Epidemiology of Tension-Type Headache

Context.— Tension-type headache is a highly prevalent condition. Because few population-based studies have been performed, little is known about its epidemiology.

Objectives.— To estimate the 1-year period prevalence of episodic tension-type headache (ETTH) and chronic tension-type headache (CTTH) in a population-based study; to describe differences in 1-year period prevalence by sex, age, education, and race; and to describe attack frequency and headache pain intensity.

Design.— Telephone survey conducted 1993 to 1994.

Setting.— Baltimore County, Maryland.

Participants.— A total of 13345 subjects from the community.

Main Outcome Measures.— Percentage of respondents with diagnoses of headache using International Headache Society criteria. Workdays lost and reduced effectiveness at work, home, and school because of headache, based on self-report.

Results.— The overall prevalence of ETTH in the past year was 38.3%. Women had a higher 1-year ETTH prevalence than men in all age, race, and education groups, with an overall prevalence ratio of 1.16. Prevalence peaked in the 30- to 39-year-old age group in both men (42.3%) and women (46.9%). Whites had a higher 1-year prevalence than African Americans in men (40.1% vs. 22.8%) and women (46.8% vs 30.9%). Prevalence increased with increasing educational levels in both sexes, reaching a peak in subjects with graduate school educations of 48.5% for men and 48.9% for women. The 1-year period prevalence of CTTH was 2.2%; prevalence was higher in women and declined with increasing education. Of subjects with ETTH, 8.3% reported lost workdays because of their headaches, while 43.6% reported decreased effectiveness at work, home, or school. Subjects with CTTH reported more lost workdays (mean of 27.4 days vs 8.9 days for those reporting lost workdays) and reduced-effectiveness days (mean of 20.4 vs 5.0 days for those reporting reduced effectiveness) compared with subjects with ETTH.

Conclusions.— Episodic tension-type headache is a highly prevalent condition with a significant functional impact at work, home, and school. Chronic tension-type headache is much less prevalent than ETTH; despite its greater individual impact, CTTH has a smaller societal impact than ETTH.

TENSION-TYPE headache is a highly prevalent condition that can be disabling. Published estimates of the prevalence of tension-type headache vary over a wide range from 1.3% to 65% in men and 2.7% to 86% in women.^1-12 Nine studies have used the widely accepted 1988 International Headache Society (IHS) criteria¹³ to assess the epidemiology of tension-type headache, but even among these studies, prevalence estimates vary widely.^1-9

There have been no large-scale population surveys in the United States describing the epidemiology of episodic tension-type headache (ETTH) or chronic tension-type headache (CTTH) as defined by the IHS criteria. The aims of our study were to estimate the 1-year period prevalence of ETTH and CTTH in a population sample; to describe the demographic factors that are associated with 1-year period prevalence; and to estimate the societal impact of ETTH and CTTH by assessing attack frequency, pain intensity, and disability.

The study population and sampling method have been previously described.¹⁴ In brief, a telephone interview survey was conducted in Baltimore County, Maryland, a demographically diverse area with regard to age, race, and household income. A total of 13345 interviews were completed for a participation rate of 77.4%.

Study participants were contacted by telephone between November 1993 and August 1994 and were questioned by trained interviewers using a computer-assisted format.¹⁴ After obtaining verbal informed consent, information was collected on demographic data and the number of headache types that were subjectively experienced. Over 80% of subjects experienced only 1 or 2 headache types, so subjects were asked about up to the 2 most severe headache types. Information was obtained on the attack frequency, pain intensity, pain location and quality, associated symptoms, and disability.

After detailed information was obtained by interview, each type of headache was classified according to IHS criteria.¹³ Subjects may have experienced 2 types of headache but met IHS tension-type headache criteria for none, 1, or both headache types. To estimate 1-year period prevalence, data are reported on those subjects who suffered headaches occurring in the prior year only. While strict application of the IHS criteria requires a diagnostic evaluation to exclude secondary causes of headache, such headaches are rare in the general population.³ The validity of the telephone interview diagnosis for ETTH was evaluated in subjects who agreed to complete a clinic visit. A total of 89.5% of those thought to have ETTH based on telephone interview data were diagnosed by a physician as having ETTH (W.F.S. and R.B.L., unpublished data, April 28, 1997).

