0895-4356/91$3.00+ 0.00 Copyright 0 1991Pergamon Pressplc
J ClinEpidemiol Vol. 44, No. 11, pp. 1147-1157,1991 Printed in Great Britain. All rights reserved
EPIDEMIOLOGY OF HEADACHE IN A GENERAL POPULATION-A PREVALENCE STUDY BIRTHE KRGGH
RASMUSSEN,* RIGMOR JENSEN, MARIANNE SCHROLL
and
JES OLESEN Glostrup Population Studies, Department of Internal Medicine C, Glostrup Hospital, University of Copenhagen and Department of Neurology, Gentofte Hospital, University of Copenhagen, Denmark (Received
in revised form 6 June 1991)
Abstract-We present the first prevalence study of specific headache entities using the operational diagnostic criteria of the International Headache Society. One thousand 25-64 year old men and women, who lived in the western part of Copenhagen County were randomly drawn from the Danish National Central Person Registry. All subjects were invited to a general health examination focusing on headache and including: a self-administered questionnaire concerning sociodemographic variables, a structured headache interview and a general physical and neurological examination. The participation rate was 76%. Information about 79% of the non-participants showed a slightly differing headache prevalence which was not quantitatively important. The following results in participants are therefore representative of the total sample. The lifetime prevalences of headache (including anybody with any form of headache), migraine, and tension-type headache were 93, 8 and 69% in men; and 99, 25 and 88% in women. The point prevalence of headache was 11% in men and 22% in women. Prevalence of migraine in the previous year was 6% in men and 15% in women and the corresponding prevalences of tension-type headache were 63 and 86%. Differences according to sex were significant with a male : female ratio of 1: 3 in migraine, and 4: 5 in tension-type headache. The prevalence of tension-type headache decreased with increasing age, whereas migraine showed no correlation to age within the studied age interval. Headache disorders are extremely prevalent and represent a major health problem, which merits increased attention. Headache
Migraine
Tension-type
headache
INTRODUCTION
Headache disorders are extremely common. Usually they are mild or infrequent but when they are severe or frequent, they cause considerable suffering as well as decreased or abolished working capacity. The prevalence of headache has long been a subject of discussion. Investigations of prevalences in selected groups (general practice, hospitals, clinics, schools and *All correspondence
should be addressed to: Dr Birthe Krogh Rasmussen, The Glostrup Population Studies, Department of Internal Medicine C, Elevboligen 7. sal, Glostrup Hospital, DK-2600 Glostrup, Denmark.
Epidemiology
Prevalence
different occupational groups) are numerous (e.g. [l-7]), whereas studies in general populations are fewer [8-141. The major problem in any study of the epidemiology of headache has been that of defining the disease entities. All previous studies have used earlier non-operational diagnostic criteria which have been outdated for several years [ 151.They were vague and widely open to individual interpretation. The international headache classification of the International Headache Society [ 161 gives operational diagnostic criteria for all headache disorders. Using these criteria, the present paper reports prevalence rates of the major headache
1147
1148
disorders general reference headache
BIIUHEKROGIIbMWSEN
and their relation to age and sex in a population. The study provides a material for future investigations of in selected groups. MATERIALS AND METHODS
et al.
and hence there were fewer self-employed. On the other hand, compared to Denmark as a whole, trades and services were over-represented as were salaried employees. Invitation
All members of the sample were invited to a general health examination by a standard letter On 23 November 1988, a random sample of containing information about the project. The 1000, 25-64 year old men and women residing project was consciously presented in very broad in 11 municipalities around the Copenhagen terms in order to avoid appreciable bias in County Hospital in Glostrup was drawn from relation to headache. It was stated that the aim the National Central Person Registry. The was to study the distribution of some important sample size was chosen to ensure adequate disorders in the general population, especially numbers of the most frequent forms of headache. The invitation was framed as an offer headache disorders. The total population of of a thorough health examination and the 1000 subjects was reduced to 975 because of importance of participation of all subjects decease (n = 3) and emigration (n = 17) in the invited was emphasized. A detailed, self-adminperiod between the date of sampling and the istered questionnaire was included to evaluate date of planned examination and an error in the demographic and psychosocial variables. If address lists (n = 5). Thus, only 975 persons necessary, examination outside working hours could be invited. A total of 740 persons (75.9%) or free transportation was offered. If the first were examined (Fig. 1). The total population of invitation evoked no response, a second was the sampling area was 325,621, which is 54% of issued. Those who still did not respond or who the Copenhagen County population and 6% of refused to participate were interviewed by the total Danish population. The representativetelephone. When they could not be reached by ness of the sampling area could be assessed by telephone, a third letter was mailed asking them comparing inhabitants in the sampling area, in to contact us by telephone or alternatively to return a postal questionnaire. Invited persons the Copenhagen County, and in all of Denmark using data from the National Statistics [17, 181. who did not turn up to the examination are The sampling area was representative of the labeled non-participants. The questions to the total Danish population as regards the age- and non-participants were selected from those of the and the information main investigation sex-distribution and marital status. Regarding obtained by telephone interviews or mailed industrial categories and employment status, questionnaires was recorded in order to evaluate farming and fishing were under-representated both in the sampling area and in the if non-participants differed from participants. Since late respondents might be more similar to Copenhagen County (0.5 vs 6.3% in Denmark), Sampling and representativeness
After second or third invitation
interviewed
Answered postal queationnalre
No infomatlon
Fig. 1. Diagram showing the distribution of the population according to type of participation.
Epidemiology of Headache
non-participants than early respondents, a comparison was made between the overall prevalences of the headache disorders in participants, who participated after the first, second or subsequent invitations. Subjects were informed that all data would be used for research purposes only and would be kept strictly confidential. The project was approved by the ethical committee for Copenhagen County. Examination
The study took place at the Glostrup Population Studies between January and July 1989. An average of 7 participants were examined per day each taking 3-4 hours. The participants were asked not to eat, drink or smoke during the last 8 hours before the examination which started in the morning. After approximately 1 hour they were offered a light meal. The questionnaire was checked for obvious errors and omissions, and the occupation and the education was recorded according to the International Standard Classification (ISIC) [ 191 with the modifications made by the Danish Statistical Department. A number of laboratory investigations were included in order to study e.g. the importance of myogenic mechanisms for the different forms of headache. The headache disorders were classified according to a structured diagnostic headache interview and a general physical as well as a neurological examination as required by the new diagnostic criteria. In the interview, questions about the occurrence of headache were phrased as follows: “Have you ever had a headache?” (lifetime prevalence) and “Do you have a headache today?” (point prevalence). The term headache included all forms of headache (migraine, tension-type headache, hangover, headache
1149
associated with fever etc.). In the following, subjects with at least one form of headache constitute the group labeled “headache”. The interview included questions about the different forms of headache and an extensive description of the headache frequency, duration, location, severity, character of pain, accompanying symptoms etc. The number of days per year with each kind of headache was recorded. The exact wording and details of the replies to individual questions about various features of migraine and tension-type headache are given in Table 1. Data about quality and quantity of each distinct headache form within the same subject were recorded. One physician took care of the clinical interview and examination for the first 2 months (R.J.), the other for the remaining 5 months (B.K.R.). Both were residents with more than 2 years of neurological training and a special interest in headache. Diagnostic criteria
All headache disorders were classified according to the operational diagnostic criteria of the International Headache Classification [ 161. The classification is hierarchically constructed. It contains 13 diagnostic groups which are subdivided to allow for coding up to a 4-digit level. It is thus possible to use the classification at different levels of sophistication. In the present paper, we present the prevalences and age- and sex-distribution of the primary headaches using the classification at the l-digit level in order to ensure adequate numbers in the subgroups. The main grouping of migraine and tension-type headache and the operational diagnostic criteria for migraine without aura (previously called common migraine) and episodic tension-type headache are given in Table 2 in order to
Table I. Interview questions used for classification of migraine and tension-tyne headache State number of migraine/tension-type headache attacks during your life until now: [l-4; 5-9; 10 or more] For how many days during the last 12 months have you suffered from migraine/tension-type headache? [O days; l-7 days; 8-14 days; 15-30 days; 31-180 days; more than 180 days] (4 Usual duration of the migraine/tension-type headache if you do not take any medicine, or if it does not work: [(l) less than 30 min; (2) between 30 min and 4 hr; (3) between 424 hr; (4) between 24-72 hr; (5) between 3-7 days; (6) more than 7 days; (7) varying from less than 30 min to more than 7 days] (4 Usual location of the migraine/tension-type headache? [(l) pain in the right part of the head only; (2) pain in the left part of the head only; (3) alternately in the right and left lateral part of the head; (4) alternately bilateral and unilateral; (5) always bilateral; (6) varies a lot] Subjects are classified with unilateral headache if they confirmed one of the answers (l), (2) or (3) (4 Which of the following types ofpain is the most characteristic description of your migraine/tension-type headache? [(l) pulsating pain; (2) pressing, tightening pain; (3) stabbing pain; (4) other] (f) How is the pain of the migraine/tension-type headache usually in case you do not take any medicine, or if it does not work? [(l) mild pain-daily activities (incl. housework) not inhibited; (2) moderate pain-inhibiting, but not preventing daily activities; (3) severe pain-daily activities suspended] (9) Does your migraine/tension-type headache get worse by climbing or walking down stairs? [yes; no] 09 Is your migraine/tension-type headache accompanied by: nausea? vomiting? loss of appetite? photophobia? phonophobia? [yes; no]
I%
1150
B~RTHEKRCIGHRASMUSSEN et al.
