Abstract
The main objective of the present investigation was to search for cervicogenic headache (CEH) after whiplash injury. Whiplash patients (n = 587), were followed for a year after their emergency service consultation. A total of 222 patients with headache after 1 month went through interview and examinations at 6 weeks, 6 months and 1 year. All included persons received a questionnaire after 1 year. De novo CEH seemed to be present in 8% at 6 weeks and in 3% at 1 year. Previous car accidents, pre-existing headache and neck pain were more frequent in chronic CEH individuals than in those in the cohort without CEH. Range of motion in the neck was reduced in 65% of chronic CEH individuals hours after the accident, compared with 41% in the cohort. Cybex inclinometer, at 6 weeks and 1 year, demonstrated reduced extension in the neck. CEH seems to be present after whiplash injury, particularly in the early phase. It seems similar to, but probably not identical to, non-whiplash CEH.
Introduction
Headache is a common phenomenon in the population at large. Nevertheless, there seems to be a marked increase in headache after accidents with whiplash mechanism, particularly in the acute phase (1–3). The descriptions of these headaches are vague and indistinct. A crucial question is whether the headache after an accident is an increase in intensity, frequency or other manifestations of a pre-existing headache or whether it is a de novo headache.
Headache originating in the neck has historically had many names. The term cervicogenic headache (CEH) was coined by Sjaastad and coworkers in 1983 (4) and its manifestations described. Criteria for CEH were issued in 1990 (5). CEH as such has been accepted by the IASP (6). The criteria for CEH to be used in future research were published in 1998 (7). Accidents with whiplash mechanism were among the first-version diagnostic criteria for CEH (5). The major part of the work on CEH has, however, been done on chronic head pain patients, regardless of whether any history of such trauma was present or not (e.g. 8, 9). To our knowledge, there has been no previous, prospective study that has brought into focus the putative presence and development of CEH in a whiplash population, based on the CEH criteria (5).
The Oslo study is a cohort study designed to search for symptoms and clinical findings during the first year after the accident. In particular, we have endeavoured to follow the course of de novo, unilateral CEH. In this communication, we will compare the cases of chronic CEH with those in the cohort.
Material and methods
During the period November 1993 to May 1995, we consecutively included in the study persons who had been involved in a car accident with possible whiplash injury and sought the emergency service of the Municipality of Oslo for this reason. The emergency service covers the whole of the city of Oslo, which has nearly 500 000 inhabitants. It is accessible to all as a public service for a small nominal fee and it is widely used because the Norwegian Health Service is mainly a public service. The vast majority of people who seek acute medical care after a whiplash injury, seek the emergency service.
Further inclusion criteria were: individuals ≥ 18 years and fluent in Norwegian. Other injuries were not allowed to interfere with the interpretation of the results. Accordingly, exclusion criteria were: loss of consciousness or direct head trauma in connection with the accident; other injuries, requiring special medical attention or definite, pathological findings on cervical spine X-ray examination (e.g. fractures); and acute or previous, relevant neurological diagnoses (e.g. cranial nerve palsies or radiculopathy).
The inclusion criteria were fulfilled by 629 patients, and 587 (93%) agreed to be included in the study. The study group consisted of 49% women and 51% men. The age ranged from 18 to 79 years (mean 35 years). The patients came to consultation within hours to 1–2 days after the accident, the vast majority within a few hours.
The primary consultation after the accident
The primary consultation consisted of a standard clinical examination by the physician on call. This included neck mobility assessment and a short-version neurological examination. Neck mobility was assessed in the sitting position according to the following rough criteria for normal movement: forward flexion: ≤ 2 fingers can be inserted between chin and chest; extension: forehead parallel to the ceiling; side flexion: c. 45°; rotation: chin on line with the shoulder. Cervical X-rays were invariably obtained. Each patient also filled in a questionnaire concerning previous illness, and frequency (daily, weekly, monthly) of neck pain, headache and sleep disturbances. They gave an estimate of the speed of the two cars involved in the collision. Low impact accidents were included, as in many whiplash materials, because patients involved in such accidents also claim economic compensation for accident-related disabilities. The speed difference was used as an indicator of trauma severity, even though such a rough assessment has many pitfalls. Headache and neck pain intensity were measured on a 1–9 point VAS scale (1 = no pain, 9=worst imaginable pain).