Episodic tension-type headache was defined using IHS criteria as headache frequency of greater than 10 lifetime attacks, but fewer than 15 attacks per month; an average attack duration of 30 minutes to 7 days (if the subject always medicates, duration criteria were ignored); and with at least 2 quality of pain features (ie, mild to moderate pain intensity, bilateral, nonpulsatile, tight band, pressing, or tightening feeling, and no exacerbation by exercise). In addition, the headache does not have the IHS-defining features of migraine (ie, nausea or photophobia and phonophobia). Chronic tension-type headache was defined using IHS criteria, which are identical to those for ETTH except that the attack frequency was 15 or more attacks per month for at least 6 months, and 1 associated symptom of nausea, photophobia, or phonophobia was permitted. Subjects who reported 2 different headaches that both met criteria for ETTH were defined as having CTTH if the sum of the attack frequencies for the 2 headaches was 15 or more attacks per month.

Two summary measures of disability were derived based on responses to 5 questions.¹⁵ Lost workdays in the past year were estimated as the product of the reported number of headaches per year; the proportion of headaches that cause the subject to miss work for all or part of the day; and the average duration of headaches. Decreased effectiveness was estimated as reduced-effectiveness day equivalents in the past year, from the product of headache frequency and duration and 2 additional questionnaire items: the proportion of headaches that cause a decreased effectiveness level and the average proportion reduction in effectiveness at work, home, or school.

Crude 1-year period prevalences for ETTH and CTTH were estimated separately for men and women by age, race, and educational level. Subjects with ETTH and CTTH were compared by age, race, educational level, pain intensity, lost workdays, and reduced-effectiveness days using analysis of variance and contingency tables using the χ² statistic.

Adjusted prevalence ratios and 95% confidence intervals were estimated using binomial regression. Prevalence ratios were derived separately for men and women and in subjects with ETTH and CTTH in 4 different binomial regression models, controlling for age, age-squared (to allow for nonlinear relations), education, and race.

For the most severe headache, 3375 (25.3%) subjects met IHS criteria for ETTH and 183 (1.4%) met IHS criteria for CTTH. For the second headache, 2606 (19.5%) met IHS criteria for ETTH and 110 (0.8%) met IHS criteria for CTTH.

Considering the 2 most severe headache types, the overall prevalence of ETTH in the past year was 38.3%, with 5108 subjects meeting diagnostic criteria for either their first or second headache types or both. The prevalence differed by sex, age, race, and educational level (Table 1). Women had higher prevalences than men in all age, race, and educational level subgroups evaluated. Prevalence peaked in the 30- to 39-year-old age group in both men (42.3%) and women (46.9%), then declined thereafter. The overall sex prevalence ratio was 1.16. Prevalence was significantly higher in whites than in African Americans in both men (40.1% vs. 22.8%) and women (46.8% vs 30.9%).

There was a strong association between prevalence and education; prevalence increased with increasing educational levels, reaching a peak in subjects with graduate school education of 48.5% in men and 48.9% in women. The variation in prevalence with age was similar among the different education groups and the higher prevalences with increasing educational levels were evident at all ages.

A total of 297 subjects met diagnostic criteria for CTTH (2.2%). Women had a higher prevalence than men (2.8% vs 1.4%), for a prevalence ratio of 2.0. The preponderance of women was maintained in all subgroups evaluated (Table 2). African Americans had a lower prevalence of CTTH than did whites, but there was no clear relation of CTTH prevalence with age. In contrast to the relation observed between ETTH and educational level, prevalence of CTTH appeared to decline with increasing education; this was most apparent in women.

