and headache—Do we see paths through the jungle? An overview and an hypothesis

Abstract

Background and overview:

It is widely accepted that cervicogenic headache (CEH) originates in the neck. In many circles, it is also accepted that neck–tongue syndrome belongs to the headaches that have their origin in the neck. For many headache researchers, the list: “headaches stemming from the neck” ends here. The objective of this overview was to explore the field and to determine whether there are grounds for adding other headaches to this list.

Discussion:

We suggest that headaches stemming from the neck possibly consist of five different subgroups: CEH, neck–tongue syndrome, tractor drivers’ headache, posterior headache subsequent to protracted neck-ache, and chronic paroxysmal hemicrania with mechanical attack precipitation. An overview of the clinical characteristics of each putative subgroup with comments is given.

Introduction

It was a true statement when, in 1989, it was stated that “there is no agreement as to the clinical presentation of headaches stemming from the neck.” ¹ Cervicogenic headache (CEH) had been described in 1983. ² That neck-ache could have local causes had probably seemed self-evident. Through the centuries, there had been no lack of ideas claiming that headache also could stem from the neck. However, the various proposals differed substantially and left the reader in great doubt. ^{3 –7} The neck may have appeared as an impenetrable jungle in this respect and at that time. The neck is extremely complex and could, presumably, produce a multifaceted variety of symptoms and signs.

We started our search for headache stemming from the neck in the early/mid-seventies ⁸ ; we must have been unbelievably fortunate when we, after some years, found a patient who could precipitate attacks at will, by rotating the neck. Mechanical precipitation later proved to be a cardinal sign of headache with cervical origin.

Over the years, it has become steadily more apparent that headache stemming from the neck may have various appearances. Still, with heightened insight, the original concept of CEH seems to fortify its position.

Various headache subgroups putatively stemming from the neck

Cervicogenic headache

CEH was described in 1983, and with that description as a guideline, diagnosis could be established with reasonable precision. ^2,9 Characteristically, a typical headache episode can be precipitated by neck movements, by awkward neck positions, or by external pressure against sensitive areas in the neck, for example, the groove behind the mastoid process and the upper part of sternocleidomastoid muscle. The unilateral pain does not shift side. Solitary attacks start in the nape of the neck, and pain spreads to frontal/ocular and periocular areas, where the headache frequently has its maximum. ¹⁰ The pain could also spread to the symptomatic side shoulder and upper extremity, usually in a vague, non-radicular way. Range of motion in the neck is reduced. The Cervicogenic Headache International Study Group (CHISG) diagnostic criteria are based on these features. ¹¹ Diagnostic blockades were not made obligatory because such blockades were not yet part of routine clinical work. For scientific work, such blockades are highly desirable. The diagnostic criteria were pruned and sharpened for the second edition. ¹² The Vågå study of headache epidemiology was our last large-scale headache study that also involved CEH. The field work for the study was carried out during the period 1995–1997. ¹³ CEH diagnoses were established during that period or during the period immediately following (see, Sjaastad and Fredriksen ¹⁴ and Sjaastad and Bakketeig ¹⁵ ). The diagnostic guidelines employed were the CHISG criteria, with which we had considerable practical experience at that time.

CEH was first mentioned in connection with the IHS diagnostic system in the 2004 version. ¹⁶ We could, therefore, not compare the two sets of diagnostic criteria in connection with the Vågå study, and we have had no opportunity since to carry out a systematic comparison.

In recent years, we have nevertheless compared the two sets of criteria, but merely on a theoretical basis. ¹⁷ Such a comparison, if it were to count fully, ought to be based on practical, diagnostic experience with both systems. In our opinion, the time seems ripe for such a comparison to be carried out. If carried out by able and impartial investigators this could be a valuable contribution.

It should be emphasized also in the present context that there is widespread misunderstanding that autonomic features play a role in the CHISG criteria, but not in the International Headache Society (IHS) criteria. ¹² The table sums up the features on which the CEH diagnosis is based. There is not a single autonomic feature in the table.

What is the evidence that CEH differs from migraine? A few main points will be mentioned: migraine pain pulsates, whereas CEH pain usually does not. Migraine drugs do not help CEH patients. The migraine pain frequently alternates side, while CEH pain does not. ¹⁸ In CEH, specific abnormal structures can be identified in the neck and dealt with surgically (see later). Vincent and Luna ¹⁹ have validated the CEH criteria.

Two specific studies in a remarkable way separate migraine and CEH. Firstly, it is striking that the plasma concentration of the neuropeptide calcitonin gene-related peptide does not increase during CEH attacks, ²⁰ in contrast to what is the case in migraine, an important observation. Secondly, pregnancy does not usually influence the tendency to headache attacks in CEH, ¹⁴ whereas it clearly does so in migraine. We believe that these observations provide strong evidence that we are faced with two different disease processes in CEH and migraine.

