Prevalence and Burden of Headache and Migraine in Germany


  • Andrea Radtke MD,

    1. From the Department of Neurology, Charité, Berlin (A. Radtke); Robert Koch Institute, Department of Epidemiology and Health Reporting, Berlin (H. Neuhauser).
    Search for more papers by this author
  • Hannelore Neuhauser MD, MPH

    1. From the Department of Neurology, Charité, Berlin (A. Radtke); Robert Koch Institute, Department of Epidemiology and Health Reporting, Berlin (H. Neuhauser).
    Search for more papers by this author

  • Conflict of Interest: None

Dr. Andrea Radtke, Department of Neurology, Virchow Klinikum, Charité, Augustenburger Platz 1, 13353 Berlin.


Objective.— To determine prevalence and burden of headache and migraine in the general population in Germany including patterns of healthcare and medication use.

Methods.— Telephone interviews were conducted on a representative sample of the general population in Germany aged ≥18 years (n = 7341). Migraine was diagnosed according to the International Headache Society criteria.

Results.— One-year prevalence was: headache 60.2%, migraine 10.6%, nonmigrainous severe headache 24.7% (women 66.6%, 15.6%, and 27.1%; men 53.0%, 5.3%, and 22.2%). Approximately 60% of headache sufferers reported severe headaches, 30% of which were migrainous. Migraineurs reported more often frequent headaches, disability, use of analgesics, and medical consultation than individuals with nonmigrainous severe headaches. Only 42% of migraineurs had consulted a physician and the majority relied exclusively on over-the-counter medication.

Conclusion.— Migraine accounts for a great part of the healthcare impact of headaches in Germany. However, the majority of migraineurs do not seek medical care and may not be optimally treated.


German National Telephone Health Interview Survey 2004


International Headache Society

Headaches and migraine as one of its most severe subtypes are common disorders everywhere in the world. Globally, almost half of the population suffer from unspecified headaches1 and prevalence of active migraine have been estimated at 14% in Europe2 and 11% worldwide.1 Headaches, and especially migraine, impose a significant burden on individual sufferers and on society through reduced quality of life, loss of productivity, and use of healthcare resources.3-12 However, despite considerable disability, many migraine sufferers do not seek medical care and even in those who do, migraine remains largely underreported13-16 and undertreated.12,14,16-18

Epidemiological studies help identify barriers for adequate care by providing estimates of the prevalence of different headache disorders, by assessing the magnitude and the distribution of the burden they inflict and by analyzing the patterns of healthcare use and treatment in affected individuals. However, while detailed epidemiological data are available for many Western European countries2,8,10,19-22 and North America,1,12,23,24 representative data on the prevalence of migraine in Germany are lacking. Two older studies from West Germany have methodological limitations and have yielded incongruent results.25,26 Representative data on the social impact of headaches and migraine including patterns of healthcare and medication use do not exist for Germany.

The aim of this study was to determine the prevalence of headache and migraine in the general adult population in the whole of Germany according to the revised International Headache Society (IHS) criteria27 and to assess the personal and social impact including healthcare utilization and treatment in migrainous and nonmigrainous headaches.


The data of this study were collected within the German National Telephone Health Interview Survey 2004 (GNT-HIS) which assessed a large representative sample of the general adult population in Germany aged 18 years or older (n = 7341; 3545 men and 3796 women) by means of standardized structured computer-assisted telephone interviews that were conducted by trained lay interviewers who completed an extensive training program, were intermittently supervised and received refresher training. Recruiting of a representative sample of the general population was achieved applying a modified random digit dialing sampling design28 combined with a next-birthday method. For this, the interview in each household reached by fixed-line telephone was carried out with the person who had next birthday. The response rate was 52.3%. The GNT-HIS covered various aspects of physical health, demographic and socioeconomic characteristics, lifestyle behavior as well as healthcare utilization. The study was approved by the Federal Data Safety Commissioner.

