Sage Journals: Discover world-class research

Abstract

We evaluated the association of body mass index (BMI) with migraine and migraine specifics in a cross-sectional study of 63 467 women aged ≥ 45 years, of whom 12 613 (19.9%) reported any history of migraine and 9195 had active migraine. Compared with women without migraine and a BMI < 23 kg/m², women with a BMI ≥ 35 kg/m² had adjusted odds ratios (ORs) (95% confidence intervals) of 1.03 (0.95, 1.12) for any history of migraine. Findings were similar for active migraineurs. Women with a BMI of ≥ 35 kg/m² had increased risk for low and high migraine frequency, with the highest estimate for women who reported daily migraine. Compared with women with the lowest associated risk (migraine frequency < 6 times/year; BMI between 27.0 and 29.9 kg/m²), women with a BMI ≥ 35 kg/m² had an OR of daily migraine of 3.11 (1.12, 8.67). Among the women with active migraine, a BMI ≥ 35 kg/m² was associated with increased risk of phonophobia and photophobia and decreased risk of a unilateral pain characteristic and migraine aura. Our data confirm previous findings that the association between BMI with migraine is limited to migraine frequency and specific migraine features.

Keywords

Migraine overweight obesity women epidemiology

The prevalence of obesity is constantly increasing in the USA and worldwide (1–3). The World Health Organization defines overweight as body mass index (BMI) between 25.0 and 29.9 kg/m² and obesity as BMI ≥ 30 kg/m² (2). Approximately 66% of adult Americans are at least overweight and over 30% are obese, and these figures are rising (1). The consequences of overweight and obesity include increased risk of diabetes, dyslipidaemia, hypertension, cardiovascular disease and cancer. Furthermore, obesity has been associated with chronic pain conditions.

An association between obesity and migraine has been the focus of attention of recent studies (4–11) that is supported by plausible biological mechanisms (12). In particular, an association between obesity and high migraine frequency, migraine features such as photophobia and phonophobia, and migraine severity has been suggested (4, 8). However, the association between BMI and migraine features is still conflicting (9, 13), and the shape of the association between BMI and migraine frequency remains unclear. A recent study has suggested that this association may be J-shaped (11).

A link between obesity and migraine would be of importance because the increasing prevalence of obesity (1–3, 14) may lead to an increase in the prevalence, frequency and severity of migraine and may be a potential target for prevention. Furthermore, such a link may increase the risk of comorbidities that have recently been associated with migraine, such as ischaemic stroke (15–17) and other ischaemic vascular events (18, 19). We thus aimed to evaluate the association between BMI and migraine, migraine frequency and migraine-related features in > 65 000 female health professionals in the USA.

Methods

Study population

This was a cross-sectional study among participants in the Women's Health Study (WHS), which was a randomized trial designed to evaluate the risks and benefits of low-dose aspirin and vitamin E in the primary prevention of cardiovascular disease and cancer among 39 876 initially apparently healthy women aged ≥ 45 years at study entry (1992–1995). The design and methods of the WHS have been described in detail previously (20–23); all participants provided written informed consent, and the Institutional Review Board of Brigham and Women's Hospital, Boston, MA approved the procedures of the WHS. In brief, invitation letters were mailed to 1757 247 female health professionals, including registered nurses, licensed vocational nurses, physicians, veterinarians, pharmacists, dieticians, dentists, dental hygienists, speech/hearing/language professionals, physical therapists and radiology technologists from 52 different states (and the District of Columbia and Commonwealth of Puerto Rico). Of those, 453 787 returned the baseline questionnaire and 194 659 indicated their willingness to participate. The 65 169 women who were eligible entered the run-in phase of the trial, in which compliance with placebos intake was evaluated over 12 weeks. A total of 39 876 female health professionals were finally randomized into the WHS (20). For the purpose of this analysis, we included information on the 65 169 women who entered the run-in phase of the trial. A total of 1464 women with missing information on BMI and 238 with missing information on migraine were excluded, leaving 63 467 participants for this analysis.

Assessment of migraine

Participants were asked on the baseline questionnaire: ‘Have you ever had migraine headaches?’ and ‘In the past year, have you had migraine headaches?’. From this information, we categorized women into ‘no history of migraine’ and ‘any history of migraine’. Furthermore, we distinguished between ‘active migraine’, which includes women with self-reported migraine in the year prior to completing the baseline questionnaire, and ‘prior migraine’, which includes women who reported ever having had a migraine but none in the year prior to completing the questionnaire. Those participants who reported active migraine were asked for further details about their migraine attacks, including attack duration of 4–72 h; unilateral location of pain; pulsating quality; inhibition of daily activities; aggravation by routine physical activity; nausea or vomiting; sensitivity to light; and sensitivity to sound. Furthermore, participants with active migraine were asked whether they had an aura or any indication a migraine is coming and the frequency of their migraine (daily, weekly, monthly, every other month, < 6 times/year). In previous studies of the WHS (18, 24, 25), we have shown good agreement with modified 1988 International Headache Society (IHS) criteria for migraine in the WHS (26). Specifically, we showed that among WHS participants who reported active migraine, 83.5% fulfilled all but one IHS criteria (code 1.7, migrainous disorder), and 46.6% fulfilled all IHS criteria for migraine (code 1.1 migraine without aura) (18).

