Abstract
The influence of pregnancy upon the head pain of cervicogenic headache (CEH) has been studied in 14 patients (number of pregnancies 25). Migraine was used as control group (n = 49; number of pregnancies 116). CEH was diagnosed according to The Cervicogenic Headache International Study Group guidelines. Migraine was diagnosed according to International Headache Society (IHS) guidelines; a further requirement was that at least eight of nine solitary IHS diagnostic requirements of migraine were present. In 79%—or more—of CEH patients, attacks seemed to appear just as usual during pregnancy; in one patient, attacks stopped completely, and in two there may have been a minor reduction of attacks. A significantly lower number of migraine patients (up to 18%) were more or less uninfluenced by pregnancy (CEH vs. migraine P < 0.0001, X2 test). The lack of response to pregnancy may be a sort of biological marker in CEH. It may also help in clinically distinguishing CEH from migraine when CEH starts early in life, i.e. prior to pregnancies.
Introduction
Migraine is known to be vastly improved during pregnancy. This was originally considered to pertain mostly to the migraine without aura variety (1–3). It was later demonstrated that it also applies to the with aura variety (4, 5). This beneficial role of pregnancy has been demonstrated not to be limited to migraine; it is also present in, for example, cluster headache (5, 6) and CPH (7). On the other hand, occasional patients with cervicogenic headache (CEH) had for more than a decade been telling us that their headache remained totally uninfluenced by pregnancy. These accounts had an anecdotal character. We have therefore gone through the files of patients with CEH, in whom the onset of CEH antedated pregnancy, to search for more solid information. Since migraine without aura is the main differential diagnostic alternative for CEH, a comparison with migraine has been made.
Materials and methods
CEH
Fourteen patients were included. Information regarding influence of pregnancy was present in the primary neurological work-up in seven cases. Nine patients were contacted by telephone or came for consultation in the summer of 2001: two patients were contacted for confirmation of the primary information.
The diagnosis of CEH was made according to the guidelines of The Cervicogenic Headache International Study Group (CHISG) (8) and IASP (9). According to the revised version (8), anaesthetic blockades are an obligatory diagnostic component in cases to be used for scientific work. In the present context, anaesthetic blockades were directed towards the greater and minor occipital nerves and/or the lower cervical nerves (8). In the old version, anaesthetic blockades were not obligatory (10). Accordingly, in three of our patients, who were examined prior to the introduction of the new guidelines, no diagnostic blockades were carried out. All patients, inclusive of the three without blockades, were considered genuine cases of CEH, clearly fulfilling the clinical criteria of its time. This also concerned patient 6 (Table 1), in whom also ‘migrainous features’ were present (see later). Invasive therapy was carried out in most of these cases and comprised: (i) radio frequency treatment of the planum occipitale (n = 6; four successful; two unsuccessful (11)); (ii) dorsal column electrical stimulation (n = 4; all: beneficial result to some extent (12)); (iii) cervical disc operation (n = 1; successful (13)). The post-operative observation period in all these cases exceeded 5 years, and in the vast majority it exceeded 10 years. The study is grossly retrospective; however, three patients also had children after the first consultation.
Cervicogenic headache: effect of pregnancy
Mean number of pregnancies: 1.79. Neck trauma in 79% of the cases. The neck traumas were all indirect, except in no. 6, where it was direct.
Neck traumas—indirect as well as direct—were recorded. Headache was purely or predominantly unilateral (8) in 13 cases. In one case, following whiplash trauma, it was bilateral (no. 8).
Migraine
All 49 patients fulfilled the IHS criteria for migraine (14), and not only that: it was a prerequisite that they had a high number of the migraine criteria stipulated by the IHS, i.e. eight or nine. No other selection of the migraine cases was carried out. The nine criteria for migraine without aura were: (i) duration of attacks 4–72 h; (ii) nausea; (iii) vomiting; (iv) phonophobia; (v) photophobia; (vi) pulsating quality; (vii) aggravation upon minor physical activity; (viii) unilaterality; (ix) moderate or severe pain. Criterion (i) was obligatory. The mean number of migraine criteria was 8.57. Patients with as well as without aura were included (Table 2, right column). In the with aura cases there was one additional migraine criterion; this additional criterion was not included in the calculation of the mean. The age of onset was between 6 and 25 years, mean 15 years.
