Obesity and migraine

Population sample.

A validated computer-assisted telephone interview (CATI)^11,12 was conducted from 1997 to 2000 in three large metropolitan U.S. areas to identify individuals with selected health problems, including primary headache disorders. The study was performed in participants of Baltimore, Atlanta, and Philadelphia, selected to facilitate the conduct of population-based clinical trials.¹³

Households were selected using random digit dialing methods. No less than 10 attempts were made to contact each household. At the time of initial telephone contact, a census of the household was obtained from the person who answered the phone. Gender and age of each eligible household member were ascertained. Finally, age-eligible individuals (≥18 years) were selected for the interview. Verbal informed consent (institutional review board approved) was obtained, and the purpose of the survey was described to the respondent.

CATI.

Respondents who agreed to participate in the research were subsequently scheduled for an interview with trained interviewers, using a CATI and computer scripts. The validity of the CATI for the diagnosis of migraine was previously assessed by comparing the telephone-based diagnosis for migraine with an independent diagnosis assigned by a headache specialist. For the diagnosis of migraine, the CATI had high sensitivity (91%) and specificity (98%).^11,12

In the CATI headache interview, participants were first asked if they had at least one headache not due to a head injury, hangover, pregnancy, or an illness such as a cold or flu. They were subsequently asked if they had at least five headaches in the previous year. For those who responded positively, they were then asked about how many different types of headache they had. Detailed questions were initially asked about the most severe headache type that the respondent had in the last 12 months. The most severe headache type was initially self-defined. If respondents had a second and different self-defined headache, the same questions were also asked about this headache.

Questions covered all the diagnostic features of migraine with and without aura, as specified by the first edition of the International Classification of Headache Disorders (ICHD-1; 1988).¹⁴ The criteria for migraine remain essentially unchanged in the second edition of the ICHD (2004).¹⁵ Each headache feature reported in the affirmative (ever vs never) was followed by a question about how often the feature occurred with the specific headache type. The response options included never, rarely, less than half the time, and half the time or more. When nausea, photophobia, or phonophobia was reported in the affirmative, a follow-up question was also asked about the severity (mild, moderate, or severe) of the symptom. Subjects with CDH, defined by 15 or more headache days per month, were excluded from the analyses.

The survey also assessed demographic information (age, gender, race, educational level, marital status) and health status (history of several other medical conditions). Respondents were requested to provide their weight and height at the time of the interview. Headache severity and headache-related disability were assessed in a 10-point scale (from no severe/disabling at all to as severe/disabling it could be). Finally, the questionnaire assessed the amount of over-the-counter, prescribed analgesic medication for headache and prescribed analgesic medication used for other pain conditions, in the prior month and prior 3 months.

Analysis.

Analyses were performed using Stata (Intercooled Stata 6.0 for Windows, College Station, TX). Data were summarized using frequency counts and descriptive statistics. BMI was calculated according to the following formula: BMI = (wt [lbs]/ht² [in]) * 703. We defined five categories based on BMI: underweight (<18.5), normal weight (18.5 to 24.9), overweight (25 to 29.9), obese (30 to 34.9), and morbidly obese (≥35). The χ² test was used to compare proportions. We modeled headache features (frequency of headache, duration of headache, headache-related disability, presence and severity of associated symptoms) as dependent variables, using BMI, use of acute or preventive medication, age, race, and socioeconomic status as dependent variables. Based on the 10-point severity scale, headaches were defined as mild (scored from 1 to 3), moderate (scored from 4 to 7), or severe (scored from 8 to 10). Similar definition was used for disability. Multivariate logistic regression was used to estimate the odds ratios (ORs) for each explanatory variable. Continuous independent variables were evaluated for nonlinearity using squared and higher-order terms. Backwards stepwise maximum likelihood estimation was used to arrive at a parsimonious model.

Description of the sample.

Our sample consists of 30,215 subjects with complete CATI headache history and who provided weight and height. The demographic characteristics of our sample are summarized in table 1. In brief, respondents were predominantly female (65%) and Caucasian (65.7%). Age ranged from 18 to 89 years (mean 38.7). The majority had completed high school. Overall, most of our sample had a normal BMI (51.3%). A total of 30.6% were overweight, 10.4% were obese, and 4.6% were morbidly obese. A minority of subjects (3.1%) was underweight (table 1).