Body mass index

Participants were asked to report their height and weight on the baseline questionnaire. Based on this information, BMI, defined as weight in kilograms divided by height in meters squared (kg/m²), was calculated. We categorized BMI values into six categories: < 23.0, 23.0–24.9, 25.0–26.9, 27.0–29.9, 30.0–34.9 and ≥ 35.0 kg/m². Further dividing the leanest group into additional categories did not indicate different associations. Self-reported and directly measured weight were highly correlated (r = 0.96) in another comparable cohort of female health professionals (27).

Statistical methods

We compared baseline characteristics of participants with respect to BMI categories using analysis of variance for continuous variables and Mantel–Haenszel test for categorical variables. We used age- and multivariable-adjusted multinominal logistic regression models to evaluate the association between BMI categories and migraine status using women with a BMI of < 23 kg/m² and who had no history of migraine as the reference group. Multinominal logistic regression models were further used to evaluate the association between BMI categories and migraine frequencies among women who reported active migraine using women with a BMI of < 23 kg/m² and with a migraine frequency of < 6 times/year as the reference group. To evaluate further the shape of the BMI–migraine frequency association, we also used the group with the lowest associated risk (migraine frequency < 6 times/year and BMI between 27.0 and 29.9 kg/m²) as reference group. To evaluate the association between BMI and migraine-related features in the subgroup of participants who indicated active migraine, multivariable-adjusted unconditional logistic regression models were used. The following migraine features were evaluated: aura, sensitivity to sound, sensitivity to light, nausea/vomiting, pulsating pain, inhibition of daily activities, aggravation by routine physical activity, and unilateral pain. We calculated prevalence odds ratios (ORs) and 95% confidence intervals (CIs) for the association between BMI categories and the various migraine outcomes.

The multivariable models were adjusted for age (continuous), smoking (never, past, current < 15 cigarettes/day, ≥ 15 cigarettes/day), exercise (never/rarely, < 1 week, 1–3 times per week, ≥ 4 times per week), alcohol consumption (never, < 1/week, weekly, and daily), history of hypertension, postmenopausal status, postmenopausal hormone use (never, past, current) and history of elevated cholesterol (≥ 240 mg/dl or use of cholesterol-lowering medication).

We incorporated a missing value indicator if the number of women with missing information on covariates was > 300 or imputed a value otherwise. For all analyses, we used SAS (version 9.1; SAS Institute Inc., Cary, NC, USA); all P-values were two-tailed and P < 0.05 was considered to be statistically significant.

Results

Of the 63 467 participants, a total of 12 613 (19.9%) reported any history of migraine, of whom 9195 reported active migraine (migraine during the year prior to the baseline questionnaire). The mean BMI was 26.3 for women without a history of migraine and 26.6 for women with any history of migraine. Table 1 summarizes the comparison of the participants’ baseline characteristics with respect to the BMI categories. Compared with women with a BMI of < 23 kg/m², women with a BMI of ≥ 35 kg/m² were younger, were more likely to have a history of migraine, to have a history of hypertension or diabetes, to currently smoke fewer cigarettes, to have a history of elevated cholesterol, to drink less alcohol and were less likely to use postmenopausal hormones and to exercise.

Table 1

Baseline characteristics according to body mass index categories in the Women's Health Study (n = 63 467)