Table 2. The grouping of migraine and tension-type headache according to the International Heada& Class&a&m and the diagnostic criteria for migraine without aura and episodic tension-type headache (I) Migraine Migraine without aura (previously common migraine) Migraine with aura (previously classic migraine) (1.2) (1.3-1.7) Other migrainous disorders (1.1)
[16]
(2) Tension-type headache (2.1) Episodic tension-type headache (previously muscle contraction headache, stress headache etc.) (2.2) Chronic tension-type headache (2.3) Tension-type like headache
Diagnostic criteria for migraine without aura:
(A) At least 5 attacks fulfilling (BHD) (B) Headache attacks lasting 4-72 hr (C) Headache has at least two of the following characteristics: (1) Unilateral location (2) Pulsating quality (3) Moderate or severe intensity (inhibits or prohibits daily activities) (4) Aggravation by walking stairs or similar routine physical activity (D) During headache at least one of the following: (1) Nausea and/or vomiting (2) Photophobia and phonophobia Diagnostic criteria for episodic tension -type headache
:
(A) At least 10 previous headache episodes fulfilling criteria (B)-(D) listed below. Number of days with such headache < 180/year (B) Headache lasting from 30min to 7 days (C) At least two of the following pain characteristics: (1) Pressing/tightening (non-pulsating) quality (2) Mild or moderate intensity (3) Bilateral location (4) No aggravation by walking stairs or similar routine physical activity (D) Both of the following: (1) No nausea or vomiting (anorexia may occur) (2) Photophobia and phonophobia are absent, or one but not the other is present
exemplify the criteria used. The last criterion for all primary headaches (not listed under each form in the table) includes the ruling out of possible organic causative factors by history, physical and neurological examinations or other appropriate investigations. It is important to notice that in the classification system subjects receive a diagnosis for each distinct headache form experienced. Thus, subjects with e.g. both migraine and tension-type headache are included in the diagnostic group of migraine as well as that of tension-type headache. Data processing and statistical methods
All data were coded daily using the SPSS (Statistical Package of Social Sciences) dataentry programme. Data were transferred to the SCIBAS computer system at Copenhagen University, Herlev Hospital. In the present publication the chi-square test with a 5% level of significance has been used, and a Mantel-Haenszel summary x2 test (M-H test) [20] was used to control for age confounding. Ninety-five per cent exact confidence intervals using the binomial distribution were used.
RESULTS
Response pattern
The participation rate was 76%. The response pattern in the study population according to age and sex appears in Table 3. Concerning the response rates, none of the age- or sex-differences were statistically significant. Participants who responded to the first, second or third invitation differed as regards age and sex, since there were more young men who required a second or third invitation before participating (sex difference: M-H test = 10.67, df = 1, p = 0.001; age difference (d): x2 = 6.76, df= 3, p = 0.079). Of the study population 12.3% (n = 120) were interviewed by telephone, 6.7% (n = 65) answered a postal questionnaire and 5.1% (n = 50) were unreachable or unwilling to give any information. Among non-participants (n = 235) information was thus obtained from 78.7% (n = 185). Prevalence of headache, migraine and tensiontype headache
The lifetime prevalence of headache (including both migraine, tension-type headache
1151
Epidemiology of Headache Table 3. Response pattern among 975 subjects in Copenhagen County Non-participants
Participants* (%)
Sex and age (yr)
Telephone interview (%)
Postal questionnaire (%)
Others (%)
Total No. in sample N = 975
Males
25-34 35-44 45-54 5564 Mean
74.6 83.0 78.5 73.8 77.9
7.6 10.6 13.3 13.6 11.3
6.8 3.6 6.7 5.8 5.6
11.0 2.8 1.5 6.8 5.2
118 141 135 103
Females 25-34 3544 45-54 554 Mean
76.5 74.3 72.4 71.9 73.9
10.1 16.2 15.8 10.4 13.4
8.4 5.9 5.5 12.5 7.7
5.0 3.7 6.3 5.2 5.0
119 136 127 96
Mean (both sexes)
75.9
12.3
6.7
5.1
*Sex difference: M-H test = 2.17, df = 1, p = 0.14. Age difference: 8, x2 = 3.91, df = 3, p = 0.27. ?, X2 = 0.76 p = 0.86.
and all other types) was 96% (709/740) and was significantly higher among females (99%) than among males (93%) (Fig. 2). Men aged 55-64 had lower prevalence of headache than men in the other age groups. Point prevalence rates were significantly higher among %A
lco-
9 All headaches combined~=.
90-
d
so-
9
7060d 50-
11
40.
30-
t;
_---
___--
zo-
9
Migraine ---
lo-
~:::yY:
’
wk‘l
--_
35:44
+
----
45kl
1
df= 3,
women compared to men with a male/female ratio of 1:2 (Table 4). The overall lifetime prevalence of migraine was 16% (119/740), among men 8% (30/387) and among women 25% (89/353). Lifetime and last year prevalence of migraine were significantly higher among women than among men (Fig. 2, Table 5). Male/female ratio was about 1: 3. There were no significant differences in migraine prevalence rates according to age. The overall lifetime prevalence of tensiontype headache was 78% (578/740), among men 69% (266/387) and among women 88% (312/353). Lifetime and last year prevalences of tension-type headache were significantly higher among women than among men with a male/female ratio about 4: 5 (Fig. 2, Table 5). Men aged 55-64 had lower lifetime- and last year prevalence of tension-type headache than men in the other age groups. Among women there were significantly decreasing lifetime- and last year prevalences of tension-type headache with increasing age.
0
55:&a
A&
Fig. 2. Lifetime prevalences of headache, tension-type headache, and migraine and 95% confidence intervals. Relation to age and sex. Headache: Sex difference: M-H test = 12.91, df = 1, p < 0.001. Age difference: 3, x2 = 5.08, df = 3, p > 0.1. 0, x2 = 2.23, df = 3, p z 0.5. Migraine: Sex difference: M-H test = 41.81, df = 1, p < 0.09001. Age difference 8, x2 = 0.31, df = 3, p > 0.9. 0, x2 = 1.99, df = 3, p > 0.5. Tension-type headache: Sex difference: M-H test = 42.09, df = 1, p < 0.00001. Age difference g, x2 = 9.99, df = 3, p < 0.02. 9, x2 = 13.53, df = 3, p < 0.004.
Table 4. Point prevalence of headache and 95% CL Point prevalence Men Age group 25-34 35-44 45-54 55-64 All ages
(;=$
$;$
Women
All (N = 740) % (CL)
13 (6-21) 10 (5-17) 10 (5-18) 9 (4-18)
31(2141) 22 (14-31) 13 (7-22) 22 (13-33)
22 (1629) 16 (11-21) 12 (8-17) 15 (l&22)
11 (8-14)
22 (18-26)
16(13-19)
Sex difference: M-H test = 16.69, df = 1, p < 0.0001. Age difference: 6, x2 = 0.51, df = 3, p > 0.9. 2, x2 = 8.43, df = 3, p < 0.05.
BIRTHEK~OGH RASMUSSEN ef al.