The prevalence of CEH (“core cases”) in the Vågå study, that is without admixture of migraine/tension-type headache (T-TH) cases, was 2.2%. ¹⁵ If cases of co-existent migraine or T-TH were included, the prevalence would be 4.1%. Another Norwegian study found a prevalence of 0.17%. ²¹ The difference between 0.17% and 2% is so marked that there in all probability are underlying, methodological problems. Basically, three CHISG criteria were used in the 0.17% study, ²¹ though it was not specified which. The study group was relatively young (30–44 years). CEH cases could also be excluded because of the inclusion criteria for frequency that were used, that is, ≥15 headache days during the last month; or ≥180 headache days in the last year. CEH is not necessarily always a chronic headache.

Bilaterality of CEH

CEH is, as already emphasized, a unilateral headache. But there are exceptions, and these are of varying nature.

• a. Spreading of attack pain to the opposite side

When attack pain is particularly long-lasting and/or intense, it may also be felt on the opposite side, but the regular side pain always dominates. ¹⁰

• b. Unilaterality on two sides

This seems to be a meaningless term, but describes a specific situation.

Several years ago, it was stated that CEH could be a C2 neuropathy. It does not seem to be quite that way. In various CEH series, ^22,23 a degenerative spine disorder with encroachment mainly upon the C5-6 and C6-7 foramina on the symptomatic side was found. These were focused operatively with foraminal widening procedure ²³ (Smith/Robinson).

Although the complaints and the corresponding pathological findings originally were unilateral, the degenerative disease goes further and can, sooner or later, attack the opposite side, either at the same level or at other, neighboring levels. The subsequent, contralateral pain is not a spreading of pain from the original source, but a de novo pain, originating on the opposite side. ²³ In other words, this is a unilaterality on two sides and entirely different from a simple spreading of the attack pain in ordinary CEH. In this situation, therapeutic measures should be directed toward the other side to be effective.

All these three conditions, the genuine, unilateral CEH, and the two bilateral variants, seem to be examples of headache stemming from the neck because of the postoperative pain freedom for prolonged periods. ^21,22,24

In the beginning, the resistance against the idea of CEH was considerable. As stated in many contexts by now, CEH seems largely to be internationally accepted. There are also symptomatic cases, such as one case of a fifth cervical nerve root schwannoma, causing a CEH-like clinical picture. ²⁵

Neck–tongue syndrome (N–TS)

This syndrome was described by Lance and Anthony in 1980. ²⁶ An attack consists of a unilateral, upper neck and/or occipital pain, usually of ultrashort duration, with subsequent, transient ipsilateral numbness in the tongue, with or without a foregoing tongue pain. In one of our own cases, there was also a lingual spasm. ²⁷ The pain may spread to the head, first and foremost to the occipital region, but also to the top of the head and exceptionally to the periocular area. The precipitation mechanism, causing this cascade of events, is an abrupt neck rotation. Solitary attacks are usually short-lasting and rare, and the patient is able to master the situation. No large series has been published, and it was originally considered that this syndrome is extremely rare. This may not necessarily be so. The prevalence of N–TS was around 0.2% in the Vågå study, ²⁷ and none of the four individuals in Vågå had consulted a physician for their complaints.

There is close temporal connection between neck rotation and pain. A cause and effect mechanism seems to be highly likely: this constellation of symptoms and signs stems from the neck. With this syndrome, the resistance against a cervical origin of symptoms and signs has been close to zero. The pathogenesis ²⁶ seems to differ widely from that of CEH, and the two clinical pictures differ entirely. This more or less proves that the neck may give rise not only to headache, but to at least two entirely different types of headache.

Tractor drivers’ headache

This headache was studied during the Vågå study of headache epidemiology. ²⁸ In this part of the world, farmers are mainly occupied with their tractors during week-long chores, spring, and autumn. The head is then rotated to the right, day in and day out. Neck- and head-complaints appear after a few days. There was sufficient information from 94 farmers. Fifteen had both neck-ache and headache; 13 had neither neck-ache nor headache, while the remainder had neck-ache only. None had a headache without also having a neck-ache. Both headache and neck-ache were of moderate intensity and bilateral, the headache mainly being localized in the occipital area, occasionally moving up to the vertex, or even anteriorly. The pain characteristics are thus entirely different from those of CEH. The underlying pathology, therefore, also probably differs fundamentally.