Questionnaire at 4 weeks; the basis for selection of patients for more extensive examinations
Four weeks after the accident all participants received a short questionnaire. Those who still had symptoms were grouped as a sub-cohort to be followed more carefully (Fig. 1). The questions concerning injury-related symptoms were measured on a Likert scale of frequency (never, occasionally, daily) and severity (weak, moderate, strong, very strong); intensity of headache and neck pain were measured on VAS-scales (1–9 points). For headache, at least ‘occasional, moderate’ complaint was an inclusion criterion. All patients still out of work were also included regardless of symptom report. This group was followed with clinical examinations and extensive questionnaires, at 6 weeks, 6 months and 12 months. This sub-cohort consisted of 222 individuals, i.e. 38% of the total cohort (Fig. 1). The number of patients dropped with time, the actual number being listed in the tables. Patients with mild symptoms tended not to attend the consultations. Individuals who did not attend for the 1-year consultation received a short questionnaire along with the rest of the cohort.
Flow chart. All participants were examined at entry, 0–3 days after the accident. They all received a questionnaire 4 weeks after the accident (□). A sub-cohort of patients reporting an appreciable amount of symptoms at 4 weeks were called in for examination (□) and larger questionnaires at 6 weeks, 6 months and 1 year (▪). The participants were also addressed at 1 year; 153 individuals attended for examination at the hospital, the reminders received a questionnaire. Total patient replies at 1 year was 80%.
All individuals reporting headache at 4 weeks, except occasional headaches of weak intensity, were thus examined at 6 weeks. Those who had a headache that seemed to fulfil the criteria for CEH were examined at least once by all three authors to ensure the correct diagnosis.
The clinical examination included a relevant neurological examination, provocative tests such as Spurling's test, stretching of neck muscles to elicit pain, identification of tender points (temporal area, jaw joint, greater (GON) and minor (MON) occipital nerves). Measurements of passive neck mobility were included, using a Cybex EDI inclinometer, known to give little variability (10). Forward flexion and movements backwards and to the sides were measured in the sitting position, while rotation was measured in the supine position. A light pressure was used at the end of each testing procedure of the passive movement.
Individuals with de novo CEH at 1 year were called in for diagnostic blocks of the occipital nerves. This was not done at an earlier stage in order not to interfere with other parts of the study.
The data were analysed on an SPSS statistical program, using standard statistical procedures. The study was approved by the regional ethics committee. All patients gave written informed consent to participate.
Results
Characteristics of the de novo CEH the first year after the accident
One-sided CEH, fulfilling the diagnostic criteria (Table 1), was found in 48 (8.2%) at 6 weeks, in 26 (4.4%) at 6 months and in 20 (c. 3.4%) at 1 year (Fig. 2). The latter 20 were regarded as patients with chronic CEH. None of them seemed to have had a pre-existing headache similar to CEH; in particular, none of them had had a unilateral headache without sideshift.
Minimum requirements for CEH diagnosis (7): Presence of criteria in present series
(Ib) Reduced ROM at first examination 0–3 days after the accident. P< 0.01, compared with rest of the cohort.
∗Pain or even attacks.
†Confirmatory evidence by diagnostic blockades is obligatory in scientific works since 1998 (7) but was not obligatory according to the 1990 criteria used for definition of CEH in this study (5).
‡Only carried out in six individuals. Furthermore, Ib and/or Ic were present in 79% of the cases.
Frequency of CEH at various stages after whiplash trauma.
The sex ratio (M : F) was c. 2 : 3 at all intervals. The mean age was 38 years at 6 weeks, 41 years at 6 months and 42 years at one year. The pain was described as non-pulsating and most often as intermittent. In some, there was dizziness. Dizziness is an integral criterion of CEH, although an unspecific one. Ipsilateral shoulder pain was present in 48% of CEH at 6 weeks.