Characteristic	Body mass index (kg/m²)						P-value
Characteristic	< 23.0	23.0–24.9	25.0–26.9	27.0–29.9	30.0–34.9	≥ 35.0
No. of participants	18 301	12 475	9744	10 105	8338	4504
Mean age, years (Std)	53.5 (7.2)	54.1 (7.2)	54.4 (7.3)	54.1 (7.0)	53.6 (6.7)	52.5 (6.2)	< 0.001
Height, m (Std)	1.6 (0.1)	1.6 (0.1)	1.6 (0.1)	1.6 (0.1)	1.6 (0.1)	1.6 (0.1)	< 0.001
Weight, kg (Std)	57.5 (5.8)	64.4 (4.9)	70.3 (5.5)	76.6 (6.4)	86.1 (7.7)	105.9 (14.0)	< 0.001
No history of migraine, %	81.2	80.1	80.0	80.0	79.5	77.6	< 0.001
Any history of migraine, %	18.8	19.9	20.0	20.0	20.5	22.4
Active migraine, %	13.8	14.4	15.2	14.3	14.5	16.4	< 0.001
Active migraine with aura, %	5.5	5.6	6.0	5.8	5.6	5.7
Active migraine without aura, %	8.3	8.8	9.2	8.5	8.8	10.7
Prior migraine, %	5.0	5.5	4.9	5.7	6.0	6.0
History of hypertension∗, %	14.2	20.1	24.2	31.5	40.8	53.6	< 0.001
History of diabetes, %	0.8	1.2	2.0	3.8	6.6	10.8	< 0.001
Smoking status, %							< 0.001
Never	49.1	48.7	48.9	50.1	51.3	51.1
Past	33.5	36.1	36.3	35.6	34.9	36.7
Current < 15 cigarettes	6.4	5.6	5.3	4.9	4.5	4.3
Current ≥ 15 cigarettes	11.0	9.6	9.5	9.3	9.3	7.9
History of elevated cholesterol ≥ 240 mg/dl	22.9	28.6	31.5	34.9	35.2	34.7	< 0.001
Alcohol intake, %							< 0.001
Rarely/never	38.2	41.3	45.2	50.2	56.8	66.7
1–3 drinks/month	12.6	13.0	13.0	13.8	13.8	12.4
1–6 drinks/week	35.5	34.0	32.1	28.4	24.4	18.0
≥ 1 drinks/day	13.8	11.7	9.7	7.6	5.1	2.9
Postmenopausal	53.3	56.0	56.9	56.0	53.8	48.8	< 0.001
Postmenopausal hormone use, %							< 0.001
Never	46.0	44.9	46.0	47.9	50.5	57.6
Past	9.0	10.6	11.0	12.0	13.1	13.7	< 0.001
Current	45.0	44.6	43.0	40.1	36.4	28.7	< 0.001
Exercise, %							< 0.001
Never	32.8	34.9	37.6	43.2	49.2	58.3
< Weekly	17.8	20.1	20.7	21.5	21.8	21.5
Weekly	34.5	33.6	32.7	28.4	24.1	16.7
Daily	14.9	11.3	9.1	6.9	4.9	3.4

^∗Defined as previous diagnosis of hypertension, systolic blood pressure ≥ 140 mmHg, diastolic blood pressure ≥ 90 mmHg, or receipt of blood pressure lowering treatment at baseline.

Table 2 summarizes the association between BMI and migraine status. After adjustment for age and compared with women without a history of migraine and with a BMI of < 23.0 kg/m², increasing BMI was associated with increased risk of any history of migraine. Women with a BMI between 30.0 and 34.9 kg/m² had an OR of any migraine history of 1.11 (95% CI 1.04, 1.19) that further increased to an OR of 1.21 (95% CI 1.12, 1.32) for women with a BMI ≥ 35.0 kg/m². The effect estimates were very similar when we made a distinction between active migraine and prior migraine. The association between BMI and migraine, however, entirely disappeared after controlling for smoking, exercise, alcohol consumption, history of hypertension, postmenopausal status, postmenopausal hormone intake and history of elevated cholesterol levels. Compared with women without migraine and a BMI < 23 kg/m², women with migraine and a BMI of ≥ 35.0 kg/m² had a multivariable-adjusted OR of 1.03 (95% CI 0.95, 1.12). Similarly, there was no association between BMI and active migraine or prior migraine (history of migraine but no indication of migraine in the 1 year prior to the baseline questionnaire).

Table 2

Associations between body mass index categories and migraine status in the Women's Health Study (n = 63 467)