As a screening question, participants were asked if they had suffered from headaches during the past 12 months and to rate the intensity of these headaches as (1) predominantly severe; (2) both severe and mild; or (3) predominantly mild. Only participants with severe headaches were then further explored for migrainous headaches and for healthcare utilization. A diagnosis of migraine based on the novel 2004 IHS criteria27 was assigned if all of the following criteria were fulfilled: (1) severe headaches in the past 12 months; (2) unilateral or pulsatile pain or pain aggravated by routine physical activity; (3) nausea/vomiting or photo- and phonophobia with headaches; and (4) duration of attacks of 4-72 hours. A simplified question was used to detect visual auras asking participants if headaches had at least on some occasion been accompanied by flickering lights, or lines, lasting for at least 5 minutes. Migraines with, and without, aura were considered mutually exclusive. Severe headaches not fulfilling the above criteria were classified as nonmigrainous severe headaches. Complete data for classification of migraine status were available for 7236 out of 7341 participants of the GNT-HIS (98.6%).

All headache sufferers were inquired about the number of headache days during the past 12 months and 4 weeks, about days during which they were unable to perform usual activities at work and in everyday life during the past 12 months and about pharmacological treatment (both prescription and over-the-counter medication) during the past 12 months and 4 weeks. Participants with severe headache were additionally questioned about medical consultation, subspecialties of medicine they had consulted, and nonpharmacological treatment of headaches in the past 12 months.

Three educational levels (lower, middle, higher level) were classified according to the educational degrees customary in Germany after 9 years, 10 years, or 12-13 years of secondary school education. Table 1 shows sociodemographic characteristics and selected health indicators of study participants compared with national population statistics in the year 2003.

Table 1.—. Sociodemographic Characteristics and Selected Health Indicators of Study Participants (n = 7341) Compared with National Statistics
 Study sampleNational statistics
  • National population statistics for age and sex and Mikrozensus 2003 for education, BMI, and smoking (national survey of 1% of all households in Germany, response rate 97%, item-response education 94.4%, BMI 73.2%, smoking 85.6%).

  • BMI = body mass index.

Age (years)  
Secondary school education  
 Higher level26.422.3
 Middle level41.327.5
 Lower level30.949.1
BMI (kg/m2)  
 BMI < 2553.950.7
 BMI 25-<30 (overweight)35.236.3
 BMI ≥ 30 (obese)10.912.9
Smoking (current daily)24.224.6

Statistical Analysis.— For statistical analyses, spss version 14.0 was used. Sampling weights were used to account for unequal sampling probabilities by household size and to reflect the distribution of the population (on December 31, 2001) with respect to age, sex, and 6 regions (the 16 German states were grouped as North, ie, Schleswig-Holstein, Free Hanseatic City of Hamburg, Free Hanseatic City of Bremen, Lower Saxony; East, ie, Mecklenburg-Western Pomerania, Saxony-Anhalt, Berlin, Brandenburg, Saxony, Thuringia; Central, ie, Hesse, Saarland, Rhineland-Palatinate; and Bavaria, Baden Wuerttemberg, and North Rhine-Westphalia). Logistic and linear regression models were calculated to test for association of age, sex, education, region (East/West and of the 6 regions, respectively, in separate models) and type of severe headache with disability days, medical consultation, and medication use among participants with severe headaches. A 2-sided significance level of P < .05 was adopted for all tests.


Prevelance.— One-year prevalence was: all headache 60.2%, total migraine 10.6%, migraine with visual aura 3.6%, severe nonmigrainous headache 24.7%, (women 66.6%, 15.6%, 5.6%, and 27.1%; men 53.0%, 5.3%, 1.5%, and 22.2%; Table 2). Migraine was 3 times more common in women than in men (P < .001). The gender difference was less marked but still significant for headaches in general (P < .001) and nonmigrainous severe headache (P < .001; Table 2). Prevalence of all headache types was highest in young and middle-aged adults up to age 50 years and declined thereafter (Table 2). While the prevalence of all migraine, including migraine with visual aura, decreased with age, the proportion of migraine with visual aura to migraine without aura increased with advancing age ( Fig.). We found almost no significant differences for headache prevalence by region with the only exception of a lower prevalence of nonmigrainous severe headache in the North region compared with Bavaria and North Rhine-Westphalia (P < .05). A higher educational level was associated with a higher prevalence of all different headache types (P < .05; Table 2). However, this social gradient remained significant only for nonmigrainous severe headaches but not for all headache and migraine after adjustment for age, sex, and region in logistic regression models (analyses not shown).