The frequency of attacks, as well as the duration of each attack in the CEH group, generally dropped with time (Table 2). Whereas the frequency and duration of the CEH were reduced with time, mean headache intensity seemed rather unaltered: 5.5 at 6 weeks, 5.5 at 6 months and 5.2 at 1 year.
Frequency and duration of headache in the chronic CEH group at 6 weeks and 1 year and in the part of the cohort reporting to have headache (unspecified) at 1 year
Pre-injury status
Prior to the accident, neck pain and headache (at least monthly) were reported by a significantly higher proportion of CEH patients than total cohort individuals. There was pre-injury neck pain in 45% of the chronic CEH and in 15% of the cohort, and pre-injury headache (non-specified) in 65% of the chronic CEH and 24% of the cohort.
Accident-related information
The estimated mean speed difference between the cars involved was of the same magnitude for chronic CEH (40 km/h) and the cohort (41 km/h). Approximately one-tenth were low-speed accidents (< 11 km/h) in both groups. There were rear-end collisions in 85% of the chronic CEH and in 67% of the cohort. As many as 70% of the chronic CEH individuals had been involved in a previous car accident compared with 45% in the cohort. There is little information available concerning headache in connection with the previous accident, apart from the fact that it was not a unilateral headache without sideshift.
The examination at the emergency service
None of the acutely reported symptoms at the initial examination (Table 3) were specific for CEH. Neck pain, stiffness in the neck and headache were reported by the majority, including those who got rid of the symptoms within weeks. The intensity of headache/neck pain was significantly higher in chronic CEH than in the cohort. Of the individuals with chronic CEH, 45% initially reported interscapular pain, 30% pain radiating to the shoulders, and 10% diffuse pain/paresthesia more distally in arms, but nobody reported pain radiating to the C6, C7 or C8 distribution areas.
Symptoms at the first examination within 3 days and at 4 weeks (all differences are significant)
∗ P < 0.01 (the P-values concern a comparison with the cohort minus the individuals with chronic CEH).
On initial examination approximately one-third of the individuals with CEH at 6 weeks and chronic CEH, had reduced extension of the neck. In the chronic CEH group, 65% had initially reduced ROM in one or more directions (Table 4), as against 41% of the cohort (P < 0.05).
Range of motion (ROM) in the neck at first examination within 3 days after injury for the group of individuals diagnosed with CEH at 6 weeks and at 1 year (chronic CEH), and for the cohort. Percentage of patients with reduced ROM based on a rough clinical estimation (see text)
∗ P<0.05.
∗∗ P=0.06 (the P-values concern a comparison with the cohort minus the number of individuals specified at the top of the respective column).
Questionnaire at 4 weeks
The questionnaire sent to all participants after 4 weeks was answered by 89%. Table 3 show that at this stage chronic CEH differed from the total cohort with regard to symptoms. The mean intensity of neck pain and headache were only slightly reduced in chronic CEH during the first 4 weeks after the accident. However, both neck pain and headache intensity had been considerably reduced in the total cohort. Thus, headache intensity was reduced to 0.3 in chronic CEH and 1.0 in the cohort. Chronic CEH individuals had significantly more problems with daily activities than the cohort individuals. One out of four CEH individuals were still on sick leave, as opposed to approximately one out of 10 in the total cohort. Soft neck collar and medication were more frequently in use among chronic CEH than cohort individuals and drug consumption was higher than for other pain patients.
Situation at 1 year
At 1 year, chronic CEH individuals had significantly more symptoms than the cohort individuals (Table 5). Headache (Table 2), neck pain, stiffness in the neck and vertigo/dizziness were particularly prominent. All but one of the chronic CEH individuals experienced stiffness in the neck. In spite of this, only 15% of chronic CEH evaluated themselves as being much worse at 1 year than prior to injury. On the other hand, only 20% evaluated themselves as at least having reverted to pre-injury status. Fifteen per cent were on sick leave, as opposed to 4.5% in the cohort. Approximately two-thirds felt that the quality of their working capacity and/or social and family life were below par because of the accident. Only one of them (5%) felt that the shortcoming in these respects was considerable. Seventy per cent of the chronic CEH individuals used pain-killers at 1 year; 15% on a daily basis.