Body mass index, kg/m²	No history of migraine		Any history of migraine			Active migraine			Prior migraine∗
No.	50854		12613			9195			3418
No.	No.	%	No.	%	OR (95% CI)	No.	%	OR (95% CI)	No.	%	OR (95% CI)
Age-adjusted
< 23.0	14860	29.2	3441	27.3	1.00	2520	27.4	1.00	921	27.0	1.00
23.0–24.9	9990	19.6	2485	19.7	1.09	1801	19.6	1.09	684	20.0	1.10
23.0–24.9	9990	19.6	2485	19.7	(1.03, 1.16)	1801	19.6	(1.02, 1.17)	684	20.0	(0.99, 1.21)
25.0–26.9	7792	15.3	1952	15.5	1.11	1479	16.1	1.17	473	13.8	0.97
25.0–26.9	7792	15.3	1952	15.5	(1.04, 1.18)	1479	16.1	(1.09, 1.25)	473	13.8	(0.86, 1.08)
27.0–29.9	8083	15.9	2022	16.0	1.10	1449	15.8	1.09	573	16.8	1.14
27.0–29.9	8083	15.9	2022	16.0	(1.03, 1.17)	1449	15.8	(1.01, 1.17)	573	16.8	(1.02, 1.27)
30.0–34.9	6633	13.0	1705	13.5	1.11	1206	13.1	1.08	499	14.6	1.21
30.0–34.9	6633	13.0	1705	13.5	(1.04, 1.19)	1206	13.1	(1.00, 1.16)	499	14.6	(1.09, 1.36)
≥ 35.0	3496	6.9	1008	8.0	1.21	740	8.1	1.20	268	7.8	1.26
					1.21			(1.10, 1.32)			(1.09, 1.45)
					(1.12, 1.32)			(1.10, 1.32)			(1.09, 1.45)
Multiple-adjusted†
< 23.0	14860	29.2	3441	27.3	1.00	2520	27.4	1.00	921	27.0	1.00
23.0–24.9	9990	19.6	2485	19.7	1.05	1801	19.6	1.05	684	20.0	1.06
23.0–24.9	9990	19.6	2485	19.7	(0.99, 1.12)	1801	19.6	(0.99, 1.13)	684	20.0	(0.95, 1.17)
25.0–26.9	7792	15.3	1952	15.5	1.05	1479	16.1	1.10	473	13.8	0.91
25.0–26.9	7792	15.3	1952	15.5	(0.98, 1.12)	1479	16.1	(1.03, 1.18)	473	13.8	(0.82, 1.03)
27.0–29.9	8083	15.9	2022	16.0	1.00	1449	15.8	0.99	573	16.8	1.03
27.0–29.9	8083	15.9	2022	16.0	(0.94, 1.06)	1449	15.8	(0.92, 1.06)	573	16.8	(0.92, 1.15)
30.0–34.9	6633	13.0	1705	13.5	0.98	1206	13.1	0.95	499	14.6	1.06
30.0–34.9	6633	13.0	1705	13.5	(0.91, 1.05)	1206	13.1	(0.88, 1.03)	499	14.6	(0.94, 1.19)
≥ 35.0	3496	6.9	1008	8.0	1.03	740	8.1	1.03	268	7.8	1.04
≥ 35.0	3496	6.9	1008	8.0	(0.95, 1.12)	740	8.1	(0.94, 1.14)	268	7.8	(0.90, 1.21)

OR, odds ratio; CI, confidence interval. Odds ratios are calculated by using a multinominal logistic regression model with women without history of migraine who had body mass index of < 23 kg/m² as reference group.

^∗Women who indicated a history of migraine but no migraine in the year preceding the baseline questionnaire.

^†Adjusted for age, smoking, exercise, alcohol consumption, history of hypertension, postmenopausal status, postmenopausal hormone use and history of elevated cholesterol.

With regard to the association between BMI and migraine frequency, we found a J-shaped pattern. Among women with active migraine and compared with women with a BMI < 23 kg/m² and a migraine frequency of < 6 times/year, the multivariable-adjusted ORs (95% CI) for women with BMI ≥ 35 kg/m² were 0.58 (0.42, 0.80) for a migraine frequency of every other month, 0.60 (0.48, 0.76) for a monthly migraine frequency, 0.59 (0.39, 0.87) for a weekly migraine frequency and 1.46 (0.63, 3.37) for a daily migraine frequency (Table 3). Using the BMI category with the lowest associated risk as the reference group (BMI between 27.0 and 29.9), emerged the J-shaped association between BMI and a daily migraine frequency (Fig. 1), with the highest risk estimate for the high-frequency group (OR 3.11, 95% CI 1.12, 8.67).

Table 3

Adjusted∗ association between body mass index and frequency of migraine among women with active migraine in the Women's Health Study (n = 9195)

