Clinical profile of cluster headaches in Japan: Low prevalence of chronic cluster headache,and uncoupling of sense and behaviour of restlessness

Abstract

Introduction: This study examined the clinical profile of cluster headaches in Japan and the potential presence of features peculiar to Japan. Most previous studies of cluster headaches have focused on Caucasian populations.

Methods: Subjects comprised 86 consecutive new cluster headache patients (68 males, 18 females; mean age, 38.4 ± 12.2 years; range, 17–73 years). Mean age at onset was 31.0 years and the ratio of males to females was 3.8:1.

Results: Chronic cluster headache was observed in 3.5% of patients. More than half of patients (68.9%) reported feelings of restlessness during headache episodes and 42.9% reported restless behaviour. Patients with uncoupling of feelings of restlessness and restless behaviour forced themselves to keep still. Similar findings were reported in a Taiwanese study.

Conclusion: Japanese patients in this study showed a relatively low prevalence of chronic cluster headaches, and uncoupling of a sense of restlessness and restless behaviour. These features of cluster headache may be more common in Japanese and Taiwanese patients than in Caucasian patients.

Introduction

Cluster headache is the most painful form of primary headaches, characterized by strictly unilateral pain localized in or around the eye and accompanied by ipsilateral autonomic features. Another characteristic feature of cluster headache is the circadian rhythmicity of the painful attacks, which can last 15–180 min if left untreated (1,2). Diagnostic criteria for cluster headaches were established by the International Headache Society and have been revised (3). During the worst attacks, the intensity of pain is excruciating and patients are usually unable to lie down and characteristically pace around (4–6). This characteristic was described as ‘a sense of restlessness or agitation’ in the revised edition of diagnostic criteria for cluster headache by the International Headache Society (3).

Previous studies have looked predominantly at Caucasian populations (1,2,7–13), and little data are available regarding ethnic differences in cluster headache (14). Some studies have been conducted in Japan, but have been published only in abstract form and/or in Japanese (15–17). Previous studies have demonstrated that the medications being overused differed from those in other countries and that this reflected social differences (18,19). The present study aimed to clarify the clinical profile of cluster headache in Japan and reveal any features that may be peculiar to the occurrence of this pathology in patients of Japanese origin.

Patients and methods

Patients

The study population comprised patients who had been diagnosed with cluster headache according to the 2nd edition of the International Classification of Headache Disorders (ICHD-II) (3) on first consultation or follow-up between November 2004 and January 2010. All patients underwent magnetic resonance imaging to rule out a symptomatic origin. Demographic and clinical characteristics of the 86 patients enrolled in the study are shown in Table 1.

Table 1.

Demographics of subjects with cluster headache

Patient characteristics	Total	Male	Female
	(n = 86)	(n = 68)	(n = 18*)
Age in years (mean ± SD) (range)	38.4 ± 12.2 (17–73)	37.9 ± 9.5 (18–65)	40.0 ± 19.5 (17–73)
Age at onset in years (mean ± SD) (range)	31.0 ± 13.8 (10–70)	30.1 ± 11.0 (10–65)	35.4 ± 22.4 (13–70)
Type of cluster headache
Episodic	83	68	15
Chronic	3	0	3

*

The ratio of males to female was 3.8:1.

Methods

Information was collected on age, sex, life-style, laterality and locations of headache, pain intensity, autonomic features, additional features (nausea, vomiting, photophobia, phonophobia, and sense and behaviours of restlessness during attacks), duration of attacks, and time of onset of attacks. Pain intensity was estimated using a visual analogue scale (VAS) to indicate the maximum intensity of pain by having the subject make a mark on a 100-mm line labelled with ‘no pain’ at one end and ‘the worst pain possible’ at the other. Severe pain was defined as a score between 80/100 and 100/100, defined by Torelli and Manzoni (6). Feelings of restlessness were distinguished from restless behaviour, as some patients reported feeling of restlessness but forced themselves to keep still. This study conformed to the ethical standards described in the Declaration of Helsinki.

Statistical analysis

We used descriptive statistics and reported mean ± SD for all measures, as well as proportions. We used the χ (2) test to compare categorical variables and the Mann–Whitney U-test to compare continuous variables. StatView for Windows v5.0 software (SAS Institute, Cary, NC, USA) was used for statistical analyses.