Symptoms at 1 year (for headache, see Table 2)
()=daily frequency of symptoms.
∗ P < 0.01 (P-value concerns a comparison with the cohort minus the individuals with chronic CEH).
The clinical examination at 6 weeks and 1 year
These examinations were performed by the authors and were more thorough than the initial examination. Range of motion in the neck (ROM) results are demonstrated in Table 6. Chronic CEH individuals had significantly reduced extension in the neck at 6 weeks and at 1 year, compared with the cohort. Otherwise, no significant differences were obtained. Rotation is the motion that seemed to change most with time, but it did so in all groups.
ROM in the neck at 6 weeks and 1 year. Cybex inclinometer investigation at 6 weeks and 1 year (see text)
∗∗ P < 0.05.
∗ P=0.07 (the P-values concern the cohort minus the individuals with chronic CEH).
At 1 year, chronic CEH individuals reported a higher frequency of pain in response to passive stretching of the neck muscles to the sides than the rest of the examined individuals. There was also a higher frequency of local pain in the neck compared with the rest of the examined individuals in response to Spurling's test (negative test because of lack of nerve root irritation).
Pain in response to pressure applied over the occipital nerves GON/MON was present in the majority of all examined individuals, both at 6 weeks and 1 year (Table 7). Radiation to the forehead in response to pressure either over the GON or MON was present in 55% of chronic CEH individuals at 1 year, significantly more frequently than in the cohort (28%). Such provoked pain is one of the cardinal features of CEH (5, 7). In contradiction to what is the case in CEH in general, neck pain was often present bilaterally and the unilateral radiating pain to the side of the headache/forehead, could be elicited from both sides (including the non-symptomatic side) in solitary cases in the present series.
Precipitation of pain over the posterior cervical nerves GON and MON at 6 weeks and 1 year. Chronic CEH and examined part of the cohort
()=unilateral pain.
∗ P < 0.01 (the P-value concerns a comparison with the cohort minus the individuals with chronic CEH).
Diagnostic GON/MON blocks were only done in c. one-third of the chronic CEH individuals at 1 year (n = 6). The main reason for not carrying out blocks in the remainder was the low level of headache during the actual office hours. Four out of the six individuals had positive response to the technically speaking satisfactory blocks, in the sense that they temporarily experienced an extensive reduction in their headaches within minutes after the block. The response rate may seem to be consistent with previous studies (11).
Discussion
Radanov and co-workers (2) found headache after whiplash injury to have a benign course, irrespective of presence or absence of pre-traumatic headache. Both frequency and intensity decreased with time. Pre-traumatic headache was a significant risk factor for whiplash-related headache. Also in the present study, the course seems favourable, as for unilateral CEH during the first post-injury year.
The individuals in the subgroup exhibiting CEH as part of their whiplash-associated disorder (WAD), were slightly older than the individuals in the total group; there was a certain female predominance, as in migraine. A relatively high proportion of CEH individuals had been involved in a previous car accident and there was a significantly higher frequency of non-specified headache and neck pain prior to the accident among CEH individuals than in the cohort. This may indicate that there is an inherent neck vulnerability at the time of injury, as well as an acquired propensity to headache generation after injury.
In 17 of the 20 patients (85%) with chronic CEH, there had been rear-end collisions. This is a somewhat higher proportion than for the total cohort (67%). Possibly, the ‘pure’ whiplash mechanism, as described by Crowe in 1928 (12), is important in producing chronic CEH.
All individuals with de novo CEH after the accident were thoroughly interviewed and none of the affected individuals seemed to have had a unilateral headache without sideshift prior to the accident, so we are probably faced with a de novo headache. The situation as regards the other individuals in the cohort is somewhat more unclear as they were not all interviewed. No attempt at classification of the pre-existing headaches has been made.