Body mass index, kg/m²	< 6 times/year		Every other month			Monthly			Weekly			Daily
No. of participants	5624		901			1955			506			59
No. of participants	No.	%	No.	%	OR (95% CI)	No.	%	OR (95% CI)	No.	%	OR (95% CI)	No.	%	OR (95% CI)
< 23.0	1487	26.4	274	30.4	1.00	560	28.6	1.00	142	28.1	1.00	17	28.8	1.00
23.0–24.9	1089	19.4	176	19.5	0.88	400	20.5	0.99	95	18.8	0.87	12	20.3	0.89
23.0–24.9	1089	19.4	176	19.5	(0.72, 1.08)	400	20.5	(0.85, 1.15)	95	18.8	(0.66, 1.14)	12	20.3	(0.42, 1.88)
25.0–26.9	919	16.3	134	14.9	0.79	313	16.0	0.91	79	15.6	0.83	7	11.9	0.58
25.0–26.9	919	16.3	134	14.9	(0.63, 0.98)	313	16.0	(0.77, 1.07)	79	15.6	(0.62, 1.11)	7	11.9	(0.24, 1.41)
27.0–29.9	889	15.8	148	16.4	0.89	300	15.4	0.88	85	16.8	0.87	6	10.2	0.47
27.0–29.9	889	15.8	148	16.4	(0.71, 1.11)	300	15.4	(0.75, 1.04)	85	16.8	(0.65, 1.16)	6	10.2	(0.18, 1.21)
30.0–34.9	739	13.1	113	12.5	0.80	257	13.2	0.87	69	13.6	0.81	6	10.2	0.58
30.0–34.9	739	13.1	113	12.5	(0.63, 1.02)	257	13.2	(0.73, 1.04)	69	13.6	(0.59, 1.11)	6	10.2	(0.22, 1.52)
≥ 35.0	501	8.9	56	6.2	0.58	125	6.4	0.60	36	7.1	0.59	11	18.6	1.46
≥ 35.0	501	8.9	56	6.2	(0.42, 0.80)	125	6.4	(0.48, 0.76)	36	7.1	(0.39, 0.87)	11	18.6	(0.63, 3.37)

OR, odds ratio; CI, confidence interval. Odds ratios are calculated by using a multinominal logistic regression model with women with a migraine frequency of < 6 times/year and a body mass index of < 23 kg/m² as reference group.

^∗Adjusted for age, smoking, exercise, alcohol consumption, history of hypertension, postmenopausal status, postmenopausal hormone use and history of elevated cholesterol.

Figure 1

Association between body mass index and daily migraine frequency. Odds ratios are calculated from a multinominal logistic regression model with women with a migraine frequency of < 6 times/year and a body mass index between 27 and 29.9 kg/m² as reference group and adjusted for age, smoking, exercise, alcohol consumption, history of hypertension, postmenopausal status, postmenopausal hormone use, history of elevated cholesterol and other migraine frequencies.

Table 4 summarizes the association between BMI and migraine-related features among the women who reported active migraine at baseline. An increasing BMI was associated with an increased risk of sensitivity to sound as well as sensitivity to light that remained when adjusting for potential confounders. Among women with active migraine and compared with women with a BMI of < 23 kg/m², women with a BMI ≥ 35 kg/m² had a multivariable-adjusted OR for sensitivity to sound of 1.42 (95% CI 1.19, 1.70) and an OR of sensitivity to light of 1.39 (95% CI 1.14, 1.68), with both significant trends across BMI categories. In addition, there was a suggestion of a trend that increasing BMI is associated with pulsating pain character (P trend = 0.05) and nausea/vomiting (P trend = 0.07), but the risk estimates in the highest BMI category were only slightly increased. On the other hand, a BMI of ≥ 35 kg/m² was associated with a decreased risk of reporting a migraine aura (adjusted OR 0.82, 95% CI 0.68, 0.98; P trend = 0.11) and having unilateral pain (OR 0.81, 95% CI 0.68, 0.97; P trend = 0.001).

Table 4

Adjusted∗ association between body mass index and migraine features among women with active migraine in the Women's Health Study (n = 9195)

Associated migraine features	Body mass index (kg/m²)						P trend†
Associated migraine features	< 23.0	23.0–24.9	25.0–26.9	27.0–29.9	30.0–34.9	≥ 35.0	P trend†
	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)
Aura	1.00	0.96	0.99	1.03	0.98	0.82	0.11
Aura	1.00	(0.84, 1.08)	(0.86, 1.13)	(0.90, 1.17)	(0.85, 1.13)	(0.68, 0.98)	0.11
Sensitivity to sound	1.00	1.14	1.12	1.08	1.29	1.42	< 0.001
Sensitivity to sound	1.00	(1.01, 1.29)	(0.98, 1.28)	(0.95, 1.24)	(1.12, 1.49)	(1.19, 1.70)	< 0.001
Sensitivity to light	1.00	1.08	1.11	1.20	1.29	1.39	< 0.001
Sensitivity to light	1.00	(0.95, 1.23)	(0.97, 1.28)	(1.04, 1.38)	(1.10, 1.50)	(1.14, 1.68)	< 0.001
Nausea/vomiting	1.00	1.00	0.95	1.03	1.11	1.15	0.07
Nausea/vomiting	1.00	(0.88, 1.13)	(0.83, 1.08)	(0.90, 1.18)	(0.96, 1.29)	(0.96, 1.38)	0.07
Pulsating pain	1.00	1.06	1.09	1.00	1.23	1.12	0.05
Pulsating pain	1.00	(0.94, 1.20)	(0.96, 1.24)	(0.88, 1.14)	(1.07, 1.43)	(0.94, 1.34)	0.05
Inhibition of daily activities	1.00	1.03	0.91	1.00	1.07	1.14	0.13
Inhibition of daily activities	1.00	(0.92, 1.17)	(0.80, 1.04)	(0.88, 1.14)	(0.93, 1.24)	(0.96, 1.36)	0.13
Aggravation by routine physical activity	1.00	1.12	0.91	1.03	1.20	1.09	0.13
Aggravation by routine physical activity	1.00	(0.99, 1.27)	(0.79, 1.04)	(0.90, 1.19)	(1.04, 1.40)	(0.91, 1.31)	0.13
Unilateral pain location	1.00	1.09	0.82	0.93	0.85	0.81	0.001
Unilateral pain location	1.00	(0.96, 1.23)	(0.72, 0.93)	(0.81, 1.06)	(0.73, 0.98)	(0.68, 0.97)	0.001

