Associations of socioeconomic status with migraine and non-migraine headache

Abstract

Background: Migraine has been linked with several measures of socioeconomic status (SES). However, results are inconsistent and data on the association between SES and non-migraine headache, migraine subtypes and migraine frequency are sparse.

Methods: We conducted a cross-sectional study among 36,858 participants in the Women’s Health Study. As proxy for SES, we calculated an SES index using annual household income and education. Migraine, migraine aura, and non-migraine headache were self-reported with good validation rates. Multinomial logistic regression models were used to evaluate the association between the SES index and the various headache forms.

Results: Of the women participating in the study, 12,140 (32.9%) reported any history of headache, 6801 (18.4%) reported any history of migraine and 5339 (14.5%) reported non-migraine headache. Women with low SES had an increased risk for all headache forms. The multivariable-adjusted odds ratios (ORs; 95% CI) were 1.22 (1.10–1.36) for non-migraine headache, 1.40 (1.28–1.54) for any migraine, 1.44 (1.23–1.69) for migraine with aura, and 1.38 (1.21–1.58) for migraine without aura. Among active migraineurs, low SES was associated with an increased OR for ≥weekly attack frequency (1.77, 1.26–2.49).

Conclusions: In this large cohort of female health professionals, low SES was associated with an increased prevalence for all headache forms and an increased migraine attack frequency.

Keywords

Migraine non-migraine headache migraine frequency socioeconomic status

Introduction

The impact of socioeconomic inequalities on various health statuses has been studied frequently (1), particularly to identify potential targets for prevention or to initiate prevention strategies. The underlying mechanisms by which socioeconomic status (SES) might be linked to diverse health outcomes are still unclear.

From a public health perspective especially, a relationship between SES and diseases with a high prevalence is of interest. Primary headache disorders are highly prevalent, particularly migraine headaches, and result in large personal and societal burden (2). Approximately 18% of females and 6% of males experience at least one migraine attack per year (3). The prevalence of active migraine changes over the lifespan with a peak in both sexes at ages 25–55, the most productive years of adulthood (4). The economic consequences of migraine are mainly reduced productivity caused by absenteeism or reduced performance at work (2).

The potential association between SES and migraine has been frequently studied with inconsistent results (5–11) and only two cross-sectional studies have evaluated migraine subtypes (12,13). In addition, studies suggest an association between low SES and chronification of headache forms (14–16). One potential mediator for this association is an increased migraine attack frequency, but only few studies have specifically addressed the relationship between SES and migraine attack frequency (6,17,18) and found inconsistent results.

The most commonly used indicators for SES are income, education, and occupation (3,5,6,8–10,17,19–21). However, socioeconomic status is a multidimensional concept, which may not be well represented by single-indicator variables. Combining SES proxies into an index has been proposed as an alternate method to measure SES and addresses the issue of multidimensionality (22–25).

We therefore aimed to evaluate the association of SES, represented by an SES index, with prevalence of migraine and non-migraine headache as well as with migraine attack frequency in a cohort of over 39,000 female health professionals.

Methods

Study population

Study subjects were all participants in the Women’s Health Study (WHS), a completed randomized trial designed to test the benefits and risks of low-dose aspirin and vitamin E in the primary prevention of cardiovascular disease and cancer among apparently healthy women. The design, methods, and results have been described in detail previously (26,27). In brief, a total of 39,876 US female health professionals aged ≥45 years at baseline (1993–1995) and without a history of cardiovascular disease, cancer, or other major illnesses were randomly assigned to active aspirin (100 mg on alternate days), active vitamin E (600 IU on alternate days), both active agents, or both placebos. All participants provided written informed consent and the Institutional Review Board of the Brigham and Women’s Hospital, Boston, Massachusetts, USA approved the procedures of the WHS. Baseline information was self-reported and collected by a mailed questionnaire that asked about many cardiovascular risk factors and lifestyle variables. Twice in the first year and yearly thereafter, participants were sent follow-up questionnaires asking about study outcomes and other information during the study period. Since the trial’s termination in 2004, observational follow-up is still ongoing.

Assessment of migraine

On the baseline questionnaire, participants were asked: ‘Have you ever had migraine headaches?’ and ‘In the past year, have you had migraine headaches?’ We distinguished between women without migraine history, with any migraine history, with active migraine (reported migraine in the year prior to completing the baseline questionnaire), and with prior migraine, which included women who reported ever having had a migraine but none in the year prior to completing the baseline questionnaire. Our ‘any migraine’ category included women with active migraine (migraine with aura, migraine without aura) and prior migraine. Those participants who reported active migraine were asked further details about their migraine attacks, including attack duration of 4–72 h; unilateral location and pulsating quality of pain; inhibition of daily activities; aggravation by routine physical activity; nausea or vomiting; sensitivity to light or sound. Furthermore, participants who reported active migraine were asked whether they had an ‘aura or any indication a migraine is coming’. Responses were used to classify women who reported active migraine into migraine with aura and migraine without aura, similar to previous studies (28,29). In addition, active migraineurs were asked details about the frequency of their migraine attacks (daily, weekly, monthly, every other month, <6 times per year).

In previous studies of the WHS, we have shown good agreement between our migraine categories and migraine classifications based on the International Headache Society criteria. Specifically, we have shown in a sub-study of the WHS that over 87% of women with active migraine could be diagnosed as having migraine without aura (71.5%) or probable migraine without aura (16.2%) when applying current revised criteria (30).