Results

Age and sex

Mean age at first consultation was 38.4 ± 12.2 years and mean age of onset was 31.0 ± 13.8 years (Table 1). Mean latency to delayed diagnosis was 7.3 ± 6.9 years (range, 0–28 years). Seventy-three patients (84.9%) had consulted other medical institutions, 11 patients (12.8%) had been referred from other doctors, and 14.0% of patients had previously been diagnosed with cluster headache. The ratio of males to females was 3.8:1. Variances of ages at first consultation and onset differed significantly between males and females (P = 0.0009, P = 0.0014: F-test); however, mean age showed no significant difference (P = 0.7229, P = 0.4255; Mann–Whitney U-test). Age distributions at onset in males and females from this study are shown in Figure 1. The peak age at onset for males was in the 3rd and 4th decades of life. However, the age-onset curve for females showed a bimodal distribution, with a major peak at 10–19 years and a second smaller peak at 60–69 years.

Figure 1.

The peak age at onset for males was in the 3rd and 4th decades of life. However, the age-onset curve for females showed a bimodal distribution, with a major peak at 10–19 years and a second smaller peak at 60–69 years.

Laterality and locations of headache

Pain was strictly unilateral in all individuals. The laterality of headache is shown in Table 2. More patients experienced only right-side attacks (45.3%) than only left-side attacks (40.7%). Right-side pain was more frequent than left-side pain in women (44.1% vs 27.8%). The most commonly reported location of cluster headache was retro-orbital (80%), followed by the temporal (52%) and occipital regions (22.1%), forehead (18.6%), upper teeth (16.3%), vertex (9.3%), nose (4.7%), jaw (4.7%), cheek (3.5%), neck (2.3%), shoulder (2.3%) and ear (2.3%).

Table 2.

Laterality of headache in study patients and comparison of pain intensity

	Total	Male	Female
Laterality (values are given as percentages)
Right-side attacks only	45.3	45.6	44.4
Left-side attacks only	40.7	44.1	27.8
Side changes within the bouts	2.3	1.5	5.6
Side changes between bouts	9.3	7.4	16.7
Side changes within and between bouts	2.3	1.5	5.6
VAS (mean ± SD) (range)	91.2 ± 11.9 (54–100)	91.8 ± 10.5 (70–100)	88.7 ± 17.3 (54–100)
Severe pain (80–100/100 on VAS)	85.7%	85.7%	85.7%

Pain intensity during attacks

Mean VAS score and the proportion of severe pain are shown in Table 2. On the VAS, 91.2% of patients with cluster headache rated the pain at the maximum intensity. A sex difference in pain intensity was not seen.

Cranial autonomic and additional features in patients with cluster headache

Lacrimation was the most consistently reported autonomic feature, followed by rhinorrhea, conjunctival injection, nasal congestion, facial sweating and ptosis (Table 3). Approximately 40% of individuals reported nausea during the acute attack, and 30.2% reported photophobia and/or phonophobia. Approximately 70% reported a sense of restlessness during the attacks, 42.9% reported restless behaviour, and 31.0% reported that physical activity exacerbated the pain (Table 3). No significant difference in pain intensity (VAS) was identified in sense of restlessness, restless behaviour or exacerbation of pain by physical activity.

Table 3.

Cranial autonomic and additional features in patients with cluster headache

	Total	Male	Female
Autonomic features
Lacrimation	66.3	64.7	72.2
Conjunctival injection	31.4	30.9	33.3
Rhinorrhea	55.8	57.4	50.0
Nasal congestion	30.2	26.5	44.4
Facial sweating	20.9	23.5	11.1
Ptosis	8.1	7.4	11.1
Additional features
Nausea	39.5	36.8	50.0
Vomiting	15.1	11.8	27.8
Photophobia	30.2	29.4	33.3
Phonophobia	30.2	27.9	38.9
Sense of restlessness	69.8	69.1	72.2
Pacing	42.9	40.9	50.0
Aggravation by physical activities	31.0	27.1	50.0

Values are given as percentages.

Duration of attacks and time of onset of attacks

Approximately half of patients (46.5%) experienced cluster attacks ranging in duration from 1 h to less than 2 h, and no difference was seen between males and females. Nearly half of patients suffered attacks at night (mostly nocturnal, 25.6%; nocturnal only, 22.1%); however, 20.9% of patients showed inconsistent times of attack onset (Table 4).

Table 4.

Duration of attacks and time of onset of attacks

	Total	Male	Female
Duration of attacks
Less than 1 hh	16.3	17.6	11.1
From 1 h to less than 2 h	46.5	47.1	44.4
From 2 h to 3 h	26.7	25.0	33.3
More than 3 h	3.5	2.9	5.6
Different with each attack	7.0	7.4	5.6
Time of onset of attacks
Day-time only	10.5	8.8	16.7
Equally nocturnal and day-time	4.7	5.9	0.0
Mostly day-time	16.3	16.2	16.7
Mostly nocturnal	25.6	22.1	38.9
Nocturnal only	22.1	25.0	11.1
Different with each cluster period	20.9	22.1	16.7

Values are given as percentages.