There was no obvious culprit to create the headache, logically speaking, except the neck rotation. In all probability, this is a headache stemming from the neck. A few days after the chore there was no more headache, and this type of headache did not reappear until the next chore.

That headache can be linked to specific tractor-work has been known from previous studies. However, no details were presented. ²⁹

Posterior headache, subsequent to protracted neck-ache ³⁰

In the wake of the insight obtained in connection with tractor drivers’ neck- and headache, other topics came to the fore. Perhaps other types of long-lasting and relatively hard, that is, unphysiological, neck strain could also give rise to headache of this, or a similar sort?

This type of headache was more or less unknown to us prior to the Vågå study. Therefore, only a general headache workup was originally carried out on individuals, later placed in this category. However, the locals kept telling about their form of neck- and head-complaints. Finally, a post hoc study could be carried out, based on data already provided in the individual records, looking for the combination of neck-ache and headache. As suspected, this combination of complaints was observed particularly frequently in certain work categories, such as carpenters, car mechanics, or electricians. ³⁰ A total of 6.6% of the local participants seemed to be affected. The bilateral headache always started in the nape of neck and spread upward. In more than half of the cases, it included the occipital area only, whereas in about 70% of the cases, both occipital and parietal areas were covered. The headache frequently ended at the top of the head. Ocular and radicular arm pain were not experienced, but bilateral, diffuse shoulder pain was present in approximately one-third of the cases. A headache episode lasted a few hours to a couple of days. ³⁰

This headache, phenomenologically, seems to have a striking resemblance to tractor drivers’ neck- and headache, the one major difference being the causative factor itself, the tractor. It may well be that future experience will demonstrate that the underlying mechanisms are so similar that the two groups ought to be combined. CEH differs essentially from both these headaches with regard to laterality and forward extension of the pain. The similarity between these two headaches to a certain extent vouches for the likelihood that also headache subsequent to protracted neck-ache derives from the neck. Affected individuals feel a kind of union between pain in neck and head and a similarity in quality, the head pain representing an extension. Moreover, there is tenderness in the musculature in the neck, that is, trapezius and splenius muscles, when applying a direct pressure of approximately 4 kg. In 12% of the cases, there was an upward spreading of discomfort, mostly to the occipital region and more rarely to the vertex. Headache attacks or exacerbations were, however, never elicited upon external pressure application.

Neck- and shoulder-complaints have been thoroughly studied by many research groups. Westgaard has summarized the work of his group. ³¹ Peculiarly enough, the engagement not infrequently seems to stop at the transition to the head.

Could “tractor drivers’ headache” and “posterior headache” (subsequent to protracted neck-ache; for short, types III and IV) be examples of “unilaterality on two sides”? and thus be variants of CEH. The arguments against this possibility may seem to be major ones:

• (A) The unilaterality on two sides is present in long-lasting CEH; CEH never starts with a bilaterality. Unilaterality always antedates the bilaterality. In types III and IV, the bilaterality is present from the onset.

• (B) Attacks cannot be precipitated mechanically in types III and IV.

• (C) In CEH, head pain spreads anteriorly to periocular area. In types III and IV, pain generally ends at the vertex. Bilateral pain, ending at the vertex, seems to be essentially different from a unilateral pain ending periocularly.

• (D) The long-term temporal pattern, in CEH, attacks appear irregularly, mostly in a chronic form whereas in types III and IV, the headache either appear in the chore (III) or in connection with particular work positions (IV).

Chronic paroxysmal hemicrania (CPH) mechanical precipitation of attacks

CPH is characterized by side-locked attacks of intense headache, day and night, with highly varying frequency and up to more than one attack per hour. ³² The relatively short-lasting pain attacks are centered around the eye and are combined with same-sided, autonomic symptoms and signs, such as lacrimation, conjunctival injection, and rhinorrhea. Indomethacin in adequate dosages takes away the attacks completely. CPH now corresponds to 3.2.2 in the International Classification of Headache Disorders 3rd edition. ³³

CPH in general was originally not known to be associated with the neck, and patients with such linkage were detected several years after CPH was described. They make up perhaps 10% of CPH cases. ³⁴ Attacks can be precipitated by forward bending or rotation of the neck, or by external pressure against, for example, the lateral part of C4-5 transverse processes, symptomatic side. Attacks appear within seconds. Lacrimation and conjunctival injection are integral parts also of mechanically precipitated attacks. However, in some patients in this group, autonomic phenomena are minimal or may at times even be lacking, making these attacks incomplete. Such abortive attacks are also shorter and more lenient than regular attacks. Indomethacin desensitizes the area of hypersensitivity.

