In this large, observational study of 1257 well-characterized patients with migraine and 496 non-migraine controls, we found that medical conditions (thyroid disease, peptic ulcer disease, and mitral valve prolapse), psychiatric conditions (depression, anxiety, insomnia, and subjective memory complaints), fibromyalgia, irritable bowel syndrome, chronic fatigue syndrome, and glaucoma were more prevalent among patients with migraine than controls. Further, a frequency-stratified analysis showed that migraine aura, low educational level, divorced or widowed marital status, unemployment status, peptic ulcer disease, mitral valve prolapse, asthma, depression, anxiety, insomnia, subjective memory complaints, fibromyalgia, irritable bowel syndrome, and chronic fatigue syndrome were associated at higher rates with chronic migraine than with episodic migraine. Moreover, an aura-stratified analysis showed that thyroid disease, anxiety, insomnia, subjective memory complaints, and chronic fatigue syndrome are strongly associated with patients with migraine regardless of whether patients had migraine with aura or without aura. However, current and ever smokers and those with peptic ulcer disease, mitral valve prolapse, depression, fibromyalgia, dysmenorrhea, irritable bowel syndrome, and glaucoma were more common among patients with migraine with aura (but not among patients with migraine without aura) compared with healthy controls. In addition, a sex-stratified analysis showed that anxiety, insomnia, subjective memory complaints, and chronic fatigue syndrome are strongly associated with patients with migraine, regardless of whether the patients were females or males. However, thyroid disease, peptic ulcer disease, mitral valve prolapse, and irritable bowel syndrome were high-risk factors in female (but not in male) patients compared with non-migraine controls.
In our study, we found a high prevalence of current smoking in patients with migraine with aura compared with the non-migraine control participants. This finding is consistent with a previous study conducted in Denmark, which found that smoking may be an important trigger of migraine attacks. Different hypotheses have emerged regarding the relationship between smoking and migraine: (a) smoking causes or triggers migraine, (b) migraine causes or triggers smoking, and (c) smoking and migraine share common origins/risk factors.
Regarding the first, the mechanism of impact of smoking on migraine, previous research revealed that smoking was a trigger for acute migraine attacks in patients with migraine. Second, concerning the mechanism of impact of migraine on smoking, a previous study demonstrated that pain increased urge ratings and shortened latencies to smoke. Smoking can be a pain-related coping behavior and pain a motivator of smoking. Third, migraine and smoking may share common pathophysiological mechanisms. In a previous experimental study, Hawkins et al. reported that nicotine caused peripheral and central sensitization by promoting an increase in neuron-glial signaling and elevating levels of 11 cytokines within the upper spinal cord, causing a significant elevation in phosphorylated extracellular signal-regulated kinases levels in primary trigeminal nociceptive neurons. Furthermore, Goadsby et al. showed that the pathophysiology of migraine is linked to the trigeminal innervation of cranial vessels and the trigeminovascular system. Collectively, the evidence suggests a bidirectional relationship between smoking and migraine and the presence of shared pathophysiological mechanisms.
In our study, the prevalence of medical conditions (thyroid disease, peptic ulcer disease, mitral valve prolapse, irritable bowel syndrome, chronic fatigue syndrome, and glaucoma) was significantly higher in patients with migraine in comparison with non-migraine controls, which is consistent with previous studies.[26–31] Additionally, aura and sex-stratified analyses revealed similar results in female patients with migraine and in patients with migraine with aura. By contrast, in males, only chronic fatigue syndrome was associated with migraine, while peptic ulcer disease, mitral valve prolapse, irritable bowel syndrome, and glaucoma were not associated with migraine without aura. Similar to our results, an epidemiologic study in Brazil reported that people with migraine-like headache attributed to hypothyroidism were responsive to levothyroxine. Further studies are warranted to establish a potential relationship between migraine and subclinical hypothyroidism. Ansari et al. showed that the Helicobacter pylori IgM titer was higher in patients with migraine than in controls. Furthermore, they observed a strong positive association between IgG antibodies and headache severity in both migraine with and without aura. Consistent with our findings, Gasbarrini et al. found a significantly higher prevalence of CagA-positive strains among patients with migraine with aura when compared with migraine without aura. Consistent with our findings, a national health insurance database study showed that the prevalence of migraine in a group of patients with irritable bowel syndrome was higher than that in a control group. A similar National Health Insurance Research Database study conducted in Taiwan found the risk of chronic fatigue syndrome in a migraine group was 1.5 times higher than that in a comparison group. With respect to the glaucoma finding, a previous population-based sample study revealed that the risk of developing open-angle glaucoma was 1.68 times higher in patients with migraine than in non-migraine controls.
Our study found that, overall, patients with migraine had a much higher prevalence of psychiatric conditions (e.g., depression, anxiety, insomnia, and subjective memory complaints), which is consistent with our previous research.[5,8,21] However, the subgroup analyses showed that depression was not related with migraine in patients with migraine without aura. This fits with a previous study that indicated that depression might be more strongly associated with migraine with aura than migraine without aura. Furthermore, a sex-stratified analysis revealed that depression was not significantly related to either female or male patients with migraine independently, whereas there was a significant association between migraine and depression in both groups when analyzed together.
Psychiatric conditions and migraine may influence each other via common pathophysiological mechanisms. First, central sensitization and dysregulation of the hypothalamic–pituitary–adrenal axis have been implicated in both affective disorders and migraine. Second, central monoaminergic disturbances seem to be common in migraine and depression/anxiety disorder. Third, dopaminergic polymorphisms may be related to both migraine and psychiatric comorbidities, as indicated by a study reporting that a specific dopamine D2 receptor genotype (DRD2 NcoI allele) was associated with comorbid migraine with aura, major depression, and generalized anxiety disorder.
In our study, we found that the prevalence of fibromyalgia in patients with migraine was higher than that in non-migraine controls. Consistent with our findings, Onder et al. showed that fibromyalgia was more frequent in patients with chronic migraine and in patients with migraine with aura. However, our sex-stratified analysis revealed that fibromyalgia was not significantly related to male or female patients with migraine. Interestingly, our study showed that patients with migraine with aura had a higher prevalence of dysmenorrhea. High-frequency migraine and fibromyalgia may interact through a shared pathophysiological mechanism. A previous study showed that patients with fibromyalgia comorbid with chronic migraine have a lower pain threshold than patients with episodic migraine, and that migraine patients treated with prophylactic medication have fewer fibromyalgia flares. This suggests that patients with migraine, fibromyalgia, or both have different degrees of central sensitization, and that migraine is a triggering factor of fibromyalgia.
Strengths and Limitations
One strength of this study was that it recruited a large number of healthy headache-free participants. Additional strengths included the demographic similarities of the study groups, the use of validated questionnaires and that all participants underwent a structured interview, as well as our consideration of unhealthy lifestyle factors and several medical, psychiatric, and pain conditions. Our robust statistical analysis also included sex-differentiated subgroups and migraine subgroups (with or without aura, episodic vs. chronic).
However, several limitations of this study must be considered. First, we employed a cross-sectional design, which restricts the causal relationship between migraine and other comorbidities. Hence, a longitudinal study is needed. Second, patients with migraine are more likely to seek medical care compared to healthy controls who