Persons with migraine had a median of 33 headache days per year, and control subjects had a median of 30 nonmigraine headache days per year. A total of 1,246 participants had 15 or more days of headache per month and were excluded from further analyses. Consistently, a total of 25,150 individuals were defined as controls. Control subjects had no headaches on the prior year or had episodic headaches (<15 days per month) not filling criteria for migraine (table 1).

Headache status and BMI.

Migraine was diagnosed in 3,791 individuals (1-year prevalence of 12.5%). Table 2 shows migraine prevalence in female and males, by demographic features. The crude prevalence of migraine was higher in underweight subjects (15.8%) than the normal weight group (13.1%; p < 0.05). Relative to normal weighted, migraine prevalence was not significantly different for overweight (11.8%), obese (11.9%), or morbidly obese (14%) subjects. This pattern held true after stratifying the prevalence by gender. For women, in the normal weight group, the prevalence of migraine was 15.5% and was not significantly different in the other weight groups. In men of normal weight, the crude prevalence of migraine was 7.2%; it was higher only in those with morbid obesity (8.8%; p < 0.01). However, after adjusting for, age, race (table 2), and education (data not shown), the prevalence of migraine was not significantly different by BMI group overall or in any age or gender.

Headache frequency and BMI.

Although the prevalence of migraine did not increase as a function of body weight, it was associated with high attack frequency (figure). With use of the normal weight group as a reference, whereas just 4.4% of them had 10 to 15 headache days per month, 5.8% of the overweight (OR = 1.3, 95% CI 1.1 to 1.9), 13.6% of the obese (OR = 2.9, 1.9 to 4.4), and 20.7% of the morbidly obese (OR = 5.7, 3.6 to 8.8) had it. Consequently, fewer individuals who were obese (66.3%; OR = 0.8, 0.6 to 0.99, p = 0.04) or morbidly obese (58.1%; OR = 0.5, 0.4 to 0.8, p < 0.01) had less than 3 headache days per month compared with the normal weighted (71.6%) (table 3). No significant changes were seen in the proportion of those with intermediate frequency of attacks (3 to 9 days per month) according to the BMI group. The number of days with headache per month experienced by underweight did not significantly differ from the values for those with normal weight.

• Download figure
• Open in new tab
• Download powerpoint

Figure. Proportion of migraine subjects with 10 or more headache days per month.

Headache features and BMI.

BMI was also a predictor of a number of headache features. The proportion of subjects reporting severe migraine attacks increased with BMI group from 53% in the normal weighted group to 57% in the overweight (OR = 1.25 vs normal weight, 95% CI 1.1 to 1.5), 59% in the obese (OR = 1.31, 1.03 to 1.66), and 65% in the morbidly obese (OR = 1.9, 1.2 to 2.4). The proportion of severe attacks in the underweight group was not statistically different from that in the normal weighted.

Similarly, obese (OR = 1.5, 1.2 to 1.9) and morbidly obese (OR = 2, 1.5 to 2.8) subjects more frequently reported having missed at least 1 day of work or school because of their headaches in the prior 3 months compared with the normal weight group. The other groups did not significantly differ.

Exacerbation by physical activities was also more common in the morbidly obese vs the normal weighted (OR = 1.7, 1.2 to 2.2), with nonsignificant trends in the overweight and underweight groups.

The proportion of subjects reporting that typically the pain was unilateral or throbbing did not significantly differ by BMI group among persons with migraine (see table E-1 on the Neurology Web site; go to www.neurology.org). Finally, the duration of pain did not vary according to the BMI group.

The BMI was significantly associated with the proportion of migraine patients that reported photophobia and phonophobia in more than 50% of their migraine attacks (table E-2). With use of the normal weight group as a reference, photophobia in 50% or more of the attacks was reported by 78% of those with normal weight vs 88% of those morbidly obese (OR = 2.0, 1.3 to 3.2). The other groups did not significantly differ. Phonophobia was reported by 75% of those with normal weight, 80% of the overweight (OR = 1.4, 1.2 to 1.7), 82% of the obese (OR = 1.5, 1.3 to 2.1), and 85% of those morbidly obese (OR = 1.8, 1.5 to 2.1). The proportion of subjects reporting nausea or visual symptoms suggestive of aura did not vary significantly by BMI group (see table E-2).

In the logistic regression model, after adjusting for covariates (gender, age, use of headache medications, sleep problems, education status, and depression), frequency of migraine attacks varied as a function of BMI (p < 0.001). BMI higher than 30 was also significantly associated with high severity of pain, high disability, and proportion of attacks accompanied by photophobia and phonophobia.