The characteristics of head pain in response to an experimental cold stimulus to the palate: An observational study of 414 volunteers

Abstract

Objective:

The objective of this article is to study the characteristics of headaches following an experimental cold stimulus and evaluate whether the ICHD-II criteria for headache attributed to ingestion or inhalation of a cold stimulus (HICS) are adequate for the diagnosis of the headache induced by ice on the palate area (ice-induced headache (IH)).

Results:

One hundred and fifty-three out of 414 (37.0%) individuals tested had IH, but only 77/153 (50.3%) satisfied the ICHD-II criteria. The frontal (60.8%) and temporal (48.4%) areas were the most affected ones, with bilateral (77.1%) predominance, often of the pulsatile type (41.2%). One hundred and forty-seven out of 379 (38.8%) individuals who presented with a previous history of primary headache presented with IH, while only six out of 35 (17.1%) who had no history of previous headache reported pain in the test (OR 3.063, 95% CI 1.241–7.557). The ice-induced headache test was positive in 115/240 (47.9%) of the migraine sufferers and in 32/139 (23%) of the tension-type headache sufferers (OR 3.076, 95% CI 1.924–4.918).

Conclusion:

IH is predominantly frontotemporal, bilateral and throbbing, most commonly affecting migraine sufferers, and the ICHD-II criteria are insufficient for classifying all individuals. We should, however, recommend caution regarding such a generalization from our findings with experimentally provoked cold stimulus headache to cold stimulus headache in general (i.e. HICS, ICHD-II).

Keywords

Headache ice-cream cold migraine prevalence

Introduction

Headache attributed to ingestion or inhalation of a cold stimulus (HICS) is a common form of headache recognized by the International Headache Society since the first classification in 1988 (1,2). In spite of being of frequent occurrence, there are few studies on HICS (3–10) and none of them has ratified the more recent International Classification of Headache Disorders 2nd Edition (ICHD II-2004) diagnostic criteria (2). They have apparently never been validated. For instance they refer to HICS as non-throbbing and acute pain that develops immediately after the cold stimulus. Nevertheless, a few studies have shown that HICS can be throbbing (8,9). It is difficult to understand what ‘acute/immediately’ means, since many individuals take more than a minute to develop HICS. The ICHD-II also established that HICS disappears within five minutes after removal of the cold stimulus. It is also a fact that several individuals have had HICS for more than five minutes. This was reported by 9.9% of those interviewed by Fuh and colleagues (6).

Unlike other headaches, HICS is not disabling but because it is of such short duration (seconds to minutes) and avoidable by removing the provoking stimulus (e.g. frozen ice treat), it seldom interferes with an individual’s daily tasks. Another HICS characteristic is that only certain individuals are susceptible, with some evidence of greater susceptibility in migraine sufferers (3,8,9).

HICS prevalence varied from 7.6% to 93% (5–11). In adult populations it was observed that 8% of those interviewed reported having had HICS (8,10); other studies reported prevalences of 15% (11), 37% (7) and 60% (9). In teenage populations prevalence ranged from 41% (6) to 79% (5).

In a literature review we found only four studies in which HICS was provoked (4,5,8,9), through the ingestion of ice (9), ice cream (4,5) or even icy water (8). When HICS prevalence is evaluated in a given population by asking whether the use of icy foods has ever provoked HICS, rates cannot be compared to those obtained during experimental triggering using cold stimuli. The variation is great, ranging from a low frequency of 7.6% in a female population between 40 and 74 years old in Sweden (8) to 93% among migraineurs (of both genders) hospitalized for several causes, who reported a previous history of HICS in the United States (3).

As the diagnostic criteria for HICS of the ICHD-II (2) were still not evaluated, the aim of this study was to provoke HICS by an experimental design of ice on the palate area (headache following an experimental cold stimulus or ice-induced headache (IH)) and to verify whether all the patients that happened to develop headache would satisfy such criteria. In the absence of a good correspondence, new criteria will be suggested, based on the results of this study. Since some criteria are based on whether the individual is a migraine sufferer, we evaluated the HICS characteristics in the two groups – migrainous and non-migrainous (tension-type headache) patients.