1152
Table 5. Prevalences in the previous year of migraine and tension-type headache and 95% CL Migraine
Tension-type headache
Age group
Men (n = 387) % (CL)
Women (n = 353) % (CL)
All (N = 740) % (CL)
(;O;$
Women (n = 353) % (CL)
(N :;40) % (CL)
25-34 354l 45-54 554
5 (l-l 1) 7 (3-13) 6 (2-12) 7 (2-15)
18 (10-27) 14 (8-22) 12 (6-20) 19 (10-30)
11 (7-17) 10 (615) 9 (5-13) 12 (8-19)
68 (57-78) 63 (5472) 70 (60-78) 49 (3760)
93 (86-98) 92 (8S97) 82 (72-89) 74 (62-84)
81 (75-86) 77 (7&82) 75 (69-81) 61 (52-69)
All ages
6 (4-9)
15 (12-19)
10 (8-13)
63 (58-68)
86 (82-90)
74 (71-77)
Men
Migraine: Sex difference: M-H test = 17.10, df = 1, p < 0.0001. Age difference: 6, x2 = 0.55, df = 3, p > 0.9. 0, x2 = 1.99, df = 3, p > 0.5. Tension-type headache: Sex difference: M-H test = 50.53, df = 1,p < 0.00001. Age difference: 6. x2 = 9.83, df = 3,p < 0.03. 9, x2 = ii.27, df = 3, p < 0.001.
None of the migraineurs had migraine on the day of examination; the prevalence of migraine in the last week before the examination was 2% (13/740) and in the last month 4% (29/740). The point prevalence of tension-type headache was 12% (91/740), the last week prevalence was 29% (212/740) and the prevalence in the last month was 48% (358/740). Among the 119 subjects who ever have had migraine, 103 (87%) had in addition at some time also had tension-type headache. Among subjects with migraine in the previous year 83% had coexisting tension-type headache. Considering subjects with migraine in the last month 62% had coexisting tension-type headache. 46% of subjects with migraine in the last week had in addition had tension-type headache in this period. Of the migraine group 15% had it 8-14 days a year and 9% had it more than 14 days a year. In the total population 2% had migraine for more than 14 days a year (Fig. 3). As regards the severity of the migraine, 85% of the migraineurs had a severe pain intensity, 14% a moderate pain and only 1% a mild pain. Of the tensiontype headache group 23% had it 8-14 days a year and 36% several times a month. In the total population 44% had tension-type headache for 1-14 days a year and 3% had chronic tensiontype headache (i.e. headache 2 180 days a year) (Fig. 3). Among subjects with tension-type headache, the pain was severe in l%, moderate in 58% and mild in 41%. We found only one case of cluster headache. Non -participants The 185 subjects interviewed by telephone or answering the postal questionnaire stated their reason for not attending the investigation. The main reasons were reluctance to lose a working day (37.5%); attention to family (7.8%); illness
(15.6%); being too healthy or having had a recent health examination (4.7%); fear of hospitals (19.5%); fear of diagnosis (10.2%) and fear of registries (4.7%). No significant differences between participants and non-participants were found regarding age, sex, and marital status. Comparisons of the answers to identical questions by the participants and by subjects interviewed by telephone or answering the postal questionnaire showed no significant differences as regards hospital admissions or self-evaluated health within the last year. Likewise present status of employment, education, and nationality were not significantly different. The lifetime prevalences of headache, migraine and tension-type headache among the non-participants who answered the telephone interview or a postal questionnaire were 92, 13 and 70%, respectively, compared to 96, 16 and 78% in participants. These differences, although slight, were statistically significant as regards headache (OR = 1.99 (1.05-3.78)) and tensiontype headache (OR = 1.55 (1.08-2.23)). No difference in the overall prevalence of headache, migraine, and tension-type headache was found between subjects who participated after the first, second or subsequent invitation.
DISCUSSION
Methodological considerations
All members of the sample were invited to a general health examination, but with the main emphasis on headache disorders. Therefore, an over-representation of persons with headache could be expected among participants. The of headache sufferers over-representation among the participants compared to non-participants could indicate that the observed
1153
Epidemiology of Headache
al %
b) %
40
8
-
7
6
30
20
1c
TENSION-TYPE
C) %
di %
40
40.
30
30.
20
20.
10
10.
HEADACHE
q
Fig. 3. Number of days with migraine/tension-type headache in the previous year. (a) percentage of those with migraine; (b) percentage of population; (c) percentage of those with tension-type headache; (d) percentage of population. 0 indicates no attacks in the previous year but earlier presence of the disorder. (Note y-axis not the same in (b))
prevalences are a little too high for the whole sample. On the other hand, the difference may be due to the different methods of data collection in participants and non-participants. Even assuming that the sampling method did not influence the prevalence results, the reported discrepancy in headache prevalence is, however, minimal because of the high participation rate. If we presume that the lifetime prevalences of headache, migraine, and tension-type headache
in non-participants of 92, 13 and 70% were valid for all non-participants, the estimated prevalences in the total study population would be 95, 15 and 76%, respectively, which is not appreciably different from the values in participants of 96, 16 and 78%. The overall morbidity and various sociodemographic parameters were uniform for participants and non-participants. Thus, the examined population can be accepted as representative of the total study population.
1154
BIRTHEKROGHRASMUSSEN et al.
The sampling area showed only minor differences compared with Copenhagen County and can probably be considered representative of this area. The most marked difference as compared to all of Denmark was an underrepresentation of industries of farming and fishing. This difference was expected and is probably without importance since a previous study found very little difference between urban and rural areas as regards headache prevalence
WI. Prevalence results
Since headache is a subjective complaint without any laboratory correlate, a study of its prevalence must rely exclusively on information given by the subjects, The threshold of recall varies probably with the severity and recency of the condition. We found very high prevalence rates of headache and of tension-type headache. Only 4% of the total population had never had a headache. Using the usual 95% confidence limits of normality, it is strictly speaking abnormal never ever to have had a headache. More important is the fact that it is meaningless to talk about tension-type headache as a disease without specifying how often it occurs. Further analysis is necessary to decide a frequency of tension-type headache beyond which people can be regarded as sufferers. Even if many had rare tension-type headache, 14% of the population had more than 30 headache days a year, and 3% had it more than half of the time. Frequent migraine was relatively rare according to our results (l-2%), but 7% of the migraineurs were in prophylactic treatment, which probably has reduced the frequency. Most migraine sufferers had l-7 migraine days a year and the number of migraineurs decreased with increased number of migraine days a year. Women had higher prevalences than men migraine, and tension-type of headache, headache, which stresses the importance of considering the different forms of headache as sex dependent. Tension-type headache showed significantly declining lifetime- and 1-year prevalences with increasing age, whereas the small numbers in the migraine group prompts caution in evaluating subgroups. One would expect lifetime prevalences to show a cumulative trend, increasing with increasing age. In contrast, we found lower lifetime prevalence in the older age groups. This could be explained by a cohort effect with the older age groups being more reluctant to
confirm that they have had headache. Secondly, it could be due to recall bias. Greater exposition to headache provoking job situations in the younger or increasing headache incidence in general are other theoretical possibilities. A final conclusion demands further elucidation including a follow up of the cohort in 5 or 10 years. The point prevalence for migraine of zero was probably artificially low, because persons who happened to have migraine on the day of appointment stayed at home and got a later appointment. The high point prevalence of headache may have been increased by the required 8 hours of fasting, abstinence from smoking, and anxiousness because of the examination. Our results of different time-period prevalences showed the expected increase with increasing observation period. Present results in relation to previous studies
The present study is the first investigation performed in a representative random population, using operational diagnostic criteria, and including a clinical interview as well as a general and a neurological examination of all participants. Only a few earlier studies have examined a random sample of the general population [8-141, whereas selected materials have often been used. Some studies used questionnaires [ 1,4,5, 12, 14,21,22] others clinical interviews [lo, 231. A physician did the interview in some previous studies [9, 10,231, lay interviewers in others [24,25]. Age and sex varied in previous studies. Another essential problem is to recognize the influence of how the questions about headache have been phrased. Some asked: “Do you suffer from headache?” [1,4, 111, others asked: “Do you (ever) have headache?’ [IO, 12, 14,22,23]. Presumably, the inclusion of the words “suffer from” gives a lower prevalence than “have”. Some reports concern lifetime prevalences [1, 10,22,23] others different time-period prevalences [5,7, 10, 13, 14,251, and some do not specify the period [l 1,261. The varying methods in earlier population studies have resulted in reports of a wide range of prevalences of the different forms of headache and have impeded systematical comparison. Thus, comparing our results to other studies presents great difficulties and should be done with caution. Nevertheless, the most comparable studies show agreement with the present results concerning prevalence rates of headache [22,23] (Table 6).