Ascertainment of non-migraine headache

Participants were asked to report on the 6-month follow-up questionnaire whether they had ever had headaches. We subsequently categorized women as having non-migraine headache if a woman reported headache but did not indicate any history of migraine similar to a previous study (29). Women who reported both migraine and non-migraine headache were classified as migraineurs.

Assessment of socioeconomic status

At enrollment, participants were asked to report the total annual household income in the year prior to the baseline questionnaire and their highest level of education. We defined six categories of income (≤US$19,999, US$20,000–29,999, US$30,000–39,999, US$40,000–49,999, US$50,000–99,999, and ≥U$ 100,000), and six levels of education (<2 years of nursing education, 2 years of nursing education, 3 years of nursing education, a bachelor’s degree, a master’s degree, and a doctoral degree. Based on this information, we created a SES index (Table 1).

Table 1.

Distribution of SES variables according to SES Index categories^a in the Women’s Health Study (N = 36,858)

		SES index
		0–3	4	5	6–7	8–10
		n = 12,805	n = 6,870	n = 5,281	n = 7,380	n = 4,522	Total
SES variable	Points^b	34.7%	18.6%	14.3%	20.0%	12.3%	N = 36,858	P value
Income (US$), %								<.0001
≤19,999	5	0.0	0.0	0.3	4.0	39.3	5.7
20,000–29,999	4	0.0	0.6	4.0	24.2	38.6	10.3
30,000–39,999	3	0.7	7.6	19.7	35.5	22.1	14.3
40,000–49,999	2	7.3	21.1	38.7	22.5	0.0	16.5
50,000–99,999	1	59.0	66.3	35.9	13.8	0.0	40.8
≥100,000	0	33.0	4.6	1.4	0.0	0.0	12.5
Contribution to index^c	–	0.34	0.33	0.38	0.44	0.48	–
Education, %								<.0001
<2 years of nursing education	5	0.0	0.0	1.4	25.0	72.8	14.1
2 years of nursing education	4	0.0	4.6	35.9	20.5	12.1	11.6
3 years of nursing education	3	7.8	66.3	38.7	44.1	15.1	31.3
Bachelor’s degree	2	41.8	21.1	19.7	9.2	0.0	23.1
Master’s degree	1	36.8	7.6	4.0	1.1	0.0	15.0
Doctoral degree	0	13.6	0.6	0.3	0.0	0.0	4.9
Contribution to the index^d	–	0.66	0.67	0.62	0.56	0.52	–

Percentages may not add up to 100 because of rounding.

Assigned point value to create the SES index.

Mean SES index score from income in SES index category divided by mean SES index score in the category.

Mean SES index score from education in SES index category divided by mean SES index score in the category.

SES: socioeconomic status.

Doctoral degree is defined as doctor of philosophy and/or medical degree.

Point values of 0–5 were assigned to each individual variable category as follows: the highest point value, point value 5, was assigned to the lowest income and education categories, respectively. The assigned point values decreased with increasing income and education levels. The highest income and education categories received the lowest point value, point value 0. We then combined the income and education variable to an overall SES index which ranged from 0–10. We divided the overall SES index into quintiles. The lowest quintile represented the highest level of SES and was used as reference category in further analysis. The assigned point scheme and the distribution of the SES variables according to the SES index categories are presented in Table 1.

Studies analyzing data from the general population often additionally include a proxy for occupation in their SES index (23,24). Since our cohort consisted of health professionals, our education and professional occupation variables are very closely related. We decided not to include occupation in our index because of the rather small variation in this traditional SES indicator.

Statistical methods

Of the 39, 876 women, a total of 3018 women were excluded [missing data on migraine status (n = 119), education (n = 669) or income (n = 2230)], leaving 36,858 participants for this analysis. Because age differed according to SES index, we used direct standardization to adjust categorical variables for age in 5-year increments to compare baseline characteristics according to SES index categories.

Multinomial logistic regression models were calculated to evaluate the association between SES index categories and headache subtypes with SES index as the independent variable and headache subtypes as the dependent variable. We ran two multinomial models, one including information on any migraine, non-migraine headache, and no history of headache and the other one including migraine with aura, migraine without aura, prior migraine, non-migraine headache, and no history of headache. Multinomial logistic regression is an extension of binary logistic regression that allows the dependent variable to have more than two categories and then simultaneously compares each of these categories to a reference category. Women within the lowest SES quintile and no history of any headache were used as reference group. Prevalence odds ratios (ORs) and 95% confidence intervals (CIs) were calculated for the association between SES index and the various headache forms.

We further evaluated the association between SES index and migraine attack frequency (daily and weekly, monthly, every other month, <6 times per year) using multinomial logistic regression models. Because of low case counts, we collapsed the SES index quintiles into a binary variable (SES index point values 8–10 versus SES index point values 0–7) for this part of the analysis. Women with a SES index between 0 and 7 with a migraine attack frequency of less than six times per year served as reference category.

We ran age- and multivariable-adjusted models. We selected covariates as potential confounders if they were related to migraine and SES status. The multivariable models were adjusted for age (continuous), race (white, non-white), marital status (single, married, divorced, widowed), geographic region (Northeast, Southeast, Midwest, West), history of hypertension, postmenopausal hormone use (never, past, current), and menopausal status (premenopausal, postmenopausal, biologically uncertain, unclear/subject unsure).To determine whether the associations were influenced by lifestyle factors, we additionally adjusted the multivariable models for smoking status (never, past, current), body mass index (BMI; <25 kg/m², 25.0–<30.0 kg/m², ≥30 kg/m²), exercise (rarely/never, <1 per week, 1–3 times per week, ≥4 times per week) and alcohol consumption (rarely/never, 1–3 drinks per month, 1–6 drinks per week, ≥1 drink per day).