In one quite extraordinary patient, fullblown, excruciatingly intense attacks could incessantly be produced by bending the head forward or by pressure against the transverse processes C4–C5 (later referred to as case 5). Refractory periods seemed to be lacking. ³⁵ The first sensation that an attack was brewing was “pulsation” in the nape of the neck, symptomatic side. Shortly thereafter, and just prior to onset of pain, her head began to nod, and rather coarse vibration could easily be observed simultaneously in the musculature in a circumscribed area about 3 cm broad [(trapezius and underlying musculature, at the level of C4–C5(?)]. On indomethacin medication, she would still have a slightly unpleasant sensation in the nape on neck movements, but no attack would follow. Just after an attack, there would be a short, spontaneously occurring, period of piloerection and goose flesh in both upper extremities, flexor side, corresponding to the C5 innervation zone, fossa cubiti.35

The many traits reminiscent of CEH (unilaterality of pain without side-shift; mechanical precipitation of attacks both by neck movements and external pressure) were more than counterbalanced by the formidable discrepancies between the CPH variety and CEH. Variant CPH had autonomic phenomena, like lacrimation, and also increased corneal indentation pulse amplitudes, the attacks being more short-lasting and much more intense than those in CEH. Indomethacin takes away the attacks in variant CPH, but not at all in CEH. In spite of a certain similarity, the CPH variant is thus far from being a true CEH. In this unique patient, a four-night sleep study was carried out, and 18 attacks were recorded: 17 of them started in rapid eye movement (REM) phase, compared to two other females with regular CPH who had 6 and 10 nocturnal attacks, respectively, none of them appeared in REM phase. ³⁶ The unique patient was delivered by a particularly troublesome breech birth.

This patient rather definitely has a headache stemming from the neck. An important question remains: are there at all any spontaneous attacks in this particular patient?

By now, the situation around CPH appears to be like: (1) regular CPH, hitherto without obvious abnormalities, linking it to the neck. (2) CPH with mechanical precipitation of attacks and hypersensitive cervical spine zones, that indomethacin can desensitize. (3) A separate version of no. (2), for example, “vibration” in the nape of the neck musculature, and REM-locked onset of attacks. Alternatives (2) and (3) seem to be connected to the neck, whereas there is no solid evidence that alternative (1) is so.

In case 5, one faces a diagnostic dilemma: headache attacks can be precipitated mechanically from the neck; this is consistent with CEH. The unilateral attacks, however, have CPH traits, with same-sided autonomic features, and are counteracted in an absolute way by indomethacin. In many headache circles, the absolute indomethacin effect has had a status of an axiom: the diagnosis then equals CPH (or: hemicrania continua). However, CPH—a primary headache—can in principle not be provoked from the neck; and CEH—a secondary headache—is not influenced by indomethacin. The two lines of arguments may seem incompatible. Is the axiom reliable? In CPH, no hypersensitive bony surfaces have been observed. In principle, the mechanism behind the indomethacin effect in CPH is unknown. In case 5, the most lateral part of the C4/C5 structures are hypersensitive and attacks can be provoked by local stimulation. Indomethacin counteracts this hypersensitivity completely, whereas placebo does not have any effect. The status of the vulnerability of the bony cervical surfaces seems to fluctuate in complete accordance with the degree of indomethacin influence. The correspondence seems to be complete and lasting. In other words, these sites may seem to be the area where indomethacin has its effect. Indomethacin, in this special case, seems to work differently from what is the case in regular CPH. The beneficial effect of indomethacin in case 5 is a different story. Case 5 still seems to belong to the headaches originating in the neck.

Our work with case 5 is to be considered only a first step to clarify its nature as also indicated in the subtitle of this article—an overview and an hypothesis.

Considerations and conclusions

It seems reasonable to characterize the three first-mentioned headache forms, that is, “genuine” CEH; the spreading of pain to the other side, as well as the “unilaterality on two sides” as CEH, and variants of CEH. Presently, there seem to be five disorders with a common denominator: all stemming from the neck. They differ clinically and although they are distinct disorders, it would probably be rational to keep the term cervicogenic headache, an established term, for the original, unilateral headache ⁹ in the CHISG 1990 and 1998 versions. ^11,12 For the other four conditions, the original terms could/should be used.

None of them are frequently occurring headache forms. But CEH itself, and posterior headache subsequent to protracted neck-ache, are both sufficiently prevalent to deserve a considerable degree of attention. If the two disorders were intermingled, and this had been the case in much of the last century, it is easy to imagine the trouble caused for the observer. Prior to 1970, it would probably have been close to impossible to put any of the five headaches herein into the correct diagnostic category with a reasonable degree of accuracy.

Is this then the end of the story? Hardly. But it is probably far from being only the beginning.