Methodology

The final sample comprised 414 volunteers consisting of 266 (64.3%) females and 148 males (35.7%), from eight to 84 years of age, with a mean age of 31 ± 12 years. They were healthy subjects without any significant or pre-existing health problems, with the exception of previous history of primary form(s) of headaches. The study was carried out between November 2008 and May 2009 in the city of Caruaru, Pernambuco, Brazil. Healthy university students and their family members were recruited to voluntarily participate in the experiment at the Associação Caruaruense de Ensino Superior. There was no financial compensation. This was a cross-sectional analytic observational study.

Sociodemographic data (age, sex), history of previous primary headache (considering the interviewee’s whole life) (Table 1) and headache characteristics were evaluated by a headache medicine specialist. A previous history of HICS was also recorded. In addition, data on the type, intensity and location of pain were also evaluated using ICHD-II criteria (2). The individuals were subdivided into three groups: i. individuals with no previous history of primary headache (n = 35); ii. individuals with a previous history of tension-type headache (n = 139); and iii. individuals with a previous history of migraine headache, including probable migraine (n = 240). When tension-type headache coexisted with migraine the individual was regarded as migraineur.

Table 1.

Demographic characteristics of the 414 volunteers.

Group	n	Age (years)
Without previous history of headache	35 (11 females, 31%)	31 ± 16
Tension-type headache	139 (79 females, 57%)	26 ± 10
Migraine	240 (176 females, 73%)*	29 ± 12^†

p < 0.001 χ²; ^†p > 0.05 Kruskal-Wallis.

Procedure of stimulation by using ice

In the attempt to provoke headache induced by ice in the volunteers, a previously described procedure (9) was used. To induce pain, ice cubes measuring approximately 20 × 20 × 35 mm (weighing 14.4 ± 0.5 g, n = 10) were placed between the hard palate area and the tongue (upper surface in contact with the palate, 15 × 20 mm, 300 mm²; lower surface in contact with the tongue, 20 × 35 mm, 700 mm²). Each volunteer was stimulated just once with a single ice cube. Before the procedure participants were told to remove any dental prosthesis that might be covering the palate. They thus had to place the ice cube on the tongue and press it against the hard palate, trying to keep it centralized. This precaution avoids teeth being stimulated by the ice. After placing the ice cube, the volunteers were instructed to close their mouth and keep their jaw motionless during the test. While the mouth was closed the volunteers were asked to indicate with the hand the onset and location of the pain. After the 90-second procedure of ice stimulation there was a 55% weight loss in the ice cube (ice cube weight after 90 seconds (s) stimulation 6.5 ± 0.7 g, n = 10).

The contact time of the ice cube on the palate was determined in 90 s. Counting began at the moment of closure of the mouth. Latency for the onset of the headache was the lapse of time from the moment the ice cube entered into contact with the palate up to the instant the subject felt the pain. Pain duration was timed from onset to relief, whether or not the ice cube was in contact with the palate. Information on the characteristics and intensity of pain were also obtained. The test was considered to be negative when no pain appeared until 10 minutes after the ice cube was removed. Of the 416 volunteers, two were excluded because they showed intolerance to ice and could not maintain the ice cube in contact with their palate for 90 s.

Individuals who reported headache after the contact of the ice with their palate had their headache characteristics compared with the criteria 13.11.2 of ICHD-II published in 2004 (2) (see Table 2).

Table 2.

13.11.2 Headache attributed to ingestion or inhalation of a cold stimulus according to the International Classification of Headache Disorders 2nd Edition (ICHD-II) (2).

Previously used term: Ice-cream headache

Description: Short-lasting pain, which may be severe, induced in susceptible individuals by the passage of cold material (solid, liquid or gaseous) over the palate and/or posterior pharyngeal wall.

