Abstract
This study was designed to compare the prevalence and clinical characteristics of ‘cold-induced headache’ between migraine and episodic tension-type headache patients. Seventy-six migraine and 38 episodic tension-type headache patients were included in the study. An experimental model of an ‘ice-cream headache’ was developed for the study. The pain occurrence period, its location and quality were recorded for each patient who felt pain in their head during the test procedure. Pain in the head occurred in 74% of migraine and 32% of ‘tension-type headache’ patients. Although the most frequent pain location was the temple in both groups of patients, this rate was greater than twofold in migraine patients when compared with episodic tension-type headache patients. While headache quality was throbbing in 71% of migraine patients, it was so in only 8% of the episodic tension-type headache patients. Considering all the results, it seems that ‘cold-stimulus headache’ is not only more frequent in migraine patients, but also its location and quality differ from ‘tension-type headache’.
Introduction
Holding ice or ice-cream in the mouth, or swallowing a cold food or drink as a bolus, may cause discomfort in the palate and throat. It may also refer pain to the forehead or temple by the trigeminal nerve and to the ears by the glossopharyngeal nerve (1, 2).
Raskin and Knittle found that 15 out of 49 subjects, who were not normally prone to headache, experienced infrequent mild ice-cream headaches at some time in their lives. In contrast, 55 out of 59 migraine patients were subjected to such headaches, which were frequent and severe in 46 of those affected. They proposed that the vascular reaction to cold in patients with erratic vasomotor regulation (migraine) may be excessive. Ice-cream headache therefore, could be no more than an epitome of the migraine mechanism (3).
In our study, an experimental model of ‘ice-cream headache’ was developed in order to compare migraine and episodic tension-type headache sufferers’ response to the cold stimulus, which was provided with contact of an ice-cube to the patients’ palate.
Materials and methods
The study was carried out at Kocaeli University, School of Medicine, in the Department of Neurology's ‘outpatient headache clinic’ between August 2001 and September 2002. Patients were seen between these dates and were questioned using a standard form. A detailed headache history was taken and a complete neurological and physical examination was accomplished. Patients were classified according to International Headache Society (IHS) criteria (4) by means of a semi-structured interview. One hundred and thirty-five patients who fulfilled the criteria of ‘migraine without aura’ (the patients who had migraine headache more than 15 days a month within 3 months were not included in the study) and ‘episodic tension-type headache’ were chosen as candidates for the study.
A research method was developed for invoking an ‘ice-cream headache’ in patients. The method was called the ‘ice test’ and approved by Kocaeli University, School of Medicine Research Ethics Committee. The test procedure and the characteristics of the cold-stimulus headache were explained to each patient and they were asked if they would volunteer to join this trial. Prior experience of cold-induced headache was not questioned.
In order to induce head pain, standard ice-cubes which had a 12 × 14 mm square estimated contact area to the palate were provided. Ice-cubes were all supplied from the same refrigerator which cools to − 12°C. Patients who had dental plates removed them before the test procedure. Patients were told to put the ice-cube on their tongue, press it onto their palate and try to keep it in the middle of the palate (palatum durum was described). This method was intended to prevent direct cold stimulus to the teeth. Patients were instructed to close their mouths when they placed the ice-cube at the described location and keep their lower jaw motionless during the test procedure. Any motion of the lower jaw was accepted as patients’ intentions to protect themselves against ice-induced head pain, and they were excluded from the study. Patients were also told to give a sign by putting up their hands as soon as possible when they felt pain on any part of their head. The period from the closure of mouth until the patient's sign was calculated. An additional time was given to the patients to localize the specific pain area with their fingers on their head after the pain occurred, with the ice-cube still in the mouth. Maximum contact time of ice to palate was determined as 90 s. The test was evaluated as negative when patients did not have pain within that period of time. The period between the emergence of pain and its disappearance was also calculated. The time elapsed until occurrence, location and quality of pain was recorded for each patient. Patients whose ice-induced pain lasted longer than 5 min were excluded from the study. The test procedure was developed and carried out on all patients by HMS after an interview with all patients, therefore, there were no interobserver differences.
Pain locations were determined mainly as orbita, temple, vertex and frontal, temporal, parietal and occipital area. These locations were determined with the cutaneous projections of the cranial bones (frontal, temporal, parietal and occipital bones). The temple area was accepted as cutaneous projections of the sphenoid bone's ala major (5). Other locations that were described by the patients were also recorded. Quality of pain was described as throbbing and non-throbbing.
None of the patients was on prophylactic therapy. Patients were told that they should feel free to stop the trial whenever they wanted. The patients taken into analysis were only those who were considered to have successfully completed the above-described test procedure by the test executor.