To assess the impact of each single SES indicator separately, we additionally calculated multinomial logistic regression models with income and education as exposure variables.

A missing value indicator was included for a covariate if the number of women with missing information was ≥100. We assigned participants with missing values to the reference category of the covariate if the number of women with missing information was less than 100. The total number of missing values for all covariates was 2375 (6.4%).

For all analysis, we used SAS version 9.1 (SAS Institute Inc., Cary, North Carolina, USA). All p values were two-tailed and p < 0.05 was considered statistically significant.

Results

Of the 36,858 women, a total of 12,140 (32.9%) women reported any history of headache; 6,801 (18.4%) women reported any history of migraine and 5,339 (14.5%) solely reported non-migraine headache. Among the 4791 (13.0%) active migraineurs (migraine in the year previous to baseline), 1906 (39.8%) women indicated migraine with aura.

In Table 2, the participants’ age-adjusted baseline characteristics according to the SES index categories are presented. Women with low SES status (SES index 8–10) had a mean age of 57.6 years and were older compared to women with high SES status (SES index 0–3). Additionally, they were more likely to have a history of migraine and non-migraine headache. Compared to women with higher SES status, women with low SES status were more likely to have BMI values ≥30 kg/m², to currently smoke cigarettes, to rarely/never consume alcohol and to rarely/never exercise.

Table 2.

Age-adjusted baseline characteristics according to SES index categories^a in the Women’s Health Study (N = 36,858)

	SES Index
	0–3	4	5	6–7	8–10
	n = 12,805	n = 6,870	n = 5,281	n = 7,380	n = 4,522
	34.7%	18.6%	14.3%	20.0%	12.3%	P value
Demographic information
Mean age in years (SD)	52.9 (5.8)	53.8 (6.3)	54.4 (6.8)	56.5 (7.9)	57.6 (8.3)	<.0001
Ethnicity, %						<.0001
White	93.8	96.2	96.0	95.7	91.9
Hispanic	1.0	0.9	0.8	1.3	2.2
African American	2.1	1.5	2.2	2.0	5.0
Other^b	3.1	1.4	1.0	1.1	0.9
Geographic location, %						<.0001
Northeast	20.8	20.7	18.0	18.1	16.4
Southeast	22.3	23.1	24.3	23.3	24.9
Midwest	32.3	35.8	34.6	38.9	41.6
West	24.6	20.4	23.1	19.7	17.1
Marital Status, %						<.0001
Single	6.8	5.6	5.5	5.1	6.6
Married	80.1	79.2	72.7	67.0	53.1
Divorced	9.5	10.9	15.7	19.5	28.0
Widowed	3.6	4.4	6.2	8.4	12.3
Headache history, %						<.0001
Migraine with aura	5.2	4.3	5.4	5.4	6.4
Migraine without aura	7.3	7.9	7.7	8.3	8.9
Prior migraine	5.1	5.2	5.0	5.6	7.1
Non-migraine headache	14.3	13.7	14.6	14.2	16.0
No history of headache	68.1	68.9	67.4	66.5	61.6
Lifestyle variables, %
Body mass index (kg/m²)						<.0001
<25	57.4	51.0	49.1	45.8	38.3
25–30	28.6	31.0	32.2	32.2	33.5
≥30	13.9	18.0	18.7	22.0	28.3
Smoking status						<.0001
Never	54.2	48.9	51.2	49.2	46.2
Past	38.0	39.6	34.5	32.7	28.5
Current	7.8	11.5	14.3	18.1	25.2
Alcohol consumption						<.0001
Rarely/never	33.9	41.2	47.9	54.5	65.2
1–3 drinks/month	13.7	13.7	13.3	13.0	10.8
1–6 drinks/week	38.4	33.9	30.3	25.5	18.4
≥1 drink/day	14.0	11.2	8.6	7.0	5.5
Exercise						<.0001
Rarely/never	31.6	38.2	38.8	43.7	49.2
<1/week	19.2	19.8	21.2	20.6	19.3
1–3 times/week	35.4	31.3	30.9	27.4	25.2
≥4 times/week	13.7	10.8	9.1	8.3	6.2
Other covariates, %
History of hypertension						<.0001
Menopausal status						<.0001
Premenopausal	30.1	28.7	26.5	25.5	21.8
Postmenopausal	54.0	54.3	55.0	54.6	54.6
Biologicaly uncertain	11.7	13.2	14.8	16.2	19.4
Unclear/subject unsure	4.1	3.9	3.8	3.6	4.1
Postmenopausal hormone use						<.0001
Never	47.4	48.1	46.5	47.5	48.3
Past	7.7	9.4	9.7	12.8	15.3
Current	44.9	42.5	43.8	39.7	36.4

Percentages may not add up to 100 because of rounding or missing values.

Other includes Asian Pacific and American Indian.

Table 3 summarizes the associations between the SES index and headache types. Women with the lowest SES status (SES index 8–10) had a multivariable-adjusted OR (95% CI) of 1.40 (1.28–1.54) for any migraine, 1.44 (1.23–1.69) for active migraine with aura, 1.38 (1.21–1.58) for active migraine without aura, 1.36 (1.17–1.58) for prior migraine and 1.22 (1.10–1.36) for non-migraine headache when compared with women with the highest SES status who did not have a headache history (Model 1). These associations did not substantially change after additional adjustment for lifestyle factors (Model 2) and were mainly driven by income (Appendix Tables 1–2).