Table 2.—. Twelve-Month Prevalence of Headache and Migraine in the General Population of Germany (in % of the Whole General Population Sample, n = 7236; 3490 men, 3746 Women)
 All headachesIHS migraineNonmigrainous severe headache
  • By definition 0%.

  • F = female; IHS = International Headache Society; M = male; T = total.

Total 12-month prevalence53.066.660.25.315.610.622.227.124.7
Headache severity         
 Predominantly severe10.018.814.
 Both severe and mild17.724.
 Predominantly mild25.323.624.4
Age (years)         
Former East-Germany48.468.
Former West-Germany54.166.460.55.515.910.922.526.824.7
 North Rhine-Westphalia54.965.660.45.015.810.624.627.125.9
Secondary school education         
 Higher level57.372.
 Middle level53.069.662.15.317.412.
 Lower level47.557.552.
Figure Fig.—.

One-year prevalence of migraines with and without aura in different age groups.

Personal and Social Burden.— Severe headaches were reported by approximately 60% of headache sufferers, 30% of which were migraine (Table 2). Almost one-quarter of headache sufferers had frequent headaches (>20 days in the previous year) and 4% reported chronic headaches, defined as >10 days in the past 4 weeks (2.6% of the general population). Disability because of headaches, defined as the inability to perform usual activities at work and in everyday life during at least one day, was reported by 16.4% of all headache sufferers (Table 3). The burden of headache was greater in individuals with migraine compared with those with nonmigrainous severe headaches. Twice as many migraine sufferers reported >20 headache days in the past year than individuals with nonmigrainous severe headaches (56.6% vs 26.4%, P < .001) and almost 3 times as many migraineurs had suffered from chronic headaches (12.3% vs 4.3%, P < .001, Table 3). Disability was 3 times more frequent in migraine than in nonmigrainous severe headache (47.4% vs 15.0%, P < .001). Four times as many migraineurs reported frequent disability (>12 days in the previous year) than nonmigraineurs (11.8% vs 2.7%, P < .001; Table 3). Individuals with migraine rated their general health significantly worse than those with nonmigrainous severe headache (P < .001, Table 3). Women suffered significantly more often from severe and frequent headaches, reported significantly more often disability, and rated their health worse than men (all differences at significance level P < .001, Table 3).

Table 3.—. Impact of Headache and Migraine (in % of Individuals with Different Types of Headache)
 All headachesIHS migraineNonmigrainous severe headache
n = 1868n = 2518n = 4386n = 185n = 584n = 769n = 775n = 1014n = 1789
  • Values do not add to 100% because participants may not have had headaches in the past 4 weeks.

  • Defined as inability to perform usual activities at work and in everyday life.

  • F = female; IHS = International Headache Society; M = male; T = total.

Days with headaches (past 12 months)
 <10 days59.743.650.528.315.818.847.036.140.8
 10-20 days24.427.426.025.524.524.830.934.232.7
 >20 days16.129.023.546.259.756.522.229.726.4
Days with headaches (past 4 weeks)
 1-3 days55.855.155.447.044.745.257.557.757.6
 4-10 days9.317.514.026.832.030.710.618.014.9
 >10 days3.
Disability during at least 1 day (past 12 months)12.619.316.444.848.347.414.215.715.0
 1-5 days8.711.010.026.824.324.810.49.910.1
 6-11 days2.
 >12 days1.94.53.410.912.
Self-rated health         
 Very good/good76.570.973.371.763.265.172.572.172.2
 Bad/very bad5.

Healthcare Utilization.— Approximately one-quarter of individuals with severe headaches had consulted a physician for their headaches in the previous 12 months, women significantly more often than men (P < .001, Table 4). This relates to a healthcare use for severe headaches of 8% of the general population in the past year (Table 5). Most headache sufferers had predominantly been treated by their general practitioner (GP) or a specialist of internal medicine, while only 12% had a neurologist as their main treating physician (Table 4). More than half of headache sufferers (53%), or 24.5% of the general population, had used analgesics for headaches in the past 4 weeks, 4% of headaches sufferers reporting very frequent use (>10 times in the past 4 weeks). Three-quarters of those who had taken medication had relied on over-the-counter-medication only (Table 4).