Diagnostic criteria:

A. Acute frontal (1) non-pulsatile headache fulfilling criteria C and D

B. Cold stimulus to palate and/or posterior pharyngeal wall due to ingestion of cold food or drink or to inhalation of cold air

C. Headache develops immediately, and only, after a cold stimulus

D. Headache resolves within 5 minutes after removal of the cold stimulus

(1) In migrainous patients, the headache may be referred to the usual site of migraine headache.

In relation to the nature of the headache, the volunteers classified it as throbbing or pulsatile, burning or blasted, weight or pressure or shock sensation. To evaluate headache intensity, a numerical visual scale was used showing figures from ‘0’ to ‘10,’ thus permitting the patients to identify their pain intensity among the indicated categories: 0, pain absence; 1 to 3, mild; 4 to 6, moderate; 7 to 9, high; and 10, unbearable. During the procedure no patient was allowed to use any prophylactic medicine for headache.

Ethical considerations

All the interviewees received detailed data concerning the purpose of the research. All doubts were clarified and each volunteer freely agreed to participate by giving his or her written consent. The research was approved by the Ethics in Research Committee of the Health Science Center of the Federal University in Pernambuco, Brazil, under protocol number 089/2008 in accordance with resolution 196/96 of the National Health Council.

Statistical analysis

Data are shown as mean ± standard deviation (SD). The Kolmogorov-Smirnov test was used to verify the type of distribution of variables to be studied. When the variables showed a normal distribution, the parametric test (Student t-test) was employed; when they did not, the Mann-Whitney non-parametric test was used. Based on the analysis of the categorical variables, we applied the chi-square test according to the expected frequency on the cells. The significance level considered was 5% (p < 0.05). For processing and analysis of data, the SPSS program version 13.0 for Windows was used.

Results

Field testing of ICHD-II criteria

Of the 414 individuals tested, 153 (37.0%) had IH (106/266 women (39.8%) vs. 47/148 men (31.8%), p = 0.126 χ²; Figure 1), but only 77 of these (50.3%) satisfied the ICHD-II criteria.

Figure 1.

The bars show the percentage of individuals with ice-induced headache in the three groups: i. individuals with no previous history of primary headache (n = 35), ii. individuals with a previous history of tension-type headache (n = 139); and iii. individuals with a previous history of migraine headache (n = 240); (a) p = 0.0006 versus no headache group (odds ratio 4.447, 95% confidence interval 1.781–11.10); (b) p < 0.0001 versus tension-type headache group (odds ratio 3.076, 95% confidence interval 1.924–4.918) (χ²).

Considering the ICHD-II (2), the following examples would not receive the diagnosis of HICS: (a) non-migrainous patients with headache induced by a cold stimulus, located in a region other than the frontal one (34.2% in our series); (b) migrainous patients with headache located in a region other than frontal one, when it was not the usual site of migraine attacks (4.3% in our series); and (c) any pulsatile headache induced by a cold stimulus (41.2% in our series).

A previous history of HICS was reported by 200 (48.3%) of the individuals studied. The experimentally provoked IH was more frequent in those individuals who reported a previous history of HICS, when compared with the individuals without a previous history of HICS (128/200 (64%) vs. 25/214 (11.7%), odds ratio (OR) 13.44 (95% confidence interval (CI) 8.092–22.32), p < 0.001 χ²).

Pain location

The most common pain location was the frontal region; 93 of 153 (60.8%) volunteers who had IH reported the frontal region as being the site of the pain, and in 21 of these the pain also included the temporal region. The temporal region was affected by pain in 74/153 (48.4%) individuals. Only seven out of 153 (4.6%) individuals reported the occipital region as being the location of IH. In 53 of the 153 (34.6%) the IH occurred only in the temporal area, and in seven out of 153 (4.6%) the headache was located exclusively in the occipital region, but not in addition to the frontal pain. In 118/153 (77.1%) volunteers the IH was bilateral (Table 3).

Table 3.

Headache aspects observed among volunteers with positive ice test (n = 153).