Complete absence of migraine attacks during pregnancy
Aura +/−, Occasionally presence of aura, occasionally not.
Migraine patients lacking information concerning pregnancy were recently contacted by telephone. One patient could not state exactly which head pain was due to migraine or to intoxication of pregnancy, which she also had during the first pregnancy. This particular pregnancy was therefore excluded. The next pregnancy was, however, included, and during that she experienced an improvement of 90–100%. Nine of the patients became pregnant after the first consultation; the series was grossly retrospective.
A full neurological examination was carried out in the CEH cases and most migraine cases and a partial examination, including cranial nerves and face/cephalic/neck structures, in most of the remaining cases.
Results
CEH
A total of 25 pregnancies were registered in 14 patients, the mean number of pregnancies being 1.8 (Table 1). In 11 of 14 patients, i.e. 79% of the cases, there seemed to be a uniform picture—that of non-interference by pregnancy upon the tendency to headache. The average age of onset was approx. 23 years. Neck trauma, prior to headache onset and mostly of an indirect nature, was common: in 11 out of 14 cases, i.e. in 79% of the cases (Table 1).
In only one patient (i.e. 7%) did pregnancy cause an absolute improvement from the second/third month onward—in both pregnancies (no. 6). This 61-year-old patient has been followed prospectively since the late twenties, i.e. from a time more than a decade ahead of the description of CEH (15). Strikingly, the headache has always been left-sided. Her case has briefly been described previously as one of partial effect of dorsal column electrical stimulation (12). During the > 8 years long post-operative period, adjustments of the stimulation parameters have bettered her situation. She might by now altogether be around 50–60% improved; she maintains a full-time job of high responsibility; analgesics consumption is vastly reduced; but she still has exacerbations—and, of note, for these sumatriptan clearly helps. During episodes of transitory apparatus failure, she has been in poor shape. She is thus in no doubt about the beneficial effect of stimulation. She clearly represents a diagnostic challenge, with possible coexistence of CEH and migraine, each of the headaches requiring separate therapy. The picture was more ‘migrainous’—with separate attacks—at the time of the pregnancies.
In two patients (nos 7 and 13, Table 1) there may have been some reduction of attacks during pregnancy.
All in all, the picture in Table 1 may be even more homogeneous than depicted. The diagnosis in case 6, although also consistent with CEH at present, may have been migraine without aura at the time of pregnancies. The assessments of possible attack reduction in cases 7 and 13 may also have been overcautious on our part. The possibility therefore exists that a clear interference by pregnancy upon attack frequency is more or less inconsistent with a CEH diagnosis.
Migraine
Number of cases improved by pregnancy
There were a total of 116 pregnancies, i.e. a mean number of 2.4 pregnancies per patient. Absence of attacks was reported by 32 migraineurs, i.e. 65% (Tables 2 and 3). In five (10%) of the 49 patients, the headache seemed to be uninfluenced by pregnancy. and in another four (8%), there seemed to be only a moderate influence (≤ 50% improvement) (Table 3). Conceivably, in some patients in the latter group, there might in reality have been no influence at all. In 82% of the cases or more, pregnancy seemed to have had an ameliorating influence on headache.
Migraine: extent of influence by pregnancy
These figures, i.e. lack of improvement in 79% CEH patients, vs. lack of improvement in 18% of the migraine patients, differ significantly (P < 0.0001, χ2 test). The figure for CEH may even be higher (see above).
Period of improvement
There was firm information concerning the duration of improvement in the vast majority of patients. Among those with information, there was an approximately even distribution between those with reduction of attacks during the ‘whole pregnancy’ on the one hand, and those with improvement from the second month/midway during the childbearing period on the other (Table 4). In some patients, the improvement continued until menstruation returned.