There were differences in the distribution of subjects by sex, age, and education, when comparing ETTH and CTTH. For sex, the prevalence ratio (women vs men) was 1.16 for ETTH compared with 2.0 for subjects with CTTH (P<.001). The proportion of subjects over 50 years was higher for CTTH (men, 18.8%; women, 29.6%) than for ETTH (men, 16.8%; women, 19.5%) (P=.03). Subjects with CTTH had less education, on average, than did subjects with ETTH (P<.001).

The average number (SD) of headaches reported per year among subjects with ETTH or CTTH was 30.0 (39.1) and 285.4 (80.7), respectively. Subjects with ETTH reported a mean (SD) pain intensity of 4.98 (1.99) on a 10-point scale. Subjects with CTTH reported a mean (SD) headache pain intensity of 5.55 (2.10). The mean pain intensity scores differed between subjects with ETTH vs CTTH (F test=22.60, P<.001).

A total of 8.3% of subjects with ETTH reported lost workdays due to their headaches; 43.6% reported reduced-effectiveness days. Among those with lost workdays, an average of 8.9 lost workdays was reported, while subjects with reduced-effectiveness days reported, on average, 5.0 reduced-effectiveness days per person.

A total of 11.8% of subjects with CTTH reported actual lost workdays due to their headaches; 46.5% reported reduced-effectiveness days. The subjects with actual lost workdays reported, on average, 27.4 lost workdays each, while subjects with reduced-effectiveness days reported, on average, 20.4 such days per person. The distribution of lost workdays and reduced-effectiveness days differed between subjects with ETTH and CTTH (P<.001 for both measures). Among subjects with CTTH and lost workdays or reduced-effectiveness days, 40% and 22.5%, respectively, reported 40 or more such days per year; the corresponding proportions for subjects with ETTH were 4.9% and 1.9%, respectively. In this sample, subjects with ETTH experienced 3791 lost workdays and 11325 reduced-effectiveness days, while subjects with CTTH experienced 959 and 2815 such days, respectively.

In this large, population-based study, the 1-year period prevalences were 38.3% for ETTH and 2.2% for CTTH. The prevalence of ETTH peaked in the fourth decade of life in both men and women, then declined thereafter. For ETTH, prevalence increased with education. For CTTH, prevalence was inversely related to education.

The data reveal that a significant majority of persons with ETTH (71.8%) experience their headaches 30 or fewer times per year. Our data are consistent with other studies, suggesting that ETTH most often occurs once or twice monthly.^1,9 In our study, pain intensity was assessed on a 10-point scale; if 1 to 3 is defined as "mild," 4 to 7 as "moderate," and 8 to 10 as "severe," then 24.9% of our study subjects experienced headaches of mild pain intensity, 62.2% experienced moderate pain, and only 12.8% had severe pain, also consistent with prior studies.^1,3 Finally, the data suggested that tension-type headaches have a significant impact on the individual and society, accounting for lost workdays and an even larger number of reduced-effectiveness days at work, home, and school.

The current data reveal that the epidemiology of ETTH and migraine is quite different. In prior studies, the prevalence of migraine increases with age, peaking between 35 and 45 years, after which it declines^14,16; for ETTH, the relation to age was not as strong. The prevalence of migraine is generally reported to be approximately 3-fold greater in women than men. Migraine prevalence is inversely related to socioeconomic status; prevalence decreases with increasing household income.^14,16 The prevalence of ETTH was dramatically and directly related to educational level, a surrogate for socioeconomic status.

The epidemiologic profiles of CTTH and ETTH also differed, mainly by age and sex. In women, the age distribution of CTTH was shifted toward older age groups (>50 years) in comparison with ETTH; although this pattern has not been previously noted, prior reports did not have an adequate sample size to address this issue. The relation of CTTH to educational level has not been previously reported. In several respects, CTTH had epidemiologic features that were intermediate between those of migraine and ETTH.

In conclusion, ETTH is highly prevalent in the population and causes frequent attacks. It has a significant but modest impact on most individuals; the aggregate societal impact is high because the condition is so prevalent. Chronic tension-type headache produces greater individual burdens, but is much less common and thus has a lower aggregate impact on society.