Abstract
There are scarce data about headache prevalence and its characteristics among elderly people. The aim was to carry out a cross-sectional study to determine the 1-year prevalence of tension-type and migraine headaches in people >65 years old in the city of São Paulo, Brazil. All 1615 people living in the study catchment area who agreed to participate in the study answered a questionnaire based in the International Headache Society criteria. Prevalence (mean and 95% confidence interval) of any type of headache in the last year was 45.6% (43.2, 48.0). Prevalence of tension-type headache in the last year was 33.1% (30.8, 35.4): 28.1% (24.6, 31.6) for men and 36.4% (33.4, 39.4) for women; for migraine headaches, prevalence in the last year was 10.6% (9.1, 12.1): 5.1% (3.4, 6.8) for men and 14.1% (11.9, 16.3) for women. One-year prevalence rates of headaches, and especially of migraine headaches, are very high among the elderly in Brazil.
Introduction
Headache is less frequent in older than in younger age groups. However, the increasing number of the aged makes headache a more common complaint among the elderly (1, 2). Solomon et al. (3), in 1990 estimated that 66% of all headaches in elderly people would be classified as primary headaches: tension-type and migraine headaches. Migraine headaches are a common medical problem, especially among women, with as many as one in six women having at least one migraine headache a year (4, 5). Despite many data on migraine in women, information about migraine prevalence and characteristics in an elderly sample of men and women is sparse, as is information about tension-type headache.
Only few studies have evaluated the prevalence of primary headaches in the elderly worldwide, most of them in Europe (6, 7). Morillo et al. have evaluated the prevalence of migraine in Latin America in a randomly selected population sample of urban communities. However, most of the sample consisted of individuals 20–49 years of age, with only a very small number of individuals >60 years old (8). A survey performed in Brazil by Queiroz et al. (9) has estimated the 1-year prevalence of headache in a sample of 625 subjects, aged 15–64 years, with a preponderance of migraine in females, divorced or widowed, at a low socioeconomic level.
To address the epidemiology of migraine and tension-type headache among the elderly, we studied the 1-year prevalence of migraine and tension-type headache in a population sample of 65-year-old men and women living in a low-income area in the city of São Paulo, Brazil.
Methods
A cross-sectional population-based study was carried out with residents of an economically deprived area of São Paulo, Brazil as part of a collaborative investigation in Brazil, Cuba and the UK., whose main objectives were the study of environmental and genetic risk factors for dementia in admixed populations in Brazil and Cuba. The study was approved by the Institutional Review Board.
Sample
The study was carried out in 66 predetermined census sectors (catchment areas) in the area of Butantã, located on the Western side of the city of São Paulo. In the catchment areas in which there were shanty towns (‘favelas’), or by the census sectors covered by the Family Health Program—with the lowest income in the area. The sample size calculation was based on the expected association between general dementia and individual racial admixture, the main objective of the study, including nearly 2000 people.
All residents ≥65 years old were included. They were identified through systematic door knocking on 21 727 households within the catchment-area boundaries, totalling 2266 people. Of these, 2072 agreed to participate. However, the questionnaire concerning headache had just begun to be applied when 457 people had already been interviewed, so 1615 people answered this questionnaire.
Procedures
The study was conducted over a 3-year period, from May 2003 to April 2005. Recruitment and interviewing of the participants were carried out by a group of interviewers trained to work in the present study. All those aged 65 and in the study catchment area who accepted to participate were interviewed. This meant that in households with two or more elderly people, all were invited to participate. All participants were assessed for dementia, socioeconomic characteristics, lifestyle and cardiovascular risk factors. A headache questionnaire using the International Headache Society (IHS-2004) criteria (10), previously validated in Brazil, was applied as part of the main questionnaire (11). Age was classified in four age strata: 65–69, 70–74, 75–79 and >80 years. Race was classified by the interviewer into four categories: White, Black, Mixed and Asian. Education was classified according to the number of school years into three categories: none, 1–3 and >4 years. Income per month (transformed from Brazilian currency to US Dollars) was classified in four categories: <110; 111–163, 164–318 and >319.
Data analysis
Data entry was carried out twice using the program EPIDATA 3.0, and the validity check was carried out to identify and correct data entry errors. Data were analysed using SPSS 14.0 (SPSS Inc., Chicago, IL, USA). One-year prevalence of headaches or headache subtypes is presented as the number of cases per 100 persons. Ninety-five percent confidence intervals were also presented for prevalence. χ2 tests were used for comparison when appropriate. χ2 for linear trend analysis was used to measure the age effects of headache prevalence. A P-value <0.05 was considered to be statistically significant.
Results
Of the 1615 people in the study, 45.6% referred headache in the last year. Of these, 71.2% were women. Prevalence of any type of headache in the last year was 33.7% for men and 53.1% for women (Table 1). Prevalence of headache was higher in women than in men (P < 0. 001).
Prevalence of primary headaches in a population sample of ageing men and women according to the 2004 criteria of International Headache Society
Tension-type headache (including all subtypes) was the most common type of headache in this population, with a female:male ratio of 1.3 : 1. Migraine headaches were also more common in women, with a female:male ratio of 2.8 : 1, although menopause had occurred in these women >10 years previously. Prevalence of tension-type headaches in the last year was 33.1% for both genders: 28.1% for men and 36.4% for women. Prevalence of migraine headaches in the last year was 10.6% for both sexes: 5.1% for men and 14.1% for women.