Migraine: period of improvement
Improvement in the subdivisions of migraine
A 100% improvement was enjoyed by a total of 32 patients, i.e. 65%. Such improvement was enjoyed by patients within all subgroups of migraine, including patients with aura (Table 2). Migraine with aura patients experienced the beneficial effect of pregnancy to approximately the same degree as patients within the other subgroups of migraine. One patient with migraine with aura and absence of migraine from the third month experienced a drastic increase in migraine attacks during the first trimester (during the first of two pregnancies).
Discussion
The present study demonstrates a striking lack of influence of pregnancy upon headache in CEH compared with migraine (P < 0.0001, χ2 test).
The force of the present CEH series
One strength of this study is that in the gross majority of the CEH cases, not only anaesthetic blocks, but even invasive therapy directed towards the cervical spine/occipital area was carried out and with more than just marginal success. This is probably as far as one can go in corroborating a CEH diagnosis at this stage of development. This study therefore in all probability has been carried out in authentic CEH patients. The fact that the outcome, as far as the influence upon pregnancy is concerned, was so clear-cut to some extent compensates for the essentially retrospective nature of the study.
Problems with case collection in a CEH study of this type
CEH is first and foremost a disorder in the female; thus, in one series, a female preponderance of 87.8% was found (16). Moreover, CEH does not seem to be uncommon (17, 18). A priori, the search for female CEH patients who had had childbirths might be considered straightforward. There is one formidable barrier in this respect: the mean age of onset of CEH, which was 35.1 ± 17.4 years in Vincent and Luna's series (16). Moreover, it usually takes several years from onset until the patient comes to our cognisance. In the usual case, the optimal time in life for childbearing generally has passed when we see these patients. A prospective study can therefore not easily be carried out. We were able to find only a limited number of young CEH patients. In 79% of the cases, there was a foregoing neck trauma. There was no selection process with regard to trauma. The results obtained in the ‘subgroup’ with traumas will not necessarily have a predictive value vs. CEH in general. However, in two of the three non-traumatic cases, and possibly in all three of them, there was no headache reparative effect of pregnancy (Table 1).
Effect of pregnancy upon migraine
The present study confirms previous observations (4, 5) that pregnancy influences migraine favourably, also the with aura variety (Table 2). In the present study, the migraine diagnosis was established by requiring at least eight of nine specific diagnostic criteria stipulated by the IHS (14). We are not aware of any other study concerning migraine/pregnancy with such rigid diagnostic criteria. Even with rigid migraine criteria, the finding of reduction of migraine attack frequency during pregnancy is upheld. Memory deficits may, nevertheless, play a role. As an example: when patients stated that the improvement/absence of attacks concerned the ‘whole pregnancy’, this might well mean that the improvement set in at any time prior to 2–3 (4) months into the pregnancy.
Possible significance of lack of influence of pregnancy in CEH
But for a few exceptions, CEH and migraine behave differently during pregnancy. ‘Vascular factors’ supposedly play a major role in migraine pathophysiology. One of the signs of pregnancy in general is a change in the pulse synchronous waves of intraocular pressure (19). These changes may reflect ‘vascular’ changes in other areas of significance, as far as migraine pathogenesis is concerned. This early reduction in pulsatile amplitudes during pregnancy presumably occurs in migraine and CEH alike. While migraine is likely to be a ‘vascular headache’, CEH is more likely to be ‘neuropathic’ in nature (20). The disease process in CEH may continue, independent of pregnancy and uninfluenced by a change in putatively significant ‘vascular’ changes.
The response to pregnancy may possibly be a marker for clinically differentiating CEH from migraine. The present study affords further strong evidence in favour of a dichotomy between migraine and CEH. The fact that CEH generally starts so late in life may, nevertheless, reduce considerably the value of the beneficial effect of pregnancy as a ‘diagnostic test’.
Footnotes
Acknowledgements
We are indebted to GlaxoSmithKline of Norway for generous support in all phases of the investigation. The authors are also grateful to the personnel at the Vågå Health Centre at Vågåmo for their aid. Last, but not least, we thank the inhabitants of the Vågå commune for their collaboration.