Discussion

Chronic cluster headache is seen in 10–21% of all cluster headache patients according to previous Caucasian reports (1,2,9), whereas chronic cluster headache was found among 3.5% of subjects in our study, and these patients with chronic cluster headache were all women. Previous Japanese studies have also reported a low prevalence of chronic cluster headache (1.2–2.9%) (15–17) and a study conducted in Taiwan reported an absence of chronic cluster headache (14). Migraine as an other primary headache appears to be less prevalent in Japan and Taiwan than in Western countries (20–22). Sakai and Igarashi (20) reported a regional difference in the prevalence of migraine in Japan, and the authors speculated on the possibility of genetic factors. Takeshima et al. (22) demonstrated that the duration of headache might tend to be shorter among Japanese migraineurs than among Caucasians and that migraineurs consume significantly more fatty/oily foods, coffee, and tea and less fish than non-headache residents based on a population-based door-to-door survey. The low prevalence of cluster headache may also be attributable to genetic or life-style factors. Torelli et al. (23) reported that heavy drinkers of alcohol and coffee were prevalent among patients with chronic cluster headache that was unremitting from onset, while patients with chronic cluster headache that evolved from episodic cluster headache were more frequently heavy smokers. Cigarette consumption is just as prevalent in Japan as in many Western countries (24,25), but alcohol and coffee consumption are lower in Japan than in many Western countries (26–28). Shizuoka City has the second lowest frequency of coffee consumption in Japan (29). Low alcohol and coffee consumption may be associated with the absence of male chronic cluster headache in our study. Chronic cluster headache patients in our study were all women, with ages at onset of 60, 63 and 70 years. Ekbom et al. (30) reported that chronic cluster headache was more frequent in women than in men (ratio, 1:0.6) over the age of 50 years. They speculated that the menstrual cycle or other endocrine factors might play a role in protecting against chronic cluster headache until middle age in women. The lack of men with chronic cluster headache may reflect the very low prevalence of chronic cluster headache in Japan, and the presence of female chronic cluster headache may be due to the ageing nature of society in Shizuoka City, particularly for females (individuals ≥ 65 years old comprise 21.5% of men and 27.2% of women in that population (31).

Our study showed that 69.8% of cluster headache patients reported feelings of restlessness. Bahra (2) found a high percentage of patients reporting a sense of agitation or restlessness (93%). Other Caucasian (9) and Taiwanese (14) studies have shown a lower prevalence of restlessness during attacks (67.6% and 51%, respectively), similar to our results. Our study revealed a discrepancy between the number of patients with feelings of restlessness (69.8%) and restless behaviours, such as pacing (42.9%). Approximately 27% of patients reported feeling restless but managed to remain still. This discrepancy did not seem to be related to intensity of pain. Pain intensity in our study appeared similar to the findings described by Torelli and Manzoni (6). The Taiwanese study (14) showed a similar discrepancy between numbers of patients reporting feelings of restlessness and those showing actual restlessness. Our study showed a higher prevalence of pain aggravated by physical activity (31.0%) than a previous Caucasian study (21.7%) (9); however, differences in results are difficult to interpret due to the low prevalence of pain aggravated by physical activity in the Taiwanese study (7%) (14). The reasons for the higher prevalence of pain aggravation caused by physical activity and for the approximately 27% of patients who remained still despite feelings of restlessness may reflect differences in ethnic, social and cultural factors. One possible reason for remaining still despite feelings of restlessness is that Eastern cultures emphasize the importance of patience and serenity whereas Western cultures may place less emphasis on such values.

Our study demonstrated that cluster headache remains unrecognized or misdiagnosed in many cases for many years. In previous published Caucasian studies, the time between the first episode and diagnosis ranged from 1 week to 48 years (median, 3 years) (32). In our study, 84.9% of patients had been admitted to other clinics, but only 14.0% had been diagnosed with cluster headache. More than half of cluster headache patients might obtain information concerning consultation with headache specialists from the internet (33). Increasing the knowledge of cluster headache among referring clinicians could improve the recognition and treatment of cluster headache. Our study showed a lower prevalence of changes in laterality within a cluster headache bout and between bouts, especially in men, compared to a Caucasian study (2). Another Japanese study also showed a low prevalence of changes in laterality both within and between cluster headache bouts (17). The low frequency of changes in laterality, particularly among men, may be a clinical characteristic in Japanese cluster headache patients.

Study limitations

One limitation of this study that must be considered is the fact that data were obtained from only a single headache clinic with a small number of patients. Aside from this limitation, a low prevalence of chronic cluster headache and the uncoupling of feelings of restlessness and restless behaviour have commonly been observed in Japanese and Taiwanese studies. These characteristics may thus reflect ethnic and/or cultural difference between Eastern and Western cluster headache patients.