Table 3.

Association between SES index and headache types (N = 36,858)

SES index categories and points	No. of subjects with the specific headache type	Age-adjusted OR (95% CI)	Model 1 (95%CI)	Model 2 (95%CI)
Any migraine
0–3 (n = 12,805)	2334	1.00	1.00	1.00
4 (n = 6870)	1211	0.98 (0.90, 1.06)	0.97 (0.90, 1.05)	0.96 (0.89, 1.04)
5 (n = 5281)	955	1.04 (0.95, 1.13)	1.03 (0.95, 1.12)	1.00 (0.92, 1.09)
6–7 (n = 7380)	1359	1.12 (1.04, 1.21)	1.12 (1.04, 1.22)	1.09 (1.01, 1.18)
8–10 (n = 4522)	942	1.41 (1.29, 1.54)	1.40 (1.28, 1.54)	1.36 (1.23, 1.49)
Migraine with aura
0–3 (n = 12,805)	685	1.00	1.00	1.00
4 (n = 6870)	308	0.86 (0.75, 0.98)	0.86 (0.75, 0.99)	0.85 (0.74, 0.98)
5 (n = 5281)	286	1.08 (0.93, 1.24)	1.08 (0.94, 1.25)	1.05 (0.90, 1.21)
6–7 (n = 7380)	369	1.08 (0.95, 1.24)	1.11 (0.97, 1.27)	1.06 (0.93, 1.22)
8–10 (n = 4522)	258	1.39 (1.19, 1.61)	1.44 (1.23, 1.69)	1.38 (1.17, 1.62)
Migraine without aura
0–3 (n = 12,805)	1000	1.00	1.00	1.00
4 (n = 6870)	561	1.08 (0.96, 1.20)	1.08 (0.96, 1.20)	1.06 (0.95, 1.18)
5 (n = 5281)	408	1.06 (0.94, 1.20)	1.07 (0.94, 1.21)	1.03 (0.91, 1.17)
6–7 (n = 7380)	561	1.15 (1.03, 1.29)	1.17 (1.05, 1.31)	1.13 (1.00, 1.26)
8–10 (n = 4522)	355	1.34 (1.18, 1.53)	1.38 (1.21, 1.58)	1.33 (1.16, 1.62)
Prior migraine
0–3 (n = 12,805)	649	1.00	1.00	1.00
4 (n = 6870)	342	0.96 (0.84, 1.10)	0.93 (0.81, 1.07)	0.93 (0.81, 1.07)
5 (n = 5281)	261	0.96 (0.83, 1.12)	0.93 (0.80, 1.08)	0.91 (0.79, 1.07)
6–7 (n = 7380)	429	1.12 (0.98, 1.27)	1.07 (0.93, 1.22)	1.05 (0.92, 1.20)
8–10 (n = 4522)	329	1.47 (1.28, 1.70)	1.36 (1.17, 1.58)	1.34 (1.15, 1.56)
Non-migraine headache
0–3 (n = 12,805)	1912	1.00	1.00	1.00
4 (n = 6870)	963	0.95 (0.87, 1.03)	0.95 (0.87, 1.03)	0.93 (0.85, 1.01)
5 (n = 5281)	774	1.03 (0.94, 1.13)	1.02 (0.93, 1.12)	0.99 (0.90, 1.08)
6–7 (n = 7380)	1009	1.02 (0.93, 1.11)	1.01 (0.93, 1.11)	0.97 (0.89, 1.07)
8–10 (n = 4522)	681	1.24 (1.12, 1.36)	1.22 (1.10, 1.36)	1.16 (1.04, 1.29)

Results from multinomial logistic regression models with the headache categories as the dependent variables and SES index categories as the independent variables. One model had the headache categories ‘any migraine’, ‘non-migraine headache’, and ‘no headache’ as outcomes and another model had the migraine subgroups, instead of any migraine, as outcome. Women without headache and who have a SES index of 0–3 served as the reference group. Model 1: adjusted for age, race, marital status, geographic region, menopausal status, hormone use, and history of hypertension. Model 2: Model 1 plus alcohol, smoking, BMI, and exercise. BMI: body mass index; SES: socioeconomic status.

The associations between SES index and migraine attack frequency are summarized in Table 4. Compared with women with higher SES status (SES index 0–7) and a migraine attack frequency of less than six times per year, women with low SES status (SES index 8–10) had a multivariable-adjusted OR of 1.77 (1.26–2.49) for having at least one migraine attack per week. This association between SES index and the various headache forms was not attenuated by lifestyle factors (Model 2). In the analysis evaluating single SES proxies, income and not education was associated with an increased attack frequency (data not shown).

Table 4.

