Abstract
Obesity affects more than a billion adults world-wide. In the United States it has been estimated that 33% of women fulfill criteria for total body obesity and 61% fulfill criteria for abdominal obesity. Both total body and abdominal obesity have been identified as independent risk factors for migraine in premenopausal women (1). Specifically, the risk of migraine in obese women has been shown to be increased by 38–60% in those with total body obesity (as estimated by body mass index (BMI)) (2–4) and to be increased 39% (OR 1.39; 95%CI: 1.25–1.56) in those with abdominal obesity (as estimated by waist circumference) (2). Furthermore, a recent evaluation of over 3700 premenopausal women in the OMEGA study demonstrated that the relative risk of migraine substantially increased with increasing severity of obesity. Women with class I obesity (BMI 30–34.9) had a 48% increased odds of migraine (OR 1.48; 95%CI: 1.12–1.96), while those with severe or class II obesity (BMI 35–39.9) had a 200% increased odds (OR 2.07; 95%CI: 1.27–3.39), and those with morbid or class III obesity (BMI ≥ 40) had a 275% increased odds of migraine (OR 2.75; 95%CI: 1.60–4.70) (4). Despite mounting evidence demonstrating an association between episodic migraine and obesity, data on treatment options for obese migraineurs are limited.
In this issue of Cephalalgia, Novak et al. evaluate the headache frequency, duration, disability and migraine-associated symptoms in 29 severe or morbidly obese, reproductive-aged women fulfilling migraine criteria before and after bariatric surgery (5). Two items are of note in this study. The first is that the headache frequency, headache-related disability and migraine-associated symptoms were evaluated in this cohort at 3 and 6 months after bariatric surgery. Their findings suggest that improvement in headache frequency and disability occurs as early as 3 months following bariatric surgery and is maintained 6 months after surgery. Of the 23 episodic migraineurs included in the study, the median monthly headache frequency declined from four headache days per month at baseline to two headache days per month 3 months following bariatric surgery. This declined further, to just one headache day per month, 6 months after surgery. The six women with chronic migraine also endorsed statistically significant and substantial decreases in headache frequency at 3 and 6 months following bariatric surgery.
In addition, and similar to their findings in regard to headache frequency, Novak et al. reported that headache-related disability (both the MIDAS and HIT-6) was significantly attenuated as early as 3 months following bariatric surgery (5). Notably, headache-related disability continued to be statistically significantly attenuated, as compared with baseline, 6 months later. Further, and for the first time, the authors reported significant improvements in both headache duration and the presence of migraine-associated symptoms (nausea, photo and phonophobia) in the attacks that continued to occur following surgery. Thus, Novak et al.'s study suggests that active interventions to treat severe and morbid obesity in reproductive-aged women migraineurs may result in an overall reduction in both migraine frequency and headache-related disability, as well as improvement in the presence of migraine-associated symptoms and a reduction in the headache duration in subsequent attacks that occur following surgery.
Only one previous study has evaluated the effect of bariatric surgery on headache frequency in morbidly obese episodic migraineurs. Similar to the current study by Novak et al., Bond et al. conducted an open, non-controlled, clinical trial of obese patients presenting for bariatric surgery (6). However, Bond et al. evaluated 24 morbidly obese, pre- and post-reproductive-aged men and women with episodic migraine presenting for bariatric surgery. Headache frequency was evaluated in the 3 months prior to bariatric surgery and again in the 3 months preceding the 6-month follow-up after surgery. Although significant, Bond et al.'s finding were less robust than those reported by Novak et al., with headache frequency declining from an average of 3.7 headache days per month before surgery to 2.2 headache days per month at the 6-month follow-up visit (6). As adipose tissue distribution has been shown to be substantially different in men vs. women and in pre vs. postmenopausal women (1), the less robust findings in Bond et al.'s study may, at least in part, be related to the inclusion of both women and men and those of post-reproductive age in their study, as compared with the current study by Novak et al., who included only women and only those of reproductive age.
The statistical precision and generalizability of results from small, uncontrolled clinical samples is limited. As such, neither the current clinical trial by Novak et al. nor the previous study by Bond et al. can definitively answer the question of whether bariatric surgery in morbidly obese episodic migraineurs is an effective migraine therapy. However, Novak et al.'s study is a timely contribution to advancing our knowledge on the potential effects of bariatric surgery in obese migraineurs and does provide the suggestion that it may result in improvements in headache frequency, headache duration, headache-related disability and the presence of migraine-associated symptoms. Larger, controlled studies will be needed to substantiate these findings and determine the true effect size of such associations.
This study (5) also highlights the need to identify whether non-surgical weight reduction modalities are effective in episodic migraineurs who do not qualify for bariatric surgery (those who fall into the classification of being overweight (BMI 25–29.9) or class I obesity (BMI 30–34.9) categories). Lack of physical activity has been demonstrated to be associated with a 21% increased risk of headache attacks in adult migraineurs (HR 1.209; p < .01) (7) and a 50% increased risk of migraine in adolescents (OR 1.5; 95%CI: 1.0–2.2) (3). In addition, limited data suggest that aerobic activity may reduce headache frequency in episodic migraineurs (8–11). So should exercise be part of the non-pharmacological approach to the management of episodic migraine? And if exercise is effective, is it exercise per se or is it weight reduction that accounts for any improvements in migraine and migraine-related symptoms?
Further, what about diet? Physicians have been debating for over a century whether abnormalities in fat or protein metabolism contribute to migraine, as well as which diets are the most effective treatment options for migraineurs. In 1873 a physician named Fothergill was quoted as saying of butter: ‘nothing more speedily and effectually gives the sick headache’. Later, in 1925 Curtis-Brown was reported to have considered that the sole cause of migraine was an inherited or acquired defect in protein metabolism, while in 1926, Marmaduke Fawkes argued that ‘the withdrawal of fats from the diet and their replacement with glucose together with large doses of alkalis, is often successful in averting the crisis [of migraine]’ (12). Today, physicians continue to suggest low fat diets (13), low protein diets (14), and even ketogenic diets as possible migraine preventive therapies (15–17). Unfortunately, these studies have been fraught with methodological limitations or variable results, and we remain without definitive scientific evidence substantiating the efficacy of any particular diet as part of migraine therapy.
In summary, Novak et al.'s study (5) highlights that obesity, as a potentially modifiable risk factor for episodic migraine, deserves special consideration by clinicians treating migraineurs, and that many questions remain to be answered. Although the existing data are suggestive, they are not substantial enough to know whether bariatric surgery will be effective in the management of episodic migraine in those with morbid obesity. However, given Novak et al.'s findings, we will wait with hopeful anticipation for larger controlled studies, with longer follow-up periods, to definitively determine the effect size between bariatric surgery and episodic migraine, and to determine whether the results are sustained long-term or whether headache frequency and disability rebounds with post-surgery complications, such as weight gain. These data should also serve to motivate the design and conduct of studies that evaluate the efficacy of non-surgical weight reduction, such as diet and exercise, for the management of migraine among obese patients of all classifications. Pending such studies, we should actively provide our episodic migraine patients with education on the risk of migraine in those who are obese; we should appropriately tailor our choices of migraine preventive medications based on our patients' obesity status, given that many have the ability to have both a positive and a negative impact on weight; and we should promote that our patients maintain healthy life-style choices in both their diet and their exercise routines.