Doctors
Dalsgaard-Nielsen,
General population
Holhragel/Norrelund,
General population
General practice
General population
Shopping center
Nikiforow, 1981, Finland [23]]]
Philips, 1976, U.K. [5j$
Rasmussen et al., 1991, Denmark, (present study) 11
1980, U.S.A. [21]$
1971, U.K. [27lt
Schnark/Hunter,
Waters/O’Connor,
General population
Church congregations
Waters, 1974/75, U.K. [14]$
Ziegler, 1977, U.S.A. [22]#
*Lay-administered interview and clinical interview of small subgroup. tQuestionnaire and clinical interview of small subgroup. $Questionnaire. §Telephone interview. l/Lay-administered interview. ]]Clinical interview and examination.
General practitioners
Waters, 1972, U.K. [q$
General population (women)
Women serving as control group for cerebrovascular disease
Markush et al., 1975, U.S.A. [24]7
1980, Denmark [I l]$
Non-clinic, telephone sampled
Duckro et al., 1989, U.S.A. [25j5
1973, Denmark [I]$
General population
General population
Crisp et al., 1977, U.K. [lO]t
Sample source
D’Alessandro er al., 1988, Italy [8]*
Study
1809
1718
882
2933
1293
740
597
200
451
1052
500
461
727
1154
Respondents (No.)
>15
>21
35-54
2&64
Adults
2564
16-60
>15
1544
40
>21
2585
Adults
>7
Age
Lifetime
Previous 1 year
Previous 1 year
Previous 1 year
Previous 1 year
Lifetime Previous 1 year Point
Previous 6 months
Lifetime Previous 1 year
Previous 1 year
?
Lifetime Past 3 months
Lifetime
Lifetime Previous 2 year
Previous 1 year
Time-period prevalence
82.6
63.5
84.2
78.4
25.7 19.5
18.0
%F 16.1
14
8 6 0
11
8
18
26
19
25 15 0
10
35
23.5
16 10 0
21
24.5
13
61
88 86 16
65
42
34.5
(mild headache) 41.7 34.0
42
69 63 9
37
28.8
%F
78 74 12
68
40
% Both
Tension-type headache % Both %M
14 22 16 (very severe headache) 11.2 20.4 15.8 8.8 15.2 12.0
9.8 8.9
9.3
%M
Migraine
14.9 23.2 (disabling headache) 40.9 50.2
7.3
16
22 11
78.7
96
89
91 77
77
%Both
99
74
83
76.5
85
93.7
46.2
%F
93
69
69
68.9
35.3
%M
Headache
Table 6. Some prevalence studies of different forms of headache in America and Western Europe
!? e Q
s
3 _. g
am $I
BIRTHE KRUGH RASMUSSENet al.
1156
Our male/female ratio among migraine sufferers is in good agreement with the most comparable earlier studies [9, 10,231. Concerning prevalence of migraine in the previous year several studies [5,23,24,27] have found higher prevalences than the present study. However, Waters [27] used a questionnaire technique and only estimated prevalences without giving confidence limits. The survey of Philips [5] was in a selected population and had a very low response rate; Markush [24] studied a non-representative sample of young women; and Nikiforow [23], although using a clinical interview and examination, studied only 200 persons. Furthermore, the 200 persons were sampled from the 77% who primarily had answered a postal headache questionnaire, and a moderate selection bias can be suspected. Our report of no correlation between migraine prevalence and age within the studied age interval (25-64) is in agreement with previous studies. In addition, these studies have shown the prevalence of migraine to decrease in subjects above the age of 64 [6,22,23,28]. Reported prevalences of tension-type headache vary. Previous studies [6,23] have reported lower prevalence rates than those found in the present study. Regarding the influence of the diagnostic criteria, it is tempting to suggest that some persons with tension-type headache may have been diagnosed as migraine in the past. The reported female preponderance and the decreasing prevalences with increasing age in tension-type headache are confirmed in some other studies [2,3,6, 14,281, even if they used other diagnostic criteria. Philips’ finding of almost equal occurrence of tension-type headache in men and women [5] seems not to hold up. CONCLUSION
The present study is the first prevalence study of specific headache entities based on a structured interview and examination by a neurologist and using internationally accepted operational diagnostic criteria in a large random sample. The representativeness of the sample improves the possibility of genemlizing prevalence rates to other populations. Our study emphasizes the important health problem that the headache disorders constitute in the general population, and it provides a basis for future studies of headache in selected groups with the purpose of revealing possible risk factors for headache.
Acknowledgements-This study received statistical support from The Danish Medical Research Council (i. No. 5.29.00.60) and was supported by grants from the Danish Health Insurance Foundation (H 1l/238-88, H 1l/262-89), The Lundbeck Foundation (86/88), GLAXO (1989), DAK medica (379), The Foundation for Experimental Research in Neurology (1988), and the Danish Migraine Society (1988).
REFERENCES 1.
2.
3. 4.
5. 6.
7. 8.
9. 10.
11. 12. 13. 14. 15. 16.
17. 18. 19. 20. 21. 22.
Dalsgaard-Nielsen T, Ulrich J. Prevalence and heredity of migraine and migrainoid headaches among 461 Danish doctors. Headache 1973; 12: 168-172. Dhopesh V, Anwar R, Herring C. A retrospective assessment of emergency depaitment patients with comolaint of headache. Headache 1979: 19: 3742. Friedman AP, von Starch TJC, Merritt HH. Migraine and tension headaches: a clinical study of two thousand cases. Neuroloev 1954: 4: 773-778. Ogunyemi AO. Preva&ce df headache among Nigerian University students. Headache 1984; 24: 127-130. Philips C. Headache in general practice. Headache 1977; 16: 322-329. Waters WE. Headache and migraine in general practitioners. The Migraine Headache and Dixarlt. Bracknell, England: Boehringer Ingelheim; 1972: 3144. Waters WE. The Epidemiology of Migraine. Bracknell, England: Boehringer IngelhGm; 1974. D’Alessandro R. Benassi G. Lenzi PL ef al. Eoiderniology of headache in the Republic of San Marino. J Neurol Neurosurg Psychiat 1988; 51: 21-27. Brewis M, Poskanzer DC, Rolland C et al. Neurological disease in an English city. Acta Neural Seand 1966; 42 (Suppl. 24): l-89. Crisp AH, Kalucy RS, McGuinness B, Ralph PC, Harris G. Some clinical, social and psychological characteristics of migraine subjects in the general population. Postgrad Med J 1977; 53: 691-697. Hollnagel H, Nsrrelund N. Headache among 40 yearolds in Glostrup. Ugeskr Laeger 1980; 142: 3071-3077. Nikiforow R, Hokkanen E. An epidemiological study of headache in an urban and a rural population in northern Finland. Headache 1978; 18: 137-145. Waters WE. Community studies of the prevalence of headache. Headache 1970; 9: 178-186. Waters WE. The Pontypridd headache survey. Headache 1974; 14: 81-90. Ad Hoc Committee on Classification of Headache. Classification of headache. JAMA 1962; 179: 717-718. Headache Classification Committee of the International Headache Society. Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalnlgla 1988; 8 (Suppl. 7). Danmarks statistik. Befolkningen i kommunerne 1. januar 1989. Statistisk Tabelvark 1989. Copenhagen; 1989. Danmarks statistik. Arbejdsmarked. Statistiske Efterretninger. 1989; 1. United Nations. International Standard Industrial Classification, Statistical Papers 1968; Series M, No. 4, Rev. 2, Add. 1. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Nat1 Cancer Inst 1959; 22: 719-748. Schnarch DM, Hunter JE. Migraine incidence in clinical vs nonclinical populations. Psychosomaties 1980; 21: 314-325. Ziegler DK, Hassanein RS, Couch JR. Characteristics of life headache histories in a nonclinic population. Neurology 1977; 27: 265-269. 1
Epidemiology of Headache 23.
Nikiforow R. Headache in a random sample of 200 persons: A clinical study of a population in northern Finland. Cephalalgia 1981; 1: 99-107. 24. Markush RE, Herbert RK, Heyman A, O’Fallon WM. Epidemiologic study of migraine symptoms in young women. Neurology 1975; 25: 430-435. 25. Duckro PN, Tait RC, Margolis RB. Prevalence of very severe headache in a large US metropolitan area. Cephalalgia 1989; 9: 199-205.
26.
1157
Levy LM. An epidemiological study of headache in an urban population in Zimbabwe. Headache 1983; 23: 2-9. 27. Waters WE, O’Connor PJ. Epidemiology of headache and migraine in women. J Neural Neurosurg Psychiat 1971; 34: 148-153. 28. Waters WE. Epidemiological data relevant to prognosis in migraine in adults. Dan Med BuB 1975; 22: 89-91.