Variables	n	%
Localization
Frontal and temporal	21	13.7
Frontal	72	47.1
Temporal	53	34.6
Occipital	7	4.6
Characteristics
Pulsatile	63	41.2
Nonpulsatile	90	58.8
Laterality
Bilateral	118	77.1
Unilateral	35	22.9
Pain intensity
Mild (1 to 3)	28	18.3
Moderate (4 to 6)	75	49.0
Severe (7 to 9)	48	31.4
Unbearable (10)	2	1.3
Mean ± SD (score)	5.5 ± 2.1
Latency
<30 s	41	26.8
30–60 s	49	32.0
1–5 min	60	39.2
>5 min	3	2.0
Mean ± SD	65.5 ± 65.4 s
Pain duration
<2 min	31	20.3
2–5 min	105	68.6
5–10 min	10	6.5
>10 min	7	4.6
Mean ± SD	207 ± 220 s

Throbbing vs. non-throbbing nature of pain

In 90 of 153 (58.8%) individuals who had IH, it was nonpulsatile (Table 3). Thus 63/153 (41.2%) subjects had a throbbing pain (50/115 (43.5%) migraine versus 11/35 (34.4%) tension-type headache subjects, p = 0.3553, χ²) (Table 4).

Table 4.

Headache aspects observed among volunteers with migraine or tension-type headache.

Variables	Migraine		Tension-type headache		p
Variables	n = 115		n = 32		p
Localization
Frontal and temporal	17	14.8	3	9.4	0.834^*
Frontal	52	45.2	16	50
Temporal	41	35.7	11	34.4
Occipital	5	4.3	2	6.3
Characteristics
Throbbing/pulsatile	50	43.5	11	34.4	0.598^*
Weight/pressure	13	11.3	3	9.4
Burning/blasted	42	36.5	13	40.6
Shock sensation	10	8.7	5	15.6
Laterality
Unilateral	24	20.9	10	31.3	0.218^†
Bilateral	91	79.1	22	68.8	0.218^†
Pain intensity
Mild (1 to 3)	17	14.8	7	21.9	0.69^*
Moderate (4 to 6)	59	51.3	16	50
Severe (7 to 9)	37	32.2	9	28.1
Unbearable (10)	2	1.7	0	0
Mean ± SD (s)	6 ± 2		5 ± 2		0.338^‡
Latency
<30 s	31	27	8	25	0.889^*
30–60 s	37	32.2	12	37.5
1–5 min	45	39.1	11	34.4
>5 min	2	1.7	1	3.1
Mean ± SD (s)	61.2 ± 51.4		69.9 ± 87		0.961^‡
Pain duration
<2 min	23	20	8	25	0.1^*
2–5 min	83	72.2	19	59.4
5–10 min	6	5.2	1	3.1
>10 min	3	2.6	4	12.5
Mean ± SD (s)	185 ± 159		269 ± 363		0.816^‡

*χ

²Pearson test; ^†χ²Yates test; ^‡Mann-Whitney test.

Pain intensity

In 50 of 153 (32.7%) individuals who had IH, pain intensity varied from high to unbearable. The pain score was 5.5 ± 2.1 (SD) (Table 3). There were no statistical differences in the pain intensity scores between migraine and tension-type headache sufferers (Table 4).

Latency for pain onset induced by ice

Latency between ice placement in contact with the palate and resulting onset of pain was from 8 s to 8 min (65.5 ± 65.4 s). In 90/153 (58.8%) individuals with IH, latency was less than 60 s. In 150/153 (98.0%) individuals with IH it was ≤5 min. Only three volunteers presented latency longer than 5 min (330 s, 320 s and 480 s, respectively) (Table 3).

Pain duration

Pain duration varied from 90 s up to 26.6 min (207 ± 220 s). It was less than 2 min in 31/153 (20.3%) and less than 5 min in 136/153 (88.9%). Ten of the 153 (6.5%) had IH with a duration of between 5 and 10 min. In seven out of 153 (4.6%) the pain lasted for more than 10 min (26.6 min, 25.6 min, 20.5 min, 18.8 min, 12.0 min, 10.8 min and 10.0 min, respectively) (Table 3).