OR, odds ratio; CI, confidence interval. Odds ratios are calculated by using an unconditional logistic regression model with women with a body mass index of < 23 kg/m² as reference group.

^∗Adjusted for age, smoking, exercise, alcohol consumption, history of hypertension, postmenopausal status, postmenopausal hormone use and history of elevated cholesterol.

^† P for linear trend.

Discussion

In this large cohort of apparently healthy women, the age-adjusted small increased risk of migraine associated with high BMI values diminished after controlling for potential confounders. We found a J-shaped association between BMI and migraine frequency, indicating that women with a BMI of < 23 kg/m² and a BMI of ≥ 35 kg/m² were at increased risk for having a daily migraine, with highest estimates for the high BMI category. Among women with active migraine, we found that increasing BMI was significantly associated with sensitivity to sound and light. In contrast, a BMI of ≥ 35 kg/m² was associated with a decreased reporting of migraine aura and a unilateral pain characteristic.

Consistent with our data, several studies have found no association between BMI and migraine (7, 9, 19, 28). For example, in a cross-sectional study of 30 215 participants, the adjusted OR for migraine was 0.9 (95% CI 0.6, 1.2) for participants with a BMI of ≥ 35.0 kg/m² compared with participants with a BMI between 18.5 and 24.9 kg/m² (7). In three other studies, no difference was found in mean BMI values of migraineurs and participants without migraine (9, 19, 28). In contrast, one small study has suggested that patients referred for obesity surgery had a higher prevalence of migraine compared with subjects with a BMI of < 25.0 kg/m² (5).

Our finding of a J-shaped association between BMI and a high frequency of migraine is consistent with a recent study of 18 989 migraine patients (11). The relative risk for having a migraine frequency of 10–14 days/month was 1.44 (95% CI 0.98, 1.96) for migraineurs with a BMI < 18.5 kg/m² and 1.74 (95% CI 1.41, 1.93) for migraineurs with a BMI of ≥ 35 kg/m² when compared with patients with migraine and a BMI between 18.5 and 24.9 kg/m². Other studies have shown increased relative risk estimates for increasing BMI levels. In cross-sectional study of 30 215 participants, obesity was associated with the number of reported headache days per month (7). Among migraineurs with an attack frequency of 10–14 days/month, the ORs were 2.9 (95% CI 1.9, 4.4) for being obese (BMI 30.0–34.9 kg/m²) and 5.7 (95% CI 3.6, 8.8) for being morbidly obese (BMI ≥ 35.0 kg/m²). In contrast, a recent study from Sweden that included 684 women who participated in a mammogram screening programme did not find any association between BMI and migraine frequency, severity and duration (9).

Two studies have indicated an association between obesity and chronic daily headache (CDH), which is defined as headaches occurring on ≥ 15 days/month. In the first study of 1134 cases of patients with CDH and 798 controls, overweight and obesity were associated with CDH after adjusting for potential confounders (4). Cases and controls were further followed for 11 months, during which time 23 incident cases of CDH occurred. After adjustment, obesity was associated with a fivefold increased risk of developing CDH (OR 5.28, 95% CI 1.3, 21.1) compared with participants with a BMI of < 25 kg/m². The second study was a large population-based cross-sectional study, in which BMI was associated with CDH (8). Compared with participants with a BMI between 18.5 and 24.9 kg/m², the OR of CDH was 2.0 (95% CI 1.4, 2.4) for participants with a BMI of ≥ 35 kg/m². Furthermore, the association between BMI and CDH was stronger for patients with transformed migraine compared with those with chronic tension-type headache.

Our data did not allow us to classify migraineurs as having CDH and no data on incident CDH were available.

As potential mechanisms for the observed association between BMI and increase in migraine frequency or CDH, inflammatory mediators, such as calcitonin gene-related peptide and interleukins, have been proposed that subsequently may influence migraine frequency and potentially cause central sensitization (12). However, the J-shaped association between BMI and migraine frequency, as found in our data and in another large cohort study (11), may suggest more complex mechanisms.