Table 4.—. Medical Consultation and Therapy for Headache and Migraine (in % of Individuals With Different Types of Headache)
 All headachesIHS migraineNonmigrainous severe headache
n = 1868n = 2518n = 4386n = 185n = 584n = 769n = 775n = 1014n = 1789
  • Information only available for severe headaches, presented values are therefore percentages of severe headache (n = 2558).

  • For example, acupuncture, autogenic training, progressive muscle relaxation.

  • F = female; IHS = International Headache Society; M = male; T = total.

Medical consultation (past 12 months)18.526.123.335.943.341.614.316.315.4
Subspecialty mainly consulted for headaches         
 Internal/general medicine74.368.870.578.570.872.471.865.970.5
 Natural medicine1.
Use of analgesics (past 12 months)77.884.381.988.090.890.175.380.678.3
Use of analgesics (past 4 weeks)45.557.453.062.571.068.941.449.746.1
 <5 times34.539.937.836.142.240.734.238.536.6
 5-10 times8.212.510.919.
 >10 times2.
Form of medication (past 4 weeks)         
 Prescription only5.
 Both prescription and nonprescription4.
 Nonprescription only35.541.939.541.845.144.334.040.037.5
Nonpharmacological therapy (past 12 months)16.222.520.124.029.528.214.318.516.7
Table 5.—. Population Prevalence of Medical Consultation and Medication Use for Different Types of Headache (in Percent of the General Population Sample, n = 7236; 3490 Men, 3746 Women)
 All headachesIHS migraineNonmigrainous severe headache
  • Information available only for severe headache.

  • F = female; IHS = International Headache Society; M = male; T = total.

Medical consultation (past 12 months)
Use of analgesics (past 12 months)33.849.441.84.714.29.616.621.819.3
Use of analgesics (past 4 weeks)17.830.824.53.311.
 <5 times14.622.618.
 5-10 times2.
 10 times0.
 Prescription only2.
 Both prescription and nonprescription1.
 Nonprescription only14.523.419.

Migraine patients had significantly more often consulted a physician, taken analgesics, and used prescription medication than participants with nonmigrainous severe headaches (all differences at significance level P < .001, Table 4). However, only 42% of migraineurs reported medical consultation for headaches in the past year and the majority had relied solely on nonprescription medication. Less than one-quarter of migraine sufferers had used prescription medication in the past 4 weeks, while 44%, or two-thirds of those who reported medication use, had exclusively taken over-the-counter drugs (Table 4). Table 5 summarizes the distribution of medical consultation and medication use for different types of headache as percentages of the whole population sample.

We investigated sociodemographic determinants of the impact of severe headaches (number of disability days, physician consultation, and use of analgesics) first in univariate analyses and then with logistic regression models (Table 6) adjusting for age, sex, education, region (East/West or the 6 regions in separate models, respectively), and type of severe headache (Table 6). Severe disability because of headache in the past 12 months (defined as at least 12 days in the past 12 months) was associated with female sex, increasing age and type of severe headache (migraine according to IHS criteria, in particular migraine with aura, Table 6) but not with education and region in univariate analysis. However, the association with female sex became nonsignificant in multivariate analysis, while a weak association with increasing age remained and a rather strong association with migraine was confirmed. Similarly, analgesics intake and medical consultation for headache appeared to be determined mostly by the type of severe headache, that is, migraine, in particular migraine with aura, while education and region were nonsignificant. Both use of analgesics and medical consultation were more common in women, but in multivariate analysis female sex showed only a rather weak association with use of analgesics and the association with medical consultation became nonsignificant. Increasing age was weakly associated with medical consultation but not with medication use (Table 6).

Table 6.—. Factors Associated With Disability, Analgesics Intake, and Medical Consultation in Participants with Severe Headaches in the Past 12 Months
 Severe disability in the past 12 monthsAnalgesics in the past 4 weeksMedical consultation for headache in the past 12 months
  • At least 12 days with disability because of headache in the past 12 months.

  • Logistic regression model adjusting for sex, age, type of severe headache.

  • §

    Nonsignificant both in a stepwise backward logistic regression model also including age, education, region, and type of severe headache, and in multivariate linear regression with the same independent variables and the number of disability days as dependent variable.