HICS in individuals with a previous history of migraine

In 13/38 (34.2%) patients without a previous history of headache or with tension-type headache with HICS, the pain was located in a region other than the frontal one, compared to 46/115 (40%) migraine patients; in only two out of 46 (4.3%) migrainous patients with headache located in a region other than the frontal one, it was not the usual site of migraine attacks (Table 4).

The ice-induced headache test was positive in 47.9% (115/240) of the migraine sufferers and in 23.0% (32/139) of the tension-type headache sufferers (χ² p < 0.0001, OR 3.076, 95% CI 1.924–4.918) (Figure 1).

As for pain characteristics, 43.5% of migraine sufferers and 34.4% of the tension-type headache individuals, respectively, reported that their pain was throbbing or pulsatile (Table 4).

One hundred and forty-seven of the 379 (38.8%) who presented a previous history of primary headache (migraine or tension-type headache) presented IH, while only six out of 35 (17.1%) who had no history of previous headache reported pain in the test (χ² p = 0.0111, OR 3.063; 95% CI 1.241–7.557).

Of the 147 with a previous history of primary headache whose ice test was positive, 71 (48.3%) reported that IH occurred at the habitual place of their previous primary headaches.

In migraineurs, the prevalence was similar between the genders (male 46.9% vs. female 48.3%).

Discussion

The present results using ice as an experimental way of provoking HICS reveal that the ICHD-II criteria are unsatisfactory for the diagnosis of this form of cold stimulus-induced headache, since only half the volunteers who developed headache after the application of ice to the palate would be thus diagnosed using such criteria.

Is all pain induced by a cold stimulus acute frontal non-pulsatile headache? Criterion ‘A’ states that the pain induced by a cold stimulus must be acute, frontal in location and of a non-pulsatile nature. However, an exception is admitted: In the case of migrainous patients, HICS may be referred to the usual site of migraine headaches.

We found that in about one third of tension-type headache subjects the IH was located in a region other than the frontal one, and in a small number of migrainous individuals the headache was also located in a region other than the frontal one, when it was not the usual site of migraine attacks. In addition, in a substantial number of subjects a pulsatile headache was induced by ice. HICS is a bilateral pain predominantly located at the frontal or temporal region (6,8).

Criterion ‘B’ clearly establishes the obvious presence of a cold stimulus to the palate and/or posterior pharyngeal wall due to the ingestion of cold food or drink or to the inhalation of cold air. Cold refers to the condition or unpleasant perception of having a low temperature (12–16).

A number of authors suggest a neurovascular mechanism in the genesis of the HICS (9,17). Migraine is a disease of the central nervous system with vascular participation, and the trigeminal nerve has an important role in migraine pathogenesis (18). Raskin and Knittle (3) suggested that HICS could represent a model for migraine because of the close relationship between HICS and migraine. The hyperexcitability of the trigeminal pathway that occurs in migraineurs and discharges of this pathway under certain circumstances (i.e. cold stimulus) could trigger a migraine-like attack. In this regard, the susceptibility of migraineurs to develop a headache induced by ice much more easily may be a form of cold allodynia or suballodinia (12), since the same stimulus does not trigger pain so readily in non-migraine sufferers.

The transient receptor potential (TRP) ionic channels are expressed in subpopulations of sensory neurons, and it has been argued that they mediate innocuous and noxious cold sensation (12–16).

An increase in blood flow and a visible flare in the area are seen after menthol (via TRP8 ionic channels) application, indicating a neurogenic vasodilatation. This may also explain a longer duration of the pain induced by ice and, perhaps, its pulsatile nature, since vascular arterial involvement and allodynia may both occur (19–22).