For tension-type headaches, 87.1% (n = 466) were confirmed cases (IHS2.1, 2.2 and 2.3) and 12.9% (n = 69) probable cases (IHS2.4). For migraine headaches, 69 (40.4%) were confirmed cases (IHS1.1, 1.2, 1.5) and 102 (59.6%) were probable cases (IHS1.6). Of these, 90 (88.2%) referred <4 h of pain. Using a modified criterion excluding duration of headache, as in previous papers (4, 5, 12–14), increased the number of confirmed cases to 159 (159/171 = 93.0%). Of the 736 people reporting any kind of headache, 30 did not fulfil criteria for tension-type or migraine headaches and they were classified as having other headaches.
Table 2 shows the prevalence of headaches according to age strata by gender. There was no difference in the prevalence of headaches according to race, number of school years or mean income. Frequency of tension-type and migraine headaches decreased with age, but not statistically significantly so.
Prevalence of tension-type and migraine headaches according to age strata by gender
Discussion
There are sparse data about prevalence of headache in ageing people. Franceschi et al. (15) in 1997 evaluated the prevalence of headache in 312 subjects >65 years old (47.4% women). Prevalence of current migraine headache was 2.0% in women with no case in men; and of tension-type headaches, 4.0% in women and 1.2% in men. In the same year, Wang et al. (16), in a population sample of elderly Chinese (≥65 years; 56.4% women), demonstrated a 1-year prevalence of migraine in men and women, respectively, of 0.7% and 4.7%; and of tension-type headache, respectively, of 20% and 46%. In 1999, the GEM Study (17) demonstrated a 1-year prevalence of migraine headaches in men and women ≥55 years old, respectively, of 5.6% and 12.4%. Henry et al. (18) in 2002 showed a prevalence of migraine headaches and migrainous disorder in France, for men and women aged >64 years, of 1.5%, 2.0%, 2.5% and 7.0%. In 2003, Camarda and Monastero (6) studied prevalence of primary headaches in 1031 participants (54.3% women) in a door-to-door survey of people ≥65 years old. One-year prevalence rates for headache were 4.6% for migraine, 16% for tension-type headaches and 1.3% for other headaches. In 2005, Prencipe et al. (7) assessed in a population sample of elderly the 1-year prevalence of migraine headaches, 13.8% for women and 7.4% for men. For tension-type headache, the 1-year prevalence was 55.1% for women and 30.9% for men.
Our results have shown a higher 1-year prevalence of migraine and tension-type headaches in our sample of elderly people. Several factors could explain our results. The population sample of ageing people who participated in the study was selected from the area in the District of Butantã that has a lower income in the area. Almost 13% of the participants live in shanty towns, and 87.8% of the heads of households had <1 year of schooling. Some studies have shown an association between low socioeconomic status and prevalence of migraine, including a previous study in Brazil (9, 19–22), not confirmed in other studies (17, 23, 24). However, it is plausible that the higher 1-year prevalence of migraine headaches could be explained by the low-income population in this sample compared with previous studies. Furthermore, Brazil is part of Latin America, and other studies on headache in elderly people have been done in Europe and China. Cultural factors could explain the higher 1-year prevalence of headaches in Brazil. Craig et al. (25) have studied the current prevalence of migraine headaches in Mexican Americans born in Mexico and the USA. Mexican Americans born in Mexico have twice the current prevalence of migraine headaches than Mexican Americans born in the USA. In the same study, the current prevalence of migraine in men and women ≥65 years old was 9.1%, more similar to the 1-year prevalence of migraine headaches in our sample of 10.6%.
In our sample, the 1-year prevalence of tension-type or migraine headaches did not change according to age strata. Frequency decreased with age, but statistically significantly so. Camarda and Monastero (6) have described a significant decrease in migraine headache prevalence with increasing age. Wang et al. (16) have also described a decrease in migraine headaches prevalence in men, but not in women. One possible explanation for the difference between our study and others could be the small number of people aged ≥75 years in our sample compared with Italy and even China. Of the 1615 participants in the present study, only 28.5% of the sample were >75 years old.
In our study, some kind of misclassification is possible, because headache diagnosis was done using a questionnaire and not personal interviews with a doctor. However, the questionnaire is based on the IHS criteria and has been validated in a previous study in a Brazilian low-income population (11). Several epidemiological studies have used lay interviewers in recent years with good results. In the present study, we have not been able to exclude symptomatic headache with certainty. However, 90.3% of people complaining of headache remembered the age in which headache began. Of these, 80.7% referred the beginning of headache to a mean age of 43.1 years (7, 24) (median 45 years; mode 50 years), suggesting a history of primary headaches of many years' duration.
Brazilian studies about prevalence of headaches have shown the high frequency of migraine headaches in our population. The recent study of Morillo et al. (8) has shown a higher frequency of migraine headaches in Brazil compared with other countries in Latin America. Other studies in different age strata and samples in Brazil have confirmed the high prevalence of migraine headaches in this country (9, 26–28).
Our findings support the conclusion of an extensive and recent review of migraine in the elderly, that stated that ‘migraine in older age groups has been neglected’ and ‘even in these age groups, effective treatment is available and careful management can improve the quality of life’ (1). Our results have confirmed in an ageing sample of men and women that the prevalence of headaches, and especially of migraine headaches, is very high in Brazil. Low income and cultural factors are a likely explanation for these findings.
Footnotes
Acknowledgements
All authors are recipients of a scholarship from Conselho Nacional de Pesquisa, CNPq, Brasília, Brazil. Grant research: Wellcome Trust, London, UK.