Association between SES index and migraine frequency (n = 4731)

SES Index categories and points	No. of subjects	Age-adjusted OR	Model 1	Model 2
Daily and weekly migraine attack frequency
0–7 (n = 4125)	195	1.00	1.00	1.00
8–10 (n = 606)	51	1.74 (1.25, 2.43)	1.77 (1.26, 2.49)	1.65 (1.16, 2.35)
Monthly migraine attack frequency
0–7 (n = 4125)	833	1.00	1.00	1.00
8–10 (n = 606)	102	0.88 (0.70, 1.11)	0.94 (0.74, 1.20)	0.96 (0.75, 1.22)
Every other month attack frequency
0–7 (n = 4125)	420	1.00	1.00	1.00
8–10 (n = 606)	50	0.82 (0.60, 1.13)	0.88 (0.64, 1.22)	0.95 (0.69, 1.32)
<6 times/year attack frequency
0–7 (n = 4125)	2677	1.00	1.00	1.00
8–10 (n = 606)	403	1.00	1.00	1.00

Model 1: adjusted for age, race, marital status, geographic region, menopausal status, hormone use, and history of hypertension. Model 2: Model 1 plus alcohol, smoking, BMI, and exercise. BMI: body mass index; SES: socioeconomic status.

Discussion

In this large cross-sectional study of women with a profession in the health sector, low SES status was associated with an increased OR for any evaluated headache form, including migraine with aura, migraine without aura, and non-migraine headache. Furthermore, active migraineurs with low SES had an increased OR for the ‘≥weekly’ migraine attack frequency. Additional adjustment for lifestyle factors did not substantially change the effect estimates suggesting that in our cohort the observed associations are unlikely to be explained by lifestyle habits. When comparing the two SES proxies, income and education, income predominately contributed to the observed relationship between SES status, headache prevalence, and migraine attack frequency.

The relationship between SES and migraine prevalence has been studied frequently. Several studies support our finding of an association between low SES and overall migraine prevalence (3,8,12,19,31). The majority of studies assessed the association between SES and migraine in general. In a cross-sectional study of 12,750 participants from the United States that distinguished between migraine subtypes, low income was associated with an increased OR for migraine with aura (OR: 1.68, 95% CI: 1.07–2.64), migraine without aura (OR: 1.56, 95% CI: 1.14–2.14), and other headaches with aura (OR:1.89, 95% CI: 1.14–3.13) (12). In a Danish cohort of 31,865 participants, low educational levels were associated with an increased multivariable-adjusted OR of 1.30 (1.08–1.56) for migraine without aura in women (13). In contrast, studies from Canada (9), Sweden (5) and the Netherlands (17) found no association between SES assessed by income or education and migraine prevalence.

We have additionally evaluated data on non-migraine headache. Our non-migraine headache category includes participants with headache other than migraine but the cases are not further classified according to diagnostic criteria. However, as tension-type headache (TTH) along with migraine is the most common primary headache disorder and other headache forms are very rare in the general population, we believe that our non-migraine headache group is primary composed of TTH. Several studies report an association between either low or high educational levels and TTH (32–34). Other studies could not confirm these results and found no association between SES and TTH or other non-migraine headache forms (18,20,32,35,36).

Only few studies have evaluated the relationship between SES and migraine attack frequency (6,17,18). Two studies are in line with our findings and report an association between low SES and frequent headache with migrainous features (18) and frequent headache (migraine and non-migrainous headache) (6), while a study from the Netherlands found no significant differences in attack frequency by SES (17). Data on headache frequency in the non-migraine headache group was not available in our study.

SES is a complex, multidimensional concept and is hard to measure. Most studies evaluating the association between SES and headache have used either income or education as a single proxy for SES. The application of a SES index instead of one single measure has several advantages. First, the SES index addresses SES as a multidimensional concept. Secondly, using a SES index reduces the dominance of one single SES measure and allows variation across categories. Indeed, the average contribution of income and education varies according to our SES index categories (Table 1). In the high SES index category, both income and education contribute equally to the combined SES index score while in the low SES index category education contributes about two-thirds. Single measures, e.g. education or income, represent different dimensions when used as indicators for SES. Education reflects a person’s individual intellectual abilities and may also be associated with different health behaviors. Compared to other SES proxies, education is a stable measure, but may rather represent earlier life circumstances. Annual household income is an economic indicator and several family members may unequally contribute and have access to it. Although income is a more recent indicator for SES, it is prone to fluctuations and may be a result of prevalent health problems (37).

The mechanisms by which SES might be related to migraine prevalence and migraine frequency are unclear. Stress linked to economic hardship and working conditions, diverse diet patterns, and different access to medical care may influence these associations. Furthermore, the direction of the associations still have to be determined. Bigal and colleagues (38) have evaluated two opposing hypotheses, social causation and social selection, for the association between SES and migraine. The social causation hypothesis suggests that factors associated with low SES increase disease prevalence. In contrast, according to the social selection theory, the disease itself affects the individual’s educational and occupational abilities and subsequently leads to lower SES (39). Results of the study suggest that social causation rather than social selection might contribute to the relationship between SES and migraine prevalence (38). However, most of the studies assessing SES and migraine, including our study, have a cross-sectional design that does not allow to address the directionality and causative pathways of the association.

Our study has several strengths including the large number of participants and the large number of individuals with various headache forms. Our migraine assessment has been shown to be valid (30) and we could distinguish migraine subtypes as well as migraine attack frequency. The WHS provides information about a large number of potential confounders that we could adjust for in our analyses.

The following limitations should be considered when interpreting our results. First, information on migraine, migraine features and migraine frequency is self-reported and under-reporting or misclassification is possible. Secondly, information on the SES variables and the potential confounders is also self-reported. However, as information on education and income are likely not related to headache-specific information, such misclassification is expected to be random. Lastly, our cohort consisted of female health professional aged ≥45 years which may limit the generalizability of our results to other populations.