The use of ice in the present model of headache may activate low temperature nociceptors not only in the palate area, but also those present on the tongue. In addition, more than half of the ice cube thaws during the test, and the icy water and the cold air present inside the oral cavity, could as well decrease the local temperature. The icy water, once ingested, would also come into contact with the posterior pharyngeal wall.

In our volunteers the latency between the contact of the ice on the palate and the onset of headache ranged from 8 s to 8 min. The vast majority (98%) presented a latency of less than 5 min. Three individuals developed HICS after 5 min, indicating that the cold stimuli can trigger headache minutes later, probably through a mechanism other than the direct activation of nociceptors by cold. We believe that 5 min is a reasonable expected lapse of time for the occurrence of a headache provoked by a cold stimulus, even though it is possible, in a few cases, for an IH to occur after a longer latency.

Ninety-seven percent of the headaches induced by ice in the present study had a duration ≤10 min. Seven of the 143 patients with IH presented an episode of headache lasting longer than 10 min. Obviously, in the case of an HICS with a duration longer than 10 min it should be borne in mind that the HICS may trigger a primary headache attack superposed (without a headache-free interval) on the HICS. In the present study this did not seem to happen with the triggering of a migraine attack, since we would have expected a duration of hours in the typical example.

The present study demonstrated that individuals with a previous history of primary headache (lifetime prevalence) are twice as likely to suffer from IH.

Only a few studies have used icy substance to provoke pain (4,5,8), and only one used ice (9). Raskin and Knittle (3) conducted their research through an interview and found prevalences of 93% and 31% in the migrainous patients and control group, respectively. Mattsson (8) induced pain by ingestion of 150 ml of icy water (0–4°C) with a straw. He found a prevalence of 8% of HICS in his sample and suggested that females with active migraine had a greater predisposition to develop HICS. Using a questionnaire, Fuh et al. (6) found a prevalence of HICS in 55.2% of migrainous patients and of 39.6% in non-migrainous ones. By provoking palate stimulation with an ice cube for 90 s, Selekler et al. (9) found a prevalence of HICS in 74% of the migrainous patients and in 32% of those with headache of the tensional episodic type.

As stated by Selekler and colleagues (9), it seems that the prevalence of HICS is influenced by the duration of the exposure and the low temperature of the stimuli. We do not believe that the 90-second duration of the stimulus used was exceedingly long, as compared to a few intraoral stimuli experienced in real-life situations of consuming a cold substance. For example, it is common for ice cream to be continuously ingested for several minutes.

We should perhaps recommend caution regarding such a generalization from our findings with experimentally provoked cold stimulus headache to cold stimulus headache in general (i.e. HICS). The circumstances here are somewhat artificial – holding an ice cube against the palate – and do not entirely replicate what occurs when people take multiple gulps of cold drinks or swallow multiple boluses of ice cream. In 48% of the subjects evaluated in the present study, a previous history of HICS was reported. And we were able to induce an IH in 64% of them. Nevertheless, the IH was much more frequent among those who reported a previous history of HICS, suggesting that the artificial nature of our experimental stimulus in some way triggered a headache in those individuals susceptible to other forms of cold stimuli. Thus, our results suggest that the ICHD should take into account that, although artificial, the induced head pain in volunteers with an experimental cold stimulus and its characteristics (i.e. IH) may represent an HICS. An important question, however, remains to be answered: Would other kinds of cold stimuli trigger different forms of HICS?

In conclusion, IH is predominantly frontotemporal, bilateral and throbbing, most commonly affecting migraine sufferers, and the ICHD-II diagnosis criteria were insufficient for classifying all patients with IH. If the IH actually represent the HICS (ICHD-II), the following modifications in the criteria, illustrated in Table 5, should be made.

Table 5.

Suggested modifications to 13.11.2 Headache attributed to ingestion or inhalation of a cold stimulus according to the International Classification of Headache Disorders 2nd Edition (ICHD II) (2).

A. Throbbing or non-throbbing pain involving frontal or temporal areas (1).

B. Cold stimulus to the palate and/or to the posterior pharynx wall, due to ingestion of icy food or drink and cold air inhalation.