J ClinEpidemiol Vol. 44, No. 11, pp. 1147-1157,1991 Printed in Great Britain. All rights reserved
EPIDEMIOLOGY OF HEADACHE IN A GENERAL POPULATION-A PREVALENCE STUDY BIRTHE KRGGH
RASMUSSEN,* RIGMOR JENSEN, MARIANNE SCHROLL
and
JES OLESEN Glostrup Population Studies, Department of Internal Medicine C, Glostrup Hospital, University of Copenhagen and Department of Neurology, Gentofte Hospital, University of Copenhagen, Denmark (Received
in revised form 6 June 1991)
Abstract-We present the first prevalence study of specific headache entities using the operational diagnostic criteria of the International Headache Society. One thousand 25-64 year old men and women, who lived in the western part of Copenhagen County were randomly drawn from the Danish National Central Person Registry. All subjects were invited to a general health examination focusing on headache and including: a self-administered questionnaire concerning sociodemographic variables, a structured headache interview and a general physical and neurological examination. The participation rate was 76%. Information about 79% of the non-participants showed a slightly differing headache prevalence which was not quantitatively important. The following results in participants are therefore representative of the total sample. The lifetime prevalences of headache (including anybody with any form of headache), migraine, and tension-type headache were 93, 8 and 69% in men; and 99, 25 and 88% in women. The point prevalence of headache was 11% in men and 22% in women. Prevalence of migraine in the previous year was 6% in men and 15% in women and the corresponding prevalences of tension-type headache were 63 and 86%. Differences according to sex were significant with a male : female ratio of 1: 3 in migraine, and 4: 5 in tension-type headache. The prevalence of tension-type headache decreased with increasing age, whereas migraine showed no correlation to age within the studied age interval. Headache disorders are extremely prevalent and represent a major health problem, which merits increased attention. Headache
Migraine
Tension-type
headache
INTRODUCTION
Headache disorders are extremely common. Usually they are mild or infrequent but when they are severe or frequent, they cause considerable suffering as well as decreased or abolished working capacity. The prevalence of headache has long been a subject of discussion. Investigations of prevalences in selected groups (general practice, hospitals, clinics, schools and *All correspondence
should be addressed to: Dr Birthe Krogh Rasmussen, The Glostrup Population Studies, Department of Internal Medicine C, Elevboligen 7. sal, Glostrup Hospital, DK-2600 Glostrup, Denmark.
Epidemiology
Prevalence
different occupational groups) are numerous (e.g. [l-7]), whereas studies in general populations are fewer [8-141. The major problem in any study of the epidemiology of headache has been that of defining the disease entities. All previous studies have used earlier non-operational diagnostic criteria which have been outdated for several years [ 151.They were vague and widely open to individual interpretation. The international headache classification of the International Headache Society [ 161 gives operational diagnostic criteria for all headache disorders. Using these criteria, the present paper reports prevalence rates of the major headache
1147
1148
disorders general reference headache
BIIUHEKROGIIbMWSEN
and their relation to age and sex in a population. The study provides a material for future investigations of in selected groups. MATERIALS AND METHODS
et al.
and hence there were fewer self-employed. On the other hand, compared to Denmark as a whole, trades and services were over-represented as were salaried employees. Invitation
All members of the sample were invited to a general health examination by a standard letter On 23 November 1988, a random sample of containing information about the project. The 1000, 25-64 year old men and women residing project was consciously presented in very broad in 11 municipalities around the Copenhagen terms in order to avoid appreciable bias in County Hospital in Glostrup was drawn from relation to headache. It was stated that the aim the National Central Person Registry. The was to study the distribution of some important sample size was chosen to ensure adequate disorders in the general population, especially numbers of the most frequent forms of headache. The invitation was framed as an offer headache disorders. The total population of of a thorough health examination and the 1000 subjects was reduced to 975 because of importance of participation of all subjects decease (n = 3) and emigration (n = 17) in the invited was emphasized. A detailed, self-adminperiod between the date of sampling and the istered questionnaire was included to evaluate date of planned examination and an error in the demographic and psychosocial variables. If address lists (n = 5). Thus, only 975 persons necessary, examination outside working hours could be invited. A total of 740 persons (75.9%) or free transportation was offered. If the first were examined (Fig. 1). The total population of invitation evoked no response, a second was the sampling area was 325,621, which is 54% of issued. Those who still did not respond or who the Copenhagen County population and 6% of refused to participate were interviewed by the total Danish population. The representativetelephone. When they could not be reached by ness of the sampling area could be assessed by telephone, a third letter was mailed asking them comparing inhabitants in the sampling area, in to contact us by telephone or alternatively to return a postal questionnaire. Invited persons the Copenhagen County, and in all of Denmark using data from the National Statistics [17, 181. who did not turn up to the examination are The sampling area was representative of the labeled non-participants. The questions to the total Danish population as regards the age- and non-participants were selected from those of the and the information main investigation sex-distribution and marital status. Regarding obtained by telephone interviews or mailed industrial categories and employment status, questionnaires was recorded in order to evaluate farming and fishing were under-representated both in the sampling area and in the if non-participants differed from participants. Since late respondents might be more similar to Copenhagen County (0.5 vs 6.3% in Denmark), Sampling and representativeness
After second or third invitation
interviewed
Answered postal queationnalre
No infomatlon
Fig. 1. Diagram showing the distribution of the population according to type of participation.
Epidemiology of Headache
non-participants than early respondents, a comparison was made between the overall prevalences of the headache disorders in participants, who participated after the first, second or subsequent invitations. Subjects were informed that all data would be used for research purposes only and would be kept strictly confidential. The project was approved by the ethical committee for Copenhagen County. Examination
The study took place at the Glostrup Population Studies between January and July 1989. An average of 7 participants were examined per day each taking 3-4 hours. The participants were asked not to eat, drink or smoke during the last 8 hours before the examination which started in the morning. After approximately 1 hour they were offered a light meal. The questionnaire was checked for obvious errors and omissions, and the occupation and the education was recorded according to the International Standard Classification (ISIC) [ 191 with the modifications made by the Danish Statistical Department. A number of laboratory investigations were included in order to study e.g. the importance of myogenic mechanisms for the different forms of headache. The headache disorders were classified according to a structured diagnostic headache interview and a general physical as well as a neurological examination as required by the new diagnostic criteria. In the interview, questions about the occurrence of headache were phrased as follows: “Have you ever had a headache?” (lifetime prevalence) and “Do you have a headache today?” (point prevalence). The term headache included all forms of headache (migraine, tension-type headache, hangover, headache
1149
associated with fever etc.). In the following, subjects with at least one form of headache constitute the group labeled “headache”. The interview included questions about the different forms of headache and an extensive description of the headache frequency, duration, location, severity, character of pain, accompanying symptoms etc. The number of days per year with each kind of headache was recorded. The exact wording and details of the replies to individual questions about various features of migraine and tension-type headache are given in Table 1. Data about quality and quantity of each distinct headache form within the same subject were recorded. One physician took care of the clinical interview and examination for the first 2 months (R.J.), the other for the remaining 5 months (B.K.R.). Both were residents with more than 2 years of neurological training and a special interest in headache. Diagnostic criteria
All headache disorders were classified according to the operational diagnostic criteria of the International Headache Classification [ 161. The classification is hierarchically constructed. It contains 13 diagnostic groups which are subdivided to allow for coding up to a 4-digit level. It is thus possible to use the classification at different levels of sophistication. In the present paper, we present the prevalences and age- and sex-distribution of the primary headaches using the classification at the l-digit level in order to ensure adequate numbers in the subgroups. The main grouping of migraine and tension-type headache and the operational diagnostic criteria for migraine without aura (previously called common migraine) and episodic tension-type headache are given in Table 2 in order to
Table I. Interview questions used for classification of migraine and tension-tyne headache State number of migraine/tension-type headache attacks during your life until now: [l-4; 5-9; 10 or more] For how many days during the last 12 months have you suffered from migraine/tension-type headache? [O days; l-7 days; 8-14 days; 15-30 days; 31-180 days; more than 180 days] (4 Usual duration of the migraine/tension-type headache if you do not take any medicine, or if it does not work: [(l) less than 30 min; (2) between 30 min and 4 hr; (3) between 424 hr; (4) between 24-72 hr; (5) between 3-7 days; (6) more than 7 days; (7) varying from less than 30 min to more than 7 days] (4 Usual location of the migraine/tension-type headache? [(l) pain in the right part of the head only; (2) pain in the left part of the head only; (3) alternately in the right and left lateral part of the head; (4) alternately bilateral and unilateral; (5) always bilateral; (6) varies a lot] Subjects are classified with unilateral headache if they confirmed one of the answers (l), (2) or (3) (4 Which of the following types ofpain is the most characteristic description of your migraine/tension-type headache? [(l) pulsating pain; (2) pressing, tightening pain; (3) stabbing pain; (4) other] (f) How is the pain of the migraine/tension-type headache usually in case you do not take any medicine, or if it does not work? [(l) mild pain-daily activities (incl. housework) not inhibited; (2) moderate pain-inhibiting, but not preventing daily activities; (3) severe pain-daily activities suspended] (9) Does your migraine/tension-type headache get worse by climbing or walking down stairs? [yes; no] 09 Is your migraine/tension-type headache accompanied by: nausea? vomiting? loss of appetite? photophobia? phonophobia? [yes; no]
I%
1150
B~RTHEKRCIGHRASMUSSEN et al.