In conclusion, data from this large cohort of female health professionals indicates an association between low SES status and any headache form. In addition, we found an association between low SES status and increased migraine attack frequency. These results support further targeted research on potential triggering factors of headache in lower SES groups and exploration of potential preventive strategies.

Footnotes

Acknowledgements

We are indebted to the participants in the Women’s Health Study for their outstanding commitment and cooperation, and to the entire Women’s Health Study staff for their expert and unfailing assistance.

Appendix Table 1.

Association between annual household income levels and headache subtypes (N = 36,858)

Income categories (US$)	No. of subjects with the specific headache type	Age-adjusted OR (95% CI)	Model 1 (95%CI)	Model 2 (95%CI)
Any migraine (n = 6,801)
≤19,999	376	1.28 (1.11, 1.47)	1.35 (1.17, 1.57)	1.24 (1.07, 1.45)
20,000–29,999	668	1.10 (0.98, 1.23)	1.13 (1.00, 1.27)	1.05 (0.92, 1.18)
30,000–39,999	946	1.04 (0.94, 1.16)	1.07 (0.96, 1.19)	1.00 (0.89, 1.12)
40,000–49,999	1113	1.00 (0.91, 1.11)	1.01 (0.91, 1.12)	0.95 (0.86, 1.06)
50,000–99,999	2795	0.99 (0.91, 1.07)	0.98 (0.90, 1.07)	0.94 (0.86, 1.03)
≥100,000	903	1.00	1.00	1.00
Migraine with aura (n = 1,906)
≤19,999	86	1.04 (0.80, 1.34)	1.19 (0.91, 1.55)	1.06 (0.81, 1.40)
20,000–29,999	191	1.07 (0.88, 1.30)	1.17 (0.95, 1.43)	1.06 (0.86, 1.30)
30,000–39,999	252	0.92 (0.77, 1.10)	0.99 (0.83, 1.19)	0.91 (0.75, 1.10)
40,000–49,999	322	0.94 (0.80, 1.11)	0.99 (0.83, 1.17)	0.91 (0.77, 1.09)
50,000–99,999	770	0.87 (0.75, 1.00)	0.88 (0.76, 1.01)	0.83 (0.72, 0.96)
≥100,000	285	1.00	1.00	1.00
Migraine without aura (n = 2,885)
≤19,999	119	1.14 (0.92, 1.43)	1.25 (1.00, 1.58)	1.13 (0.89, 1.43)
20,000–29,999	264	1.15 (0.97, 1.36)	1.21 (1.02, 1.45)	1.10 (0.92, 1.32)
30,000–39,999	394	1.11 (0.95, 1.29)	1.15 (0.99, 1.35)	1.06 (0.90, 1.24)
40,000–49,999	458	1.02 (0.88, 1.17)	1.04 (0.90, 1.21)	0.96 (0.82, 1.11)
50,000–99,999	1270	1.07 (0.95, 1.21)	1.08 (0.96, 1.22)	1.02 (0.90, 1.15)
≥100,000	380	1.00	1.00	1.00
Prior migraine (n = 2,010)
≤19,999	171	1.55 (1.25, 1.93)	1.47 (1.17, 1.85)	1.42 (1.12, 1.80)
20,000–29,999	213	1.06 (0.87, 1.29)	1.00 (0.81, 1.22)	0.97 (0.79, 1.19)
30,000–39,999	300	1.09 (0.91, 1.30)	1.03 (0.86, 1.24)	1.01 (0.84, 1.22)
40,000–49,999	333	1.05 (0.88, 1.25)	1.01 (0.84, 1.20)	0.98 (0.82, 1.18)
50,000–99,999	755	0.99 (0.85, 1.15)	0.96 (0.82, 1.11)	0.94 (0.81, 1.10)
≥100,000	238	1.00	1.00	1.00
Non-migraine headache (n = 5,339)
≤19,999	306	1.55 (1.33, 1.82)	1.63 (1.38, 1.92)	1.51 (1.28, 1.79)
20,000–29,999	516	1.26 (1.11, 1.44)	1.29 (1.13, 1.48)	1.21 (1.05, 1.39)
30,000–39,999	749	1.23 (1.09, 1.39)	1.26 (1.11, 1.42)	1.18 (1.04, 1.34)
40,000–49,999	860	1.15 (1.02, 1.29)	1.16 (1.03, 1.31)	1.09 (0.97, 1.23)
50,000–99,999	2299	1.20 (1.09, 1.33)	1.20 (1.09, 1.33)	1.15 (1.04, 1.27)
≥100,000	609	1.00	1.00	1.00

Results from multinomial logistic regression models with the headache categories as the dependent variables and income categories as the independent variables. One model had the headache categories ‘any migraine’, ‘non–migraine headache’, and ‘no headache’ as outcomes and another model had the migraine subgroups instead of any migraine as outcome. Women without headache and who have an income of ≥100,000 served as reference group. Model 1: adjusted for age, race, marital status, geographic region, menopausal status, hormone use, history of hypertension. Model 2: Model 1 plus alcohol, smoking, BMI, exercise. BMI: body mass index.

Appendix Table 2.