CEH is unlikely to be a misinterpreted migraine
Between 20% and 30% of CEH patients (13) have been found also to meet the IHS criteria for migraine without aura (14). Can CEH in our series be misdiagnosed migraine? The current criteria for migraine without aura state that a solitary attack lasts 4–72 h (14). At 1 year only 25% of the CEH individuals in this study regularly had attacks lasting 4–72 h (Table 2). There was no need for immobility during headache, a typical trait in migraine. The pain was thus seemingly of lower intensity than in migraine. Moreover, headache was not associated with heightened sensitivity to sensory input. Nausea was associated with the headache in half of the CEH cases. Most importantly, the pain was non-pulsating. There is thus little to substantiate a migraine diagnosis. There were positive criteria for CEH (see Table 1). From a clinical point of view the evidence pointed towards CEH.
Clinical findings in CEH in this series
In the acute phase, CEH seems to be a part of the triad: neck pain/stiffness in the neck/headache. This symptom triad should be regarded as part of the acute WAD and can not be used to predict any kind of chronicity. In the majority of patients the symptomatology related to this triad diminished with time as the neck function improved. Reduced neck mobility has been reported to be a common finding in chronic CEH in general (10). Reduced ROM, particularly backwards, pain upon flexion to the sides, with normal or very slightly reduced rotation, were parts of the picture in some CEH cases after whiplash. Reduced backward motion might not have been found in other studies because the flexion and extension are reported as a sum. Reduced early phase rotation (Table 4) did not seem to predict disability.
CEH follows an expected biological curve during the first year after whiplash injury
Contrary to the findings in many whiplash studies, in a study from Lithuania (15) no increase in headache was found two years after a car accident. This has been used by others as support for the idea that increased disability after an accident with whiplash injury only forms a link in the chain of events leading to insurance claims.
In the present study there is evidence of de novo CEH after whiplash accidents. The characteristics of the headache in the present cases were neither congruent with those of psychosomatic disorders nor with disorders without a demonstrable pathological substrate. The fact that motion in only one or two directions seemed to be decreased in the majority of cases, and that in particular extension was affected, as might be expected after rear-end collisions, suggests that there may be organic changes in the neck in these individuals. The number of affected individuals, as well as the magnitude of discomfort, decreased along the time axis and followed an expected curve for disability after trauma. The downward slope of the frequency curve of CEH may seem to flatten out with time. For that reason, it may seem somewhat unlikely that in this cohort the curve would reach zero level at, for example, 2 years. A higher-power study than the Lithuanian study may be needed to exclude any disability with an incidence as low as CEH after whiplash, especially as it is a study of motor vehicle accidents and not a study of the much more limited cohort of individuals who seek medical care after such an accident.
CEH after common whiplash is probably not only unilateral
For scientific studies concerning CEH, it is recommended to include only unilateral CEH (5). One inclusion criterion in the present study was therefore unilaterality without sideshift. Patients with bilateral headache would accordingly automatically be excluded. In individuals in the remainder of the cohort (excluding the chronic CEH), there was still at 1 year more headache than prior to accident. The possibility exists – and not only theoretically – that in some of these cases there could also be a post-whiplash CEH. Bilateral CEH could be as frequent or even more frequent than unilateral CEH after whiplash.
CEH is probably best understood as a syndrome, and the symptomatology of CEH following in the wake of whiplash might differ somewhat from that of CEH without any previous whiplash. Besides the possibility that it can be bilateral, there may be less movement-restriction in the neck, and in the case where it is present it may be limited to one or two directions. Mechanically precipitated pain-attack in unilateral CEH in general shows a clear preference for the symptomatic side. In post-whiplash, unilateral CEH cases, there seemingly is a tendency that the pain can be provoked also from the non-symptomatic side.
The present study throws a critical light upon the homogeneity of the CEH group as such. CEH studies hitherto carried out have included CEH cases irrespective of aetiology, stage and severity. The time of studies of ‘unqualified CEH’ without any closer definition of the target group may have come to a close.
Footnotes
Acknowledgements
Medical Division, HQ Defence Command Norway and in particular Dr Jon Ivar Brevik, generously assisted the authors with the study design and the statistical analyses.