Crude OR (95% CI)
Male1 1 1 
Age (10-year increase)1.21.1-
Type of severe headache      
 Nonmigrainous severe1 1 1 
 Migraine without aura3.42.2-
Migraine with aura7.44.8-
 Higher1 1 1 
Adjusted OR (95% CI)
Male1 1 1 
Age (10-year increase)1.31.1-
Type of severe headache      
 Nonmigrainous severe1 1 1 
 Migraine without aura4.22.7-
 Migraine with aura7.54.8-


Prevalence of Headache and Migraine.— Our study provides the first prevalence estimate of migrainous and nonmigrainous headaches for the entire adult population of Germany. The 1-year prevalence for headache in general was 60.2%, for nonmigrainous severe headache 24.7% and for migraine 10.6%. These rates are well in accordance with the reported global prevalence of active headache which, according to a recent-meta-analysis, was estimated at 51% for headaches in general and 11% for migraine, with prevalence being highest in European countries and North America and lowest in Asia and Africa.1 Another meta-analysis, including only European data, found a somewhat lower 1-year prevalence of 46% for unspecified headache and a slightly higher rate of 14% for migraine.2 In agreement with previous studies, all headache types were most frequent in the young and middle-aged. However, although the prevalence of migraine decreased with age, roughly 4% of participants aged 60 years and older still suffered from migraine, suggesting that migraine remains a substantial health problem in older age. Women were more affected by all types of headaches than men, the gender difference being most marked in migraine.1-3,22,25,29 Two older studies on prevalence of headaches including migraine in Germany have been incongruent. A small, not strictly population-based questionnaire study on 2000 men and women in former West Germany found a remarkably low prevalence of 23.4% for headache in general and 3.6% for migraine. However, the time span of the presented prevalence rate was not specified, and, although molded on the 1988 criteria of the IHS-classification for migraine,30 the questionnaire items exploring migraine characteristics were formulated more strictly than the original criteria intended.26 A second questionnaire survey with a large, however most likely more selective sample from West Germany reported similar rates as our study with 71% headache in general and 11.3% IHS migraine (n = 4061).25 The presented data were lifetime prevalences which, although more prone to recall bias, would have been expected to be higher than our 12-month prevalences. The diagnostic criteria for migraine were based on the previous IHS criteria and were similar to ours with the exception that, in our study, only individuals with severe headaches were eligible for a diagnosis of migraine, which may have lowered our prevalence rate.

Similar to previous studies,3,19,21,31,32 the prevalence of migraine with visual aura was 3.6% or one-third of total migraine with a gradual increase of this proportion with age. Of note, we did not investigate aura symptoms other than visual. Nevertheless, there is evidence that visual auras may serve as a good surrogate for any migraine aura, as visual disturbances have been reported to be the most common neurological symptoms in migraine, occurring in 99% of patients with migraine with aura (n = 16333). As in this study migraines with and without aura were considered mutually exclusive, the prevalence of migraine without aura is likely to be underestimated because some patients may suffer from migraines both with and without aura.

Personal and Social Burden.— Our data confirm that headaches constitute a considerable public health problem in Germany as they are frequent, often severe and associated with substantial disability. About 60% headache sufferers reported severe headaches and one-quarter reported more than 20 headache days in a year. Four percent of headache patients, or 2.6% of the general population, suffered from chronic headache, defined as more than 10 headache days in the previous 4 weeks. Although this category does not fulfill the IHS criteria for chronic daily headache, which is defined as >15 days of headache per month over a period of 3 months, our results closely approach those of previous studies reporting a population prevalence of 3-4% for chronic daily headache.19,34,35 The burden inflicted by headaches was greatest in migraine as reflected by a greater number of headache days, including a higher proportion of chronic headache, and a higher degree of disability compared with nonmigraineurs. Twice as many migraine patients reported frequent headaches (>20 days/year) and 4 times as many migraineurs reported frequent disability (>12 days/year) as compared with individuals with nonmigrainous severe headaches. This, in line with previous studies, shows that migraine headaches are more disabling than other headaches.4,29 As has been shown before,36 women were more severely affected than men by all headache types with regard to headache severity, frequency, and functional impairment.