Association between highest education levels and headache subtypes (N = 36,858)

Education categories	No. of subjects with the specific headache type	Age-adjusted OR (95% CI)	Model 1 (95%CI)	Model 2 (95%CI)
Any migraine (n = 6801)
<2 years of nursing education	1160	1.16 (1.01, 1.33)	1.10 (0.96, 1.27)	1.06 (0.92, 1.22)
2 years of nursing education	935	1.06 (0.92, 1.22)	1.00 (0.87, 1.16)	0.97 (0.84, 1.12)
3 years of nursing education	1879	0.79 (0.69, 0.90)	0.78 (0.68, 0.88)	0.76 (0.67, 0.87)
Bachelor’s degree	1531	0.84 (0.74, 0.96)	0.83 (0.73, 0.95)	0.81 (0.71, 0.93)
Master’s degree	934	0.78 (0.68, 0.90)	0.78 (0.68, 0.89)	0.77 (0.67, 0.87)
Doctoral degree	362	1.00	1.00	1.00
Migraine with aura (n = 1906)
<2 years of nursing education	339	1.02 (0.82, 1.27)	0.98 (0.79, 1.23)	0.93 (0.74, 1.17)
2 years of nursing education	283	0.95 (0.76, 1.19)	0.90 (0.71, 1.13)	0.86 (0.68, 1.08)
3 years of nursing education	465	0.60 (0.49, 0.74)	0.60 (0.48, 0.74)	0.58 (0.47, 0.72)
Bachelor’s degree	431	0.71 (0.57, 0.87)	0.70 (0.56, 0.86)	0.68 (0.55, 0.84)
Master’s degree	268	0.67 (0.53, 0.84)	0.67 (0.54, 0.84)	0.66 (0.53, 0.83)
Doctoral degree	120	1.00	1.00	1.00
Migraine without aura (n = 2885)
<2 years of nursing education	468	1.24 (1.02, 1.53)	1.24 (1.01, 1.53)	1.20 (0.97, 1.48)
2 years of nursing education	395	1.16 (0.94, 1.42)	1.14 (0.92, 1.41)	1.10 (0.89, 1.36)
3 years of nursing education	805	0.93 (0.77, 1.13)	0.95 (0.78, 1.16)	0.93 (0.76, 1.13)
Bachelor’s degree	683	0.99 (0.81, 1.20)	1.01 (0.83, 1.23)	0.98 (0.81, 1.20)
Master’s degree	399	0.88 (0.71, 1.08)	0.89 (0.72, 1.09)	0.88 (0.71, 1.08)
Doctoral degree	135	1.00	1.00	1.00
Prior Migraine (n = 2010)
<2 years of nursing education	353	1.21 (0.96, 1.51)	1.06 (0.84, 1.33)	1.04 (0.83, 1.32)
2 years of nursing education	257	1.07 (0.85, 1.36)	0.95 (0.75, 1.21)	0.94 (0.74, 1.20)
3 years of nursing education	609	0.82 (0.67, 1.02)	0.76 (0.61, 0.94)	0.75 (0.61, 0.94)
Bachelor’s degree	417	0.82 (0.66, 1.02)	0.77 (0.61, 0.96)	0.76 (0.61, 0.95)
Master’s degree	267	0.80 (0.63, 1.00)	0.76 (0.60, 0.97)	0.76 (0.60, 0.97)
Doctoral degree	107	1.00	1.00	1.00
Non-migraine headache (n = 5339)
<2 years of nursing education	799	1.14 (0.98, 1.34)	1.11 (0.95, 1.30)	1.04 (0.88, 1.22)
2 years of nursing education	628	1.02 (0.87, 1.20)	0.98 (0.84, 1.16)	0.93 (0.79, 1.10)
3 years of nursing education	1550	0.94 (0.81, 1.08)	0.94 (0.81, 1.09)	0.90 (0.77, 1.04)
Bachelor’s degree	1283	1.01 (0.87, 1.18)	1.01 (0.87, 1.17)	0.97 (0.84, 1.13)
Master’s degree	827	0.99 (0.85, 1.16)	1.00 (0.85, 1.17)	0.98 (0.83, 1.14)
Doctoral degree	252	1.00	1.00	1.00

Results from multinomial logistic regression models with the headache categories as the dependent variables and the education categories as the independent variables. One model had the headache categories any migraine, non-migraine headache, and no headache as outcomes and another model had the migraine subgroups instead of any migraine as outcome. Women without headache and who have a doctoral degree served as reference group. Model 1: adjusted for age, race, marital status, geographic region, menopausal status, hormone use, history of hypertension. Model 2: Model 1 plus alcohol, smoking, BMI, exercise. BMI: body mass index.

References

Mackenbach

Stirbu

Roskam

. Socioeconomic inequalities in health in 22 European countries. N Engl J Med 2008; 358: 2468–2481.

Mäki

Vahtera

Virtanen

. Sickness absence among female employees with migraine and co-existing conditions. Cephalalgia 2008; 28: 1136–1144.

Lipton

Bigal

Diamond

. Migraine prevalence, disease burden, and the need for preventive therapy. Neurology 2007; 68: 343–349.

Lipton

Bigal

. The epidemiology of migraine. Am J Med 2005; 118(Suppl 1): 3S–10S.

Dahlöf

Linde

. One-year prevalence of migraine in Sweden: a population-based study in adults. Cephalalgia 2001; 21: 664–671.

Hagen

Vatten

Stovner

. Low socio-economic status is associated with increased risk of frequent headache: a prospective study of 22718 adults in Norway. Cephalalgia 2002; 22: 672–679.

Kryst

Scherl

. A population-based survey of the social and personal impact of headache. Headache 1994; 34: 344–350.