Four patients who had severe headaches when exposed to cold stimulus refused to participate in the study – two because of upper respiratory system infection and two because of Ramadan. Eleven patients were excluded because of inaccuracies in the test procedure. Another two patients whose headaches lasted 25 and 80 min, respectively, after the ice test were also excluded. The test results of the patients who reported ambiguous sensations other than head pain were evaluated as negative and they were also included in statistical analysis. Finally, 38 patients with episodic tension-type headache and 76 patients with migraine (total 114 patients) were included in the study.
In statistical analysis, while age was determined as a numeric variable, gender, diagnosis, location and quality of pain were determined as categorical variables. Because pain occurrence periods were given at 10-s intervals, this data was also accepted as a categorical variable. Mean, standard deviation, maximum and minimum values of age were calculated for ‘migraine’ and ‘episodic tension-type headache’ patients. Distribution of categorical variables in the ‘migraine’ and ‘episodic tension-type headache’ groups was worked out with chi-square tests and Fisher's exact test. Significance level was determined as 0.05. Statistical analysis of study was carried out by Statistical Package for the Social Sciences (SPSS) program version 10 (SAS Institute, Cary, NC, USA).
Results
The group of episodic tension-type headache patients consisted of 21 females and 17 males, aged 16–78, mean age 38.00 ± 15.41. Migraine patients’ group consisted of 67 females and nine males, aged 14–61, mean age 35.57 ± 11.45.
Among 114 patients, the ‘ice test’ induced head pain in 68 patients. While 56 of them (48 female, 85.7%; eight male, 14.3%) were in ‘migraine group’, 12 (six female, 50%; six male 50%) were in ‘tension-type headache group’ (x2 = 18.66; P < 0.001; d.f. = 1). Confounding the effect of gender differences to the cold stimulus was tested using the Mantel-Haenszel method and no significance was found (P = 0.286, d.f. = 1, OR = 0.599–5.679).
Pain location was in the temple in 35 (51.5%) out of 68 patients whose headache was induced by the ‘ice test’. The locations other than the temple reach the maximum 16.2% and ‘expected count’ in 20 cells (83.3%) was less than five. In order to carry out a statistical analysis, rows were combined and two categories were formed (temple region and regions other than temple). The pain induced by ice was seen most frequently in the temple in the patients with migraine (57.1%). Pain in the temple was 25% in patients with episodic tension-type headache (Table 1). This difference between the groups was statistically significant (x2 = 4.09; d.f. = 1; P = 0.043).
Pain locations in two groups
While ‘ice test’-induced head pain was of throbbing quality in 40 out of the 56 migraine patients (71.4%), it was so in one of the 12 patients with episodic tension-type headache (8.3%). This difference between groups was also statistically significant (Fisher's exact test P < 0.001).
In order to test the distribution of pain emergence between two groups, rows were combined and two categories were formed, 0–30 and 31–70 s (pain was not evoked in any of the patients after 70 s); there was no statistically significant distribution in respect to pain emergence between the two groups. (Fisher's exact test P = 0.096) (Fig. 1).
‘Ice test’-induced pain intervals in two groups. ETTH, episodic tension-type headache.
Discussion
In IHS criteria, cold-stimulus headache is coded as 4.3 on a two-digit level, and ‘ice-cream headache’ is coded on a three-digit level (4.3.2), which is named as ‘ingestion of a cold stimulus’, and its criteria were composed. This type of pain is described by IHS as ‘a pain produced in susceptible individuals by the passage of cold material, solid or liquid, over the palate and posterior pharyngeal wall’ (4).
According to another description (1, 2), an ice-cream headache has two components. Firstly, the material to be ingested must be cold and secondly, the material to be ingested must be in the bolus form. The bolus form may provide a widespread contact area as the tongue pushes it through the palate and then the pharynx. But in the ‘ice test’, the ice-cube provides only a small contact area on the palate and tongue, but not on the pharynx. The amount of melting ice which is swallowed during the test procedure is probably not enough to irritate the pharynx. But longer contact time of the ice-cube may, in time, help the spreading of cooling to the palate and neighbouring areas (especially the soft palate). So, the ‘ice-test’ procedure does not lead to the same sequence of events in the ingestion of the bolus of cold material. Another point is that the contribution of the tongue to the provocation of pain is unknown. In the ‘ice test’, in order to apply the ice to the palatum durum; the tongue may also be involved in the pain provocation procedure. Therefore, we don’t claim that the pain evoked by the ‘ice-test’ procedure is the equivalent of an ‘ice-cream headache’.