Healthcare Utilization.— Migraine accounts for a considerable proportion of the healthcare burden inflicted by headaches. Overall, about one-quarter of headache sufferers, or 8% of the general population, consulted a physician for severe headaches. Most headache patients recurred to their GP or a specialist of internal medicine, and only a minority, including migraine patients, were treated mainly by a neurologist (12%). Migraine sufferers consulted more than twice as often as individuals with nonmigrainous severe headaches (42% vs 15%). Our data also reveal that despite frequent headaches and considerable disability, less than half of migraineurs sought professional care. A limitation of this finding is, however, that we only assessed medical consultation for migraine within the past year. If a migraine is sufficiently treated, regular physician visits may not be necessary which may be one of the reasons for the relatively low consultation rate for migraine headaches in one year.

Although analgesics were frequently used, the majority of headache sufferers, including migraine patients, solely relied on over-the-counter medication. Prescription medication was taken by less than 25% of migraineurs. These results are in line with previous studies showing that only about 50% of migraine patients see a doctor for headaches within a year,8,14,23,36,37 and that at least half of migraineurs treat their migraine exclusively with over-the-counter drugs.14,18,21,23,37 Apart from headache type being migraine, only increasing age was identified as a predictor for medical consultation in multivariate analysis, while educational level and region showed no association. Although women consulted more often than men, there was no association of female sex and consultation after adjusting for variables such as headache type, age, educational level, and region, suggesting that additional factors may have an influence on female consultation behavior.

Given the evidence for effective treatment with migraine-specific medication and preventive treatment strategies, our results suggest that a great proportion of migraine patients in Germany may remain suboptimally treated. These patients could benefit from consulting a physician and getting access to specific treatment. Raising awareness about the need to adequately diagnose migrainous headaches and the availability of effective migraine treatment among both patients and physicians could thus help improve medical care of affected individuals.

Strengths and Limitations of the Study.— This study surveys a large representative sample of the general population of Germany including the area of former East Germany. As the great majority of the German population is covered by fixed-line telephones and has few cell phone-only users (94.5% vs 4.2%, data according to the National Statistical Office in 2003), our fixed-line telephone sampling method is likely to have reached a representative portion of the general population. The demographic characteristics and selected healthcare indicators of responders were very similar to the distribution in the general population according to German national statistics in 2003 with the exception of an underrepresentation of participants with a lower educational level and overrepresentation of those with middle and higher educational level. There is disagreement in the literature about the prevalence of migraine among different socioeconomic classes.21,22,24,25,29 In line with a previous German study,25 our data show a higher migraine prevalence in the middle and higher socioeconomic groups, which are somewhat overrepresented in the study sample. This may lead to an overestimation of migraine prevalence; however, this overestimation is likely to be rather small as the prevalence differences between the socioeconomic groups were not large (Table 2). We report 1-year prevalences because they are less prone to recall bias than lifetime prevalences. Diagnostic criteria for migraine were based on the novel IHS criteria.27 Yet there is some potential for misclassification of headaches. Only participants reporting severe headaches were eligible for further evaluation of migraine. This may have resulted in an underestimation of the true migraine prevalence. As in any epidemiological study that solely relies on interviews without individual neurological examination, symptomatic headache could not be excluded with certainty. Symptomatic headaches, however, are rare compared with primary headaches; therefore, it is unlikely that this should have considerably modified our results.


The results of our study confirm that headache and migraine are frequent conditions in Germany that are associated with a substantial individual and social burden, especially in women. Migraine constitutes a considerable part of this burden as migraineurs are more strongly affected by headaches than individuals with nonmigrainous severe headaches as indicated by a greater number of headache days and a higher degree of disability. Accordingly, migraine accounts for a great proportion of the healthcare impact of headaches, reflected by a higher rate of medical consultation and medication use. Given the considerable adverse impact of migraine, however, the proportion of migraine patients who seek professional care is still low, and many patients may not receive optimal treatment. To improve the burden of headache and migraine in Germany, public health measures should include raising awareness among physicans and patients about the necessity to correctly identify different types of headaches in order to administer adequate treatment and counseling.


Acknowledgment: We gratefully acknowledge the funding of the GNT-HIS by the German Ministry of Health.