Lipton

Stewart

Diamond

. Prevalence and burden of migraine in the United States: data from the American Migraine Study II. Headache 2001; 41: 646–657.

O'Brien

Goeree

Streiner

. Prevalence of migraine headache in Canada: a population-based survey. Int J Epidemiol 1994; 23: 1020–1026.

10.

Queiroz

Peres

Piovesan

. A nationwide population-based study of migraine in Brazil. Cephalalgia 2009; 29: 642–649.

11.

Stang

Yanagihara

Swanson

. A population-based study of migraine headaches in Olmsted County, Minnesota. Case ascertainment and classification. Neuroepidemiology 1991; 10: 297–307.

12.

Carson

Rose

Sanford

. Lifetime prevalence of migraine and other headaches lasting 4 or more hours: the Atherosclerosis Risk in Communities (ARIC) study. Headache 2004; 44: 20–28.

13.

Tfelt-Hansen

Skytthe

. Association between migraine, lifestyle and socioeconomic factors: a population-based cross-sectional study. J Headache Pain 2011; 12: 157–172.

14.

Buse

Manack

Serrano

. Sociodemographic and comorbidity profiles of chronic migraine and episodic migraine sufferers. J Neurol Neurosurg Psychiatry 2010; 81: 428–432.

15.

Lipton

. Tracing transformation: chronic migraine classification, progression, and epidemiology. Neurology 2009; 72: S3–S7.

16.

Scher

Stewart

Ricci

. Factors associated with the onset and remission of chronic daily headache in a population-based study. Pain 2003; 106: 81–89.

17.

Launer

Terwindt

Ferrari

. The prevalence and characteristics of migraine in a population-based cohort: the GEM study. Neurology 1999; 53: 537–542.

18.

Scher

Stewart

Liberman

. Prevalence of frequent headache in a population sample. Headache 1998; 38: 497–506.

19.

Fernández-de-Las-Peñas

Hernández-Barrera

Carrasco-Garrido

. Population-based study of migraine in Spanish adults: relation to socio-demographic factors, lifestyle and co-morbidity with other conditions. J Headache Pain 2010; 11: 97–104.

20.

Lyngberg

Rasmussen

Jørgensen

. Incidence of primary headache: a Danish epidemiologic follow-up study. Am J Epidemiol 2005; 161: 1066–1073.

21.

Molarius

Tegelberg

Ohrvik

. Socio-economic factors, lifestyle, and headache disorders – a population-based study in Sweden. Headache 2008; 48: 1426–1437.

22.

Clarke

Miller

Chang

. Racial and social class gradients in life expectancy in contemporary California. Soc Sci Med 2010; 70: 1373–1380.

23.

James-Todd

Tehranifar

Rich-Edwards

. The impact of socioeconomic status across early life on age at menarche among a racially diverse population of girls. Ann Epidemiol 2010; 20: 836–842.

24.

von Lengerke

John

Mielck

. Excess direct medical costs of severe obesity by socioeconomic status in German adults. Psychosoc Med 2010; 7: Doc01–Doc01.

25.

Yang

Carmichael

Canfield

. Socioeconomic status in relation to selected birth defects in a large multicentered US case-control study. Am J Epidemiol 2008; 167: 145–154.

26.

Rexrode

Lee

Cook

. Baseline characteristics of participants in the Women's Health Study. J Womens Health Gend Based Med 2000; 9: 19–27.

27.

Ridker

Cook

Lee

. A randomized trial of low-dose aspirin in the primary prevention of cardiovascular disease in women. N Engl J Med 2005; 352: 1293–1304.

28.

Kurth

Kase

Schürks

. Migraine and risk of haemorrhagic stroke in women: prospective cohort study. BMJ 2010; 341: c3659–c3659.

29.

Kurth

Slomke

Kase

. Migraine, headache, and the risk of stroke in women: a prospective study. Neurology 2005; 64: 1020–1026.

30.

Schürks

Buring

Kurth

. Agreement of self-reported migraine with ICHD-II criteria in the Women's Health Study. Cephalalgia 2009; 29: 1086–1090.

31.

Stewart

Lipton

Celentano

. Prevalence of migraine headache in the United States. Relation to age, income, race, and other sociodemographic factors. JAMA 1992; 267: 64–69.

32.

Köseoglu

Naçar

Talaslioglu

. Epidemiological and clinical characteristics of migraine and tension type headache in 1146 females in Kayseri, Turkey. Cephalalgia 2003; 23: 381–388.

33.

Queiroz

Peres

Piovesan

. A nationwide population-based study of tension-type headache in Brazil. Headache 2009; 49: 71–78.

34.

Schwartz

Stewart

Simon

. Epidemiology of tension-type headache. JAMA 1998; 279: 381–383.

35.

Katsarava

Dzagnidze

Kukava

. Primary headache disorders in the Republic of Georgia: prevalence and risk factors. Neurology 2009; 73: 1796–1803.

36.

Rasmussen

. Migraine and tension-type headache in a general population: psychosocial factors. Int J Epidemiol 1992; 21: 1138–1143.

37.

Daly

Duncan

McDonough

. Optimal indicators of socioeconomic status for health research. Am J Public Health 2002; 92: 1151–1157.

38.

Bigal

Lipton

Winner

. Migraine in adolescents: association with socioeconomic status and family history. Neurology 2007; 69: 16–25.

39.

Hudson

. Socioeconomic status and mental illness: tests of the social causation and selection hypotheses. Am J Orthopsychiatry 2005; 75: 3–18.