The only other study that evaluated the association between BMI and migraine features is the previously mentioned cross-sectional study (7). Among the subgroup of patients with migraine (n = 3791), a BMI of ≥ 35 kg/m² was associated with more severe headaches, worsened by physical activity, high disability, photophobia and phonophobia. In addition, women with a BMI of < 18.5 kg/m² had an increased risk of unilateral pain, although this finding was not statistically significant (OR 1.31, 95% CI 0.8, 1.2). Furthermore, in this study there was no association between BMI and migraine aura status. Our data confirm the association between high BMI levels and photo- as well as phonophobia and the inverse association between BMI and unilateral pain characteristic. In contrast, we found a modest inverse association with migraine aura status and we did not find associations between BMI and migraine-related inhibition of daily activities or migraine aggravation by physical activity. The nature of subgroup analyses as well as lack of clear biological mechanisms to explain the association between BMI and migraine-specific features, however, indicate caution in the interpretation of these findings.

The present study has several strengths, including the large number of participants and number of women with migraine as well as the standardized data ascertainment. In addition, we collected information about many migraine features, including migraine aura and migraine frequency. Furthermore, information about a large number of potential confounders was available, allowing us to adjust the association between BMI and migraine. In addition, the homogeneous nature of WHS participants further reduced potential confounding.

Several limitations should be considered when interpreting our results. First, information on migraine, migraine frequency and migraine features was self-reported, and misclassification is possible. However, in previous studies using data from this cohort, good agreement between migraine classification and modified 1988 IHS criteria for migraine (26) were found (18, 24, 25). In addition, our prevalence of migraine with aura is very close to that found in other population-based studies (29, 30), in particular to the 37% figure reported in the American Migraine Study II (29), although this study included participants who were ≥ 12 years old. Furthermore, we cannot exclude the possibility that migraine status was underreported. However, we have no reason to believe that underreporting would depend on specific BMI categories and this is therefore an unlikely explanation of our findings. Second, information on body weight and height was also self-reported. However, studies evaluating the validity of self-reported BMI have found excellent correlations (31–33). Third, despite our ability to control for a large number of potential confounders, we cannot exclude the possibility of residual and unmeasurable confounding since our study is observational. Further, our data were cross-sectional, not allowing us to determine the time sequence of the association between BMI and migraine. Lastly, the study population was limited to female health professionals aged ≥ 45 years, the vast majority of whom were White. Thus, the present results may not necessarily be extrapolated to other female populations, and different associations between BMI and migraine as well as migraine-specific features in younger women and men cannot be excluded.

In summary, the overwhelming evidence from epidemiological data does not support a causal relationship between BMI and occurrence of migraine. In contrast, we confirm previous findings of an association between BMI and migraine frequency and specific migraine features that warrant future investigations.

Footnotes

Competing interests

None to declare. Full disclosures for each of the authors have been made available to the Editor.

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. The Women's Health Study is supported by grants from the National Heart, Lung, and Blood Institute (HL-43851), and the National Cancer Institute (CA-47988).

References

1 National Center for Chronic Disease Prevention and Health Promotion. Overweight and obesity: U.S. obesity trends 1985–2006. [WWW document]. URL http://www.cdc.gov/nccdphp/dnpa/obesity/trend/maps/index.htm (accessed 24 April 2008).

2 World Health Organization. Global strategy on diet, physical activity, and health: obesity and overweight. [WWW document]. URL http://www.who.int/dietphysicalactivity/publications/facts/obesity/en/ (accessed 24 April 2008).

Manson

Skerrett

Greenland

VanItallie

TB.

The escalating pandemics of obesity and sedentary lifestyle. A call to action for clinicians. Arch Intern Med 2004; 164:249–58.

Scher

Stewart

Ricci

Lipton

RB.

Factors associated with the onset and remission of chronic daily headache in a population-based study. Pain 2003; 106:81–9.

Peres

Lerario

Garrido

Zukerman

Primary headaches in obese patients. Arq NeuroPsiquiatr 2005; 63:931–3.

Horev

Wirguin

Lantsberg

Ifergane

A high incidence of migraine with aura among morbidly obese women. Headache 2005; 45:936–8.

Bigal

Liberman

Lipton

RB.

Obesity and migraine: a population study. Neurology 2006; 66:545–50.

Bigal

Lipton

RB.

Obesity is a risk factor for transformed migraine but not chronic tension-type headache. Neurology 2006; 67:252–7.

Mattsson

Migraine headache and obesity in women aged 40–74 years: a population-based study. Cephalalgia 2007; 27:877–80.

10.