Table 2. The grouping of migraine and tension-type headache according to the International Heada& Class&a&m and the diagnostic criteria for migraine without aura and episodic tension-type headache (I) Migraine Migraine without aura (previously common migraine) Migraine with aura (previously classic migraine) (1.2) (1.3-1.7) Other migrainous disorders (1.1)
[16]
(2) Tension-type headache (2.1) Episodic tension-type headache (previously muscle contraction headache, stress headache etc.) (2.2) Chronic tension-type headache (2.3) Tension-type like headache
Diagnostic criteria for migraine without aura:
(A) At least 5 attacks fulfilling (BHD) (B) Headache attacks lasting 4-72 hr (C) Headache has at least two of the following characteristics: (1) Unilateral location (2) Pulsating quality (3) Moderate or severe intensity (inhibits or prohibits daily activities) (4) Aggravation by walking stairs or similar routine physical activity (D) During headache at least one of the following: (1) Nausea and/or vomiting (2) Photophobia and phonophobia Diagnostic criteria for episodic tension -type headache
:
(A) At least 10 previous headache episodes fulfilling criteria (B)-(D) listed below. Number of days with such headache < 180/year (B) Headache lasting from 30min to 7 days (C) At least two of the following pain characteristics: (1) Pressing/tightening (non-pulsating) quality (2) Mild or moderate intensity (3) Bilateral location (4) No aggravation by walking stairs or similar routine physical activity (D) Both of the following: (1) No nausea or vomiting (anorexia may occur) (2) Photophobia and phonophobia are absent, or one but not the other is present
exemplify the criteria used. The last criterion for all primary headaches (not listed under each form in the table) includes the ruling out of possible organic causative factors by history, physical and neurological examinations or other appropriate investigations. It is important to notice that in the classification system subjects receive a diagnosis for each distinct headache form experienced. Thus, subjects with e.g. both migraine and tension-type headache are included in the diagnostic group of migraine as well as that of tension-type headache. Data processing and statistical methods
All data were coded daily using the SPSS (Statistical Package of Social Sciences) dataentry programme. Data were transferred to the SCIBAS computer system at Copenhagen University, Herlev Hospital. In the present publication the chi-square test with a 5% level of significance has been used, and a Mantel-Haenszel summary x2 test (M-H test) [20] was used to control for age confounding. Ninety-five per cent exact confidence intervals using the binomial distribution were used.
RESULTS
Response pattern
The participation rate was 76%. The response pattern in the study population according to age and sex appears in Table 3. Concerning the response rates, none of the age- or sex-differences were statistically significant. Participants who responded to the first, second or third invitation differed as regards age and sex, since there were more young men who required a second or third invitation before participating (sex difference: M-H test = 10.67, df = 1, p = 0.001; age difference (d): x2 = 6.76, df= 3, p = 0.079). Of the study population 12.3% (n = 120) were interviewed by telephone, 6.7% (n = 65) answered a postal questionnaire and 5.1% (n = 50) were unreachable or unwilling to give any information. Among non-participants (n = 235) information was thus obtained from 78.7% (n = 185). Prevalence of headache, migraine and tensiontype headache
The lifetime prevalence of headache (including both migraine, tension-type headache
1151
Epidemiology of Headache Table 3. Response pattern among 975 subjects in Copenhagen County Non-participants
Participants* (%)
Sex and age (yr)
Telephone interview (%)
Postal questionnaire (%)
Others (%)
Total No. in sample N = 975
Males
25-34 35-44 45-54 5564 Mean
74.6 83.0 78.5 73.8 77.9
7.6 10.6 13.3 13.6 11.3
6.8 3.6 6.7 5.8 5.6
11.0 2.8 1.5 6.8 5.2
118 141 135 103
Females 25-34 3544 45-54 554 Mean
76.5 74.3 72.4 71.9 73.9
10.1 16.2 15.8 10.4 13.4
8.4 5.9 5.5 12.5 7.7
5.0 3.7 6.3 5.2 5.0
119 136 127 96
Mean (both sexes)
75.9
12.3
6.7
5.1
*Sex difference: M-H test = 2.17, df = 1, p = 0.14. Age difference: 8, x2 = 3.91, df = 3, p = 0.27. ?, X2 = 0.76 p = 0.86.
and all other types) was 96% (709/740) and was significantly higher among females (99%) than among males (93%) (Fig. 2). Men aged 55-64 had lower prevalence of headache than men in the other age groups. Point prevalence rates were significantly higher among %A
lco-
9 All headaches combined~=.
90-
d
so-
9
7060d 50-
11
40.
30-
t;
_---
___--
zo-
9
Migraine ---
lo-
~:::yY:
’
wk‘l
--_
35:44
+
----
45kl
1
df= 3,
women compared to men with a male/female ratio of 1:2 (Table 4). The overall lifetime prevalence of migraine was 16% (119/740), among men 8% (30/387) and among women 25% (89/353). Lifetime and last year prevalence of migraine were significantly higher among women than among men (Fig. 2, Table 5). Male/female ratio was about 1: 3. There were no significant differences in migraine prevalence rates according to age. The overall lifetime prevalence of tensiontype headache was 78% (578/740), among men 69% (266/387) and among women 88% (312/353). Lifetime and last year prevalences of tension-type headache were significantly higher among women than among men with a male/female ratio about 4: 5 (Fig. 2, Table 5). Men aged 55-64 had lower lifetime- and last year prevalence of tension-type headache than men in the other age groups. Among women there were significantly decreasing lifetime- and last year prevalences of tension-type headache with increasing age.
0
55:&a
A&
Fig. 2. Lifetime prevalences of headache, tension-type headache, and migraine and 95% confidence intervals. Relation to age and sex. Headache: Sex difference: M-H test = 12.91, df = 1, p < 0.001. Age difference: 3, x2 = 5.08, df = 3, p > 0.1. 0, x2 = 2.23, df = 3, p z 0.5. Migraine: Sex difference: M-H test = 41.81, df = 1, p < 0.09001. Age difference 8, x2 = 0.31, df = 3, p > 0.9. 0, x2 = 1.99, df = 3, p > 0.5. Tension-type headache: Sex difference: M-H test = 42.09, df = 1, p < 0.00001. Age difference g, x2 = 9.99, df = 3, p < 0.02. 9, x2 = 13.53, df = 3, p < 0.004.
Table 4. Point prevalence of headache and 95% CL Point prevalence Men Age group 25-34 35-44 45-54 55-64 All ages
(;=$
$;$
Women
All (N = 740) % (CL)
13 (6-21) 10 (5-17) 10 (5-18) 9 (4-18)
31(2141) 22 (14-31) 13 (7-22) 22 (13-33)
22 (1629) 16 (11-21) 12 (8-17) 15 (l&22)
11 (8-14)
22 (18-26)
16(13-19)
Sex difference: M-H test = 16.69, df = 1, p < 0.0001. Age difference: 6, x2 = 0.51, df = 3, p > 0.9. 2, x2 = 8.43, df = 3, p < 0.05.
BIRTHEK~OGH RASMUSSEN ef al.