C. Headache arising within 5 minutes after exposure to a cold stimulus at the palate and/or pharynx.

D. Headache that disappears within 10 minutes after removal of the cold stimulus (2).

(1) In a few cases the occipital region may be affected.

(2) The headache may, exceptionally, last up to 30 minutes. Based on the Gaussian definition that a normal range is habitually established as the range of measurements ± 2 SD above and below the mean (23).

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest statement

None declared.

References

Headache Classification Committee of the International Headache Society. Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalalgia 1988; 8: 1–96.

ICHD-II Headache Classification Subcommittee of the International Headache Society. The International Classification of Headache Disorders: 2nd edition. Cephalalgia 2004; 24: 9–160.

Raskin

Knittle

. Ice cream headache and orthostatic symptoms in patients with migraine. Headache 1976; 16: 222–225.

Bird

MacGregor

Wilkinson

. Ice cream headache—site, duration, and relationship to migraine. Headache 1992; 32: 35–38.

Kaczorowski

. Ice cream evoked headaches (ICE-H) study: randomised trial of accelerated versus cautious ice cream eating regimen. BMJ 2002; 325: 1445–1446.

Fuh

Wang

. Ice-cream headache—a large survey of 8359 adolescents. Cephalalgia 2003; 23: 977–981.

Drummond

Lance

. Neurovascular disturbances in headache patients. Clin Exp Neurol 1984; 20: 93–99.

Mattsson

. Headache caused by drinking cold water is common and related to active migraine. Cephalalgia 2001; 21: 230–235.

Selekler

Erdogan

Budak

. Prevalence and clinical characteristics of an experimental model of ‘ice-cream headache' in migraine and episodic tension-type headache patients. Cephalalgia 2004; 24: 293–297.

10.

Sjaastad

Bakketeig

. Hydrogen sulphide headache and other rare, global headaches: Vaga study. Cephalalgia 2006; 26: 466–476.

11.

Rasmussen

Olesen

. Symptomatic and nonsymptomatic headaches in a general population. Neurology 1992; 42: 1225–1231.

12.

Sand

Zhitniy

Nilsen

. Thermal pain thresholds are decreased in the migraine preattack phase. Eur J Neurol 2008; 15: 1199–1205.

13.

Cortright

Szallasi

. TRP channels and pain. Curr Pharm Des 2009; 15: 1736–1749.

14.

McKemy

. How cold is it? TRPM8 and TRPA1 in the molecular logic of cold sensation. Mol Pain 2005; 1: 16–16.

15.

Reid

. ThermoTRP channels and cold sensing: What are they really up to?. Pflugers Arch 2005; 451: 250–263.

16.

Wasner

Schattschneider

Binder

. Topical menthol—a human model for cold pain by activation and sensitization of C nociceptors. Brain 2004; 127: 1159–1171.

17.

Sleigh

. Ice cream headache. Cerebral vasoconstriction causing decrease in arterial flow may have role. BMJ 1997; 315: 609–609.

18.

Dodick

. Examining the essence of migraine—is it the blood vessel or the brain? A debate. Headache 2008; 48: 661–667.

19.

Geppetti

Rossi

Chiarugi

Benemi

. Antidromic vasodilatation and the migraine mechanism. J Headache Pain 2012; 13: 103–111.

20.

Bevilaqua-Grossi

Lipton

Napchan

. Temporomandibular disorders and cutaneous allodynia are associated in individuals with migraine. Cephalalgia 2010; 30: 425–432.

21.

Tietjen

Brandes

Peterlin

. Allodynia in migraine: Association with comorbid pain conditions. Headache 2009; 49: 1333–1344.

22.

Guy

Marques

Orliaguet

. Are there differences between cephalic and extracephalic cutaneous allodynia in migraine patients?. Cephalalgia 2010; 30: 881–886.

23.

Lang

Secic

. How to report statistics in medicine Philadelphia: Port City Press, 1997.