10 Tietjen

Peterlin

Brandes

Hafeez

Hutchinson

Martin

. Depression and anxiety: effect on the migraine–obesity relationship. Headache 2007; 47:866–75.

11.

Bigal

Tsang

Loder

Serrano

Reed

Lipton

RB.

Body mass index and episodic headaches: a population-based study. Arch Intern Med 2007; 167:1964–70.

12.

Bigal

Lipton

Holland

Goadsby

PJ.

Obesity, migraine, and chronic migraine: possible mechanisms of interaction. Neurology 2007; 68:1851–61.

13.

13 Bigal

Gironda

Tepper

Feleppa

Rapoport

Sheftell

. Headache prevention outcome and body mass index. Cephalalgia 2006; 26:445–50.

14.

Hedley

Ogden

Johnson

Carroll

Curtin

Flegal

KM.

Prevalence of overweight and obesity among US children, adolescents, and adults, 1999–2002. JAMA 2004; 291:2847–50.

15.

15 Kurth

Slomke

Kase

Cook

Lee

Gaziano

. Migraine, headache, and the risk of stroke in women: a prospective study. Neurology 2005; 64:1020–6.

16.

16 Stang

Carson

Rose

Ephross

Shahar

. Headache, cerebrovascular symptoms, and stroke: the atherosclerosis risk in communities study. Neurology 2005; 64:1573–7.

17.

17 MacClellan

Giles

Cole

Wozniak

Stern

Mtichell

. Probable migraine with visual aura and risk of ischemic stroke: the stroke prevention in young women study. Stroke 2007; 38:2438–45.

18.

Kurth

Gaziano

Cook

Logroscino

Diener

Buring

JE.

Migraine and risk of cardiovascular disease in women. JAMA 2006; 296:283–91.

19.

19 Kurth

Gaziano

Cook

Bubes

Logroscino

Diener

. Migraine and risk of cardiovascular disease in men. Arch Intern Med 2007; 167:795–801.

20.

Rexrode

Lee

Cook

Hennekens

Buring

JE.

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

21.

21 Ridker

Cook

Lee

Gordon

Gaziano

Manson

. A randomized trial of low-dose aspirin in the primary prevention of cardiovascular disease in women. N Engl J Med 2005; 352:1293–304.

22.

22 Lee

Cook

Gaziano

Gordon

Ridker

Manson

. Vitamin E in the primary prevention of cardiovascular disease and cancer: the Women's Health Study: a randomized controlled trial. JAMA 2005; 294:56–65.

23.

23 Cook

Lee

Gaziano

Gordon

Ridker

Manson

. Low-dose aspirin in the primary prevention of cancer: the Women's Health Study: a randomized controlled trial. JAMA 2005; 294:47–55.

24.

Bensenor

Cook

Lee

Chown

Hennekens

Buring

JE.

Low-dose aspirin for migraine prophylaxis in women. Cephalalgia 2001; 21:175–83.

25.

25 Cook

Bensenor

Lotufo

Lee

Skerrett

Chown

. Migraine and coronary heart disease in women and men. Headache 2002; 42:715–27.

26.

26 Headache Classification Committee of the International Headache Society. Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalalgia 1988; 8:1–96.

27.

27 Willett

Stampfer

Bain

Lipnick

Speizer

Rosner

. Cigarette smoking, relative weight, and menopause. Am J Epidemiol 1983; 117:651–8.

28.

Scher

Terwindt

Picavet

Verschuren

Ferrari

Launer

LJ.

Cardiovascular risk factors and migraine: the GEM population-based study. Neurology 2005; 64:614–20.

29.

Lipton

Stewart

Diamond

Reed

Prevalence and burden of migraine in the United States: data from the American Migraine Study II. Headache 2001; 41:646–57.

30.

Launer

Terwindt

Ferrari

MD.

The prevalence and characteristics of migraine in a population-based cohort: the GEM study. Neurology 1999; 53:537–42.

31.

Himes

Hannan

Wall

Neumark-Sztainer

Factors associated with errors in self-reports of stature, weight, and body mass index in Minnesota adolescents. Ann Epidemiol 2005; 15:272–8.

32.

32 Frazer

Ballinger

Cox

Hinds

Stuve

Gibbs

. A second generation human haplotype map of over 3.1 million SNPs. Nature 2007; 449:851–61.

33.

Pavlakis

Phillips

DiMauro

De Vivo

Rowland

LP.

Mitochondrial myopathy, encephalopathy, lactic acidosis, and strokelike episodes: a distinctive clinical syndrome. Ann Neurol 1984; 16:481–8.

Body Mass Index,Migraine,Migraine Frequency and Migraine Features in Women

Abstract

Keywords

Methods

Study population

Assessment of migraine

Body mass index

Statistical methods

Results

Discussion

Footnotes

Competing interests

Acknowledgements

References