1152
Table 5. Prevalences in the previous year of migraine and tension-type headache and 95% CL Migraine
Tension-type headache
Age group
Men (n = 387) % (CL)
Women (n = 353) % (CL)
All (N = 740) % (CL)
(;O;$
Women (n = 353) % (CL)
(N :;40) % (CL)
25-34 354l 45-54 554
5 (l-l 1) 7 (3-13) 6 (2-12) 7 (2-15)
18 (10-27) 14 (8-22) 12 (6-20) 19 (10-30)
11 (7-17) 10 (615) 9 (5-13) 12 (8-19)
68 (57-78) 63 (5472) 70 (60-78) 49 (3760)
93 (86-98) 92 (8S97) 82 (72-89) 74 (62-84)
81 (75-86) 77 (7&82) 75 (69-81) 61 (52-69)
All ages
6 (4-9)
15 (12-19)
10 (8-13)
63 (58-68)
86 (82-90)
74 (71-77)
Men
Migraine: Sex difference: M-H test = 17.10, df = 1, p < 0.0001. Age difference: 6, x2 = 0.55, df = 3, p > 0.9. 0, x2 = 1.99, df = 3, p > 0.5. Tension-type headache: Sex difference: M-H test = 50.53, df = 1,p < 0.00001. Age difference: 6. x2 = 9.83, df = 3,p < 0.03. 9, x2 = ii.27, df = 3, p < 0.001.
None of the migraineurs had migraine on the day of examination; the prevalence of migraine in the last week before the examination was 2% (13/740) and in the last month 4% (29/740). The point prevalence of tension-type headache was 12% (91/740), the last week prevalence was 29% (212/740) and the prevalence in the last month was 48% (358/740). Among the 119 subjects who ever have had migraine, 103 (87%) had in addition at some time also had tension-type headache. Among subjects with migraine in the previous year 83% had coexisting tension-type headache. Considering subjects with migraine in the last month 62% had coexisting tension-type headache. 46% of subjects with migraine in the last week had in addition had tension-type headache in this period. Of the migraine group 15% had it 8-14 days a year and 9% had it more than 14 days a year. In the total population 2% had migraine for more than 14 days a year (Fig. 3). As regards the severity of the migraine, 85% of the migraineurs had a severe pain intensity, 14% a moderate pain and only 1% a mild pain. Of the tensiontype headache group 23% had it 8-14 days a year and 36% several times a month. In the total population 44% had tension-type headache for 1-14 days a year and 3% had chronic tensiontype headache (i.e. headache 2 180 days a year) (Fig. 3). Among subjects with tension-type headache, the pain was severe in l%, moderate in 58% and mild in 41%. We found only one case of cluster headache. Non -participants The 185 subjects interviewed by telephone or answering the postal questionnaire stated their reason for not attending the investigation. The main reasons were reluctance to lose a working day (37.5%); attention to family (7.8%); illness
(15.6%); being too healthy or having had a recent health examination (4.7%); fear of hospitals (19.5%); fear of diagnosis (10.2%) and fear of registries (4.7%). No significant differences between participants and non-participants were found regarding age, sex, and marital status. Comparisons of the answers to identical questions by the participants and by subjects interviewed by telephone or answering the postal questionnaire showed no significant differences as regards hospital admissions or self-evaluated health within the last year. Likewise present status of employment, education, and nationality were not significantly different. The lifetime prevalences of headache, migraine and tension-type headache among the non-participants who answered the telephone interview or a postal questionnaire were 92, 13 and 70%, respectively, compared to 96, 16 and 78% in participants. These differences, although slight, were statistically significant as regards headache (OR = 1.99 (1.05-3.78)) and tensiontype headache (OR = 1.55 (1.08-2.23)). No difference in the overall prevalence of headache, migraine, and tension-type headache was found between subjects who participated after the first, second or subsequent invitation.
DISCUSSION
Methodological considerations
All members of the sample were invited to a general health examination, but with the main emphasis on headache disorders. Therefore, an over-representation of persons with headache could be expected among participants. The of headache sufferers over-representation among the participants compared to non-participants could indicate that the observed
1153
Epidemiology of Headache
al %
b) %
40
8
-
7
6
30
20
1c
TENSION-TYPE
C) %
di %
40
40.
30
30.
20
20.
10
10.
HEADACHE
q
Fig. 3. Number of days with migraine/tension-type headache in the previous year. (a) percentage of those with migraine; (b) percentage of population; (c) percentage of those with tension-type headache; (d) percentage of population. 0 indicates no attacks in the previous year but earlier presence of the disorder. (Note y-axis not the same in (b))
prevalences are a little too high for the whole sample. On the other hand, the difference may be due to the different methods of data collection in participants and non-participants. Even assuming that the sampling method did not influence the prevalence results, the reported discrepancy in headache prevalence is, however, minimal because of the high participation rate. If we presume that the lifetime prevalences of headache, migraine, and tension-type headache
in non-participants of 92, 13 and 70% were valid for all non-participants, the estimated prevalences in the total study population would be 95, 15 and 76%, respectively, which is not appreciably different from the values in participants of 96, 16 and 78%. The overall morbidity and various sociodemographic parameters were uniform for participants and non-participants. Thus, the examined population can be accepted as representative of the total study population.
1154
BIRTHEKROGHRASMUSSEN et al.
The sampling area showed only minor differences compared with Copenhagen County and can probably be considered representative of this area. The most marked difference as compared to all of Denmark was an underrepresentation of industries of farming and fishing. This difference was expected and is probably without importance since a previous study found very little difference between urban and rural areas as regards headache prevalence
WI. Prevalence results
Since headache is a subjective complaint without any laboratory correlate, a study of its prevalence must rely exclusively on information given by the subjects, The threshold of recall varies probably with the severity and recency of the condition. We found very high prevalence rates of headache and of tension-type headache. Only 4% of the total population had never had a headache. Using the usual 95% confidence limits of normality, it is strictly speaking abnormal never ever to have had a headache. More important is the fact that it is meaningless to talk about tension-type headache as a disease without specifying how often it occurs. Further analysis is necessary to decide a frequency of tension-type headache beyond which people can be regarded as sufferers. Even if many had rare tension-type headache, 14% of the population had more than 30 headache days a year, and 3% had it more than half of the time. Frequent migraine was relatively rare according to our results (l-2%), but 7% of the migraineurs were in prophylactic treatment, which probably has reduced the frequency. Most migraine sufferers had l-7 migraine days a year and the number of migraineurs decreased with increased number of migraine days a year. Women had higher prevalences than men migraine, and tension-type of headache, headache, which stresses the importance of considering the different forms of headache as sex dependent. Tension-type headache showed significantly declining lifetime- and 1-year prevalences with increasing age, whereas the small numbers in the migraine group prompts caution in evaluating subgroups. One would expect lifetime prevalences to show a cumulative trend, increasing with increasing age. In contrast, we found lower lifetime prevalence in the older age groups. This could be explained by a cohort effect with the older age groups being more reluctant to
confirm that they have had headache. Secondly, it could be due to recall bias. Greater exposition to headache provoking job situations in the younger or increasing headache incidence in general are other theoretical possibilities. A final conclusion demands further elucidation including a follow up of the cohort in 5 or 10 years. The point prevalence for migraine of zero was probably artificially low, because persons who happened to have migraine on the day of appointment stayed at home and got a later appointment. The high point prevalence of headache may have been increased by the required 8 hours of fasting, abstinence from smoking, and anxiousness because of the examination. Our results of different time-period prevalences showed the expected increase with increasing observation period. Present results in relation to previous studies
The present study is the first investigation performed in a representative random population, using operational diagnostic criteria, and including a clinical interview as well as a general and a neurological examination of all participants. Only a few earlier studies have examined a random sample of the general population [8-141, whereas selected materials have often been used. Some studies used questionnaires [ 1,4,5, 12, 14,21,22] others clinical interviews [lo, 231. A physician did the interview in some previous studies [9, 10,231, lay interviewers in others [24,25]. Age and sex varied in previous studies. Another essential problem is to recognize the influence of how the questions about headache have been phrased. Some asked: “Do you suffer from headache?” [1,4, 111, others asked: “Do you (ever) have headache?’ [IO, 12, 14,22,23]. Presumably, the inclusion of the words “suffer from” gives a lower prevalence than “have”. Some reports concern lifetime prevalences [1, 10,22,23] others different time-period prevalences [5,7, 10, 13, 14,251, and some do not specify the period [l 1,261. The varying methods in earlier population studies have resulted in reports of a wide range of prevalences of the different forms of headache and have impeded systematical comparison. Thus, comparing our results to other studies presents great difficulties and should be done with caution. Nevertheless, the most comparable studies show agreement with the present results concerning prevalence rates of headache [22,23] (Table 6).