A similar protocol was carried out by Bird et al. (6) who applied ice-cream to the palate instead of an ice-cube, which was used in our procedure. In the method of Bird et al. there are two remarkable points. Firstly, the ice-cream applied and ingested to controls was colder than the one applied and ingested to the migraine group (−26°C, − 15°C, respectively). Secondly, some of the patients with migraine were on prophylactic therapy. Bird et al. admitted that a colder and larger quantity of ice-cream should have been used in their study, and this could be the reason for the lower prevalence of ‘ice-cream headache’ in migraineurs than in controls.
In this study, although the ice applied to palate is not colder than the ice-cream Bird et al. used, a longer application time of the ice to the palate allows spreading of cooling to the neighbouring areas. The contact location and area of the ice to palate is more stable than the ice-cream because ice is a firm substance. Another important point in this study is that none of the patients were on prophylactic therapy.
In 8% of his study group, Mattsson (7) accomplished inducing pain by ingesting 150 ml ice water (0–4°C) with a straw. He suggested that active migraine was a significant predictor of cold-water-induced headache. He did not find any statistical difference in the location, quality and duration of cold-water-induced pain between participants with and without active migraine.
Mattsson admitted that the cold-water test was moderately well standardized because the exact location of the end of the straw against the palate, and the magnitude of water flow, could not be controlled in that study. In addition to that, participants with prior experience of cold-induced headaches may have adjusted to ingest the water with the intention of avoiding pain associated with cold-water-induced headache (7). However, as previously mentioned, in our study, movement of the patients’ lower jaw was accepted as patients’ intention to prevent them from the cold-induced headache and they were excluded from the study by the examiner.
In Mattsson's study (7), 8% of pain occurrence was less than in both those of Bird et al. (6) (26% in migraineurs, 40% in controls) and in our study (74% in migraineurs, 32% in episodic tension-type headache patients). In our study and that of Bird et al., a colder substance was applied to patients in order to evoke pain. It seems that using colder material to induce pain increases the rate of positive results. The rate of positive results is even higher in our study. This higher rate may be related to prolonged contact time of the cold stimulus.
When the first 30 s of our study is considered (the period in which Bird et al. applied the cold stimulus to migrainuers and controls), pain was invoked in 44 (38.6%) of 114 patients (the whole group) and in 39 (51.3%) of 76 migraineurs. In the study of Bird et al., these percentages were 29.1 and 26%, respectively. Kaczorowski and Kaczorowski (8) have found that cold stimulation of the palate induced by gobbling up ice-cream more than doubles the likelihood of developing an ice-cream headache among middle school students. We can deduct from this data that repetitive cold stimulus with high pace, instead of prolonged contact of cold stimulus, could also increase the occurrence rate of ice-cream headache. We therefore state that the application method of the cold material may be an important factor in eliciting positive results.
We suggest that a cold-stimulus headache should not be evaluated only in terms of its existence among primary headache groups but also because, according to our study, it has some different characteristics in migraineurs. The cold-induced headache of our method is more common in migraineurs than in patients with tension-type headache; while the pain's quality is throbbing in migraineurs, it is dull in tension-type headache patients. Although the pain is felt most frequently in the temple in both groups, this rate is much higher in migraine patients.
There are some limitations in our study. The executor of the test was not blind to the diagnosis of the patients. Patients were not asked about their recent analgesic use. Moreover, the number of the patients with episodic tension-type headache who were subject to the test was limited and the whole group taken in this study dominantly consisted of females. Although we excluded patients who obviously did not press the ice-cube against their palate, some patients may still have pressed the ice-cubes more firmly than other patients. With our method, it was not possible to determine ice-cube's pressure on the palate. Such differences may have an impact on the frequency and characteristics of pain induced by the ice in the two groups of patients, and this may be a threat to the validity of our findings.
In this study, we evaluated the effect of cold stimulus to the palate in two primary headache disorders, and we documented the differences between them. We don’t claim that the pain evoked by the ‘ice-test’ procedure is the equivalent of ‘ice-cream headache’ and we did not examine the validity of the ‘ice test’ in true ‘ice-cream headache’ sufferers. Therefore, the findings must be considered with caution due to the difference(s) between the ‘experimental model’ and ‘true’ ice-cream headaches.
Migraine is a neurovascular disease and the trigeminal nerve has a unique role in the pathogenesis of migraine. A specific portion of trigeminal pathways may be activated as a reflex for seconds or minutes by sudden cooling of the pharynx or may discharge migraine headache for hours. This suggests that hyperexcitabilty of trigeminal pathways persists between migraine attacks and that periodic discharge of these pathways could initiate migraine headache (9). Apart from trigeminal hyperexcitability, the partial defect of pain control mechanisms in migraine patients was also suggested because of their inclination to ice-cream headache (10).
This study leads us to think that the interaction between the trigeminal and vascular system is specific to migraine patients. We hope that some of the characteristics of cold-stimulus headache in migraine patients described in this study will be supported by further studies.