Doctors
Dalsgaard-Nielsen,
General population
Holhragel/Norrelund,
General population
General practice
General population
Shopping center
Nikiforow, 1981, Finland [23]]]
Philips, 1976, U.K. [5j$
Rasmussen et al., 1991, Denmark, (present study) 11
1980, U.S.A. [21]$
1971, U.K. [27lt
Schnark/Hunter,
Waters/O’Connor,
General population
Church congregations
Waters, 1974/75, U.K. [14]$
Ziegler, 1977, U.S.A. [22]#
*Lay-administered interview and clinical interview of small subgroup. tQuestionnaire and clinical interview of small subgroup. $Questionnaire. §Telephone interview. l/Lay-administered interview. ]]Clinical interview and examination.
General practitioners
Waters, 1972, U.K. [q$
General population (women)
Women serving as control group for cerebrovascular disease
Markush et al., 1975, U.S.A. [24]7
1980, Denmark [I l]$
Non-clinic, telephone sampled
Duckro et al., 1989, U.S.A. [25j5
1973, Denmark [I]$
General population
General population
Crisp et al., 1977, U.K. [lO]t
Sample source
D’Alessandro er al., 1988, Italy [8]*
Study
1809
1718
882
2933
1293
740
597
200
451
1052
500
461
727
1154
Respondents (No.)
>15
>21
35-54
2&64
Adults
2564
16-60
>15
1544
40
>21
2585
Adults
>7
Age
Lifetime
Previous 1 year
Previous 1 year
Previous 1 year
Previous 1 year
Lifetime Previous 1 year Point
Previous 6 months
Lifetime Previous 1 year
Previous 1 year
?
Lifetime Past 3 months
Lifetime
Lifetime Previous 2 year
Previous 1 year
Time-period prevalence
82.6
63.5
84.2
78.4
25.7 19.5
18.0
%F 16.1
14
8 6 0
11
8
18
26
19
25 15 0
10
35
23.5
16 10 0
21
24.5
13
61
88 86 16
65
42
34.5
(mild headache) 41.7 34.0
42
69 63 9
37
28.8
%F
78 74 12
68
40
% Both
Tension-type headache % Both %M
14 22 16 (very severe headache) 11.2 20.4 15.8 8.8 15.2 12.0
9.8 8.9
9.3
%M
Migraine
14.9 23.2 (disabling headache) 40.9 50.2
7.3
16
22 11
78.7
96
89
91 77
77
%Both
99
74
83
76.5
85
93.7
46.2
%F
93
69
69
68.9
35.3
%M
Headache
Table 6. Some prevalence studies of different forms of headache in America and Western Europe
!? e Q
s
3 _. g
am $I
BIRTHE KRUGH RASMUSSENet al.
1156
Our male/female ratio among migraine sufferers is in good agreement with the most comparable earlier studies [9, 10,231. Concerning prevalence of migraine in the previous year several studies [5,23,24,27] have found higher prevalences than the present study. However, Waters [27] used a questionnaire technique and only estimated prevalences without giving confidence limits. The survey of Philips [5] was in a selected population and had a very low response rate; Markush [24] studied a non-representative sample of young women; and Nikiforow [23], although using a clinical interview and examination, studied only 200 persons. Furthermore, the 200 persons were sampled from the 77% who primarily had answered a postal headache questionnaire, and a moderate selection bias can be suspected. Our report of no correlation between migraine prevalence and age within the studied age interval (25-64) is in agreement with previous studies. In addition, these studies have shown the prevalence of migraine to decrease in subjects above the age of 64 [6,22,23,28]. Reported prevalences of tension-type headache vary. Previous studies [6,23] have reported lower prevalence rates than those found in the present study. Regarding the influence of the diagnostic criteria, it is tempting to suggest that some persons with tension-type headache may have been diagnosed as migraine in the past. The reported female preponderance and the decreasing prevalences with increasing age in tension-type headache are confirmed in some other studies [2,3,6, 14,281, even if they used other diagnostic criteria. Philips’ finding of almost equal occurrence of tension-type headache in men and women [5] seems not to hold up. CONCLUSION
The present study is the first prevalence study of specific headache entities based on a structured interview and examination by a neurologist and using internationally accepted operational diagnostic criteria in a large random sample. The representativeness of the sample improves the possibility of genemlizing prevalence rates to other populations. Our study emphasizes the important health problem that the headache disorders constitute in the general population, and it provides a basis for future studies of headache in selected groups with the purpose of revealing possible risk factors for headache.
Acknowledgements-This study received statistical support from The Danish Medical Research Council (i. No. 5.29.00.60) and was supported by grants from the Danish Health Insurance Foundation (H 1l/238-88, H 1l/262-89), The Lundbeck Foundation (86/88), GLAXO (1989), DAK medica (379), The Foundation for Experimental Research in Neurology (1988), and the Danish Migraine Society (1988).
REFERENCES 1.
2.
3. 4.
5. 6.
7. 8.
9. 10.
11. 12. 13. 14. 15. 16.
17. 18. 19. 20. 21. 22.
Dalsgaard-Nielsen T, Ulrich J. Prevalence and heredity of migraine and migrainoid headaches among 461 Danish doctors. Headache 1973; 12: 168-172. Dhopesh V, Anwar R, Herring C. A retrospective assessment of emergency depaitment patients with comolaint of headache. Headache 1979: 19: 3742. Friedman AP, von Starch TJC, Merritt HH. Migraine and tension headaches: a clinical study of two thousand cases. Neuroloev 1954: 4: 773-778. Ogunyemi AO. Preva&ce df headache among Nigerian University students. Headache 1984; 24: 127-130. Philips C. Headache in general practice. Headache 1977; 16: 322-329. Waters WE. Headache and migraine in general practitioners. The Migraine Headache and Dixarlt. Bracknell, England: Boehringer Ingelheim; 1972: 3144. Waters WE. The Epidemiology of Migraine. Bracknell, England: Boehringer IngelhGm; 1974. D’Alessandro R. Benassi G. Lenzi PL ef al. Eoiderniology of headache in the Republic of San Marino. J Neurol Neurosurg Psychiat 1988; 51: 21-27. Brewis M, Poskanzer DC, Rolland C et al. Neurological disease in an English city. Acta Neural Seand 1966; 42 (Suppl. 24): l-89. Crisp AH, Kalucy RS, McGuinness B, Ralph PC, Harris G. Some clinical, social and psychological characteristics of migraine subjects in the general population. Postgrad Med J 1977; 53: 691-697. Hollnagel H, Nsrrelund N. Headache among 40 yearolds in Glostrup. Ugeskr Laeger 1980; 142: 3071-3077. Nikiforow R, Hokkanen E. An epidemiological study of headache in an urban and a rural population in northern Finland. Headache 1978; 18: 137-145. Waters WE. Community studies of the prevalence of headache. Headache 1970; 9: 178-186. Waters WE. The Pontypridd headache survey. Headache 1974; 14: 81-90. Ad Hoc Committee on Classification of Headache. Classification of headache. JAMA 1962; 179: 717-718. Headache Classification Committee of the International Headache Society. Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalnlgla 1988; 8 (Suppl. 7). Danmarks statistik. Befolkningen i kommunerne 1. januar 1989. Statistisk Tabelvark 1989. Copenhagen; 1989. Danmarks statistik. Arbejdsmarked. Statistiske Efterretninger. 1989; 1. United Nations. International Standard Industrial Classification, Statistical Papers 1968; Series M, No. 4, Rev. 2, Add. 1. Mantel N, Haenszel W. Statistical aspects of the analysis of data from retrospective studies of disease. J Nat1 Cancer Inst 1959; 22: 719-748. Schnarch DM, Hunter JE. Migraine incidence in clinical vs nonclinical populations. Psychosomaties 1980; 21: 314-325. Ziegler DK, Hassanein RS, Couch JR. Characteristics of life headache histories in a nonclinic population. Neurology 1977; 27: 265-269. 1
Epidemiology of Headache 23.
Nikiforow R. Headache in a random sample of 200 persons: A clinical study of a population in northern Finland. Cephalalgia 1981; 1: 99-107. 24. Markush RE, Herbert RK, Heyman A, O’Fallon WM. Epidemiologic study of migraine symptoms in young women. Neurology 1975; 25: 430-435. 25. Duckro PN, Tait RC, Margolis RB. Prevalence of very severe headache in a large US metropolitan area. Cephalalgia 1989; 9: 199-205.
26.
1157
Levy LM. An epidemiological study of headache in an urban population in Zimbabwe. Headache 1983; 23: 2-9. 27. Waters WE, O’Connor PJ. Epidemiology of headache and migraine in women. J Neural Neurosurg Psychiat 1971; 34: 148-153. 28. Waters WE. Epidemiological data relevant to prognosis in migraine in adults. Dan Med BuB 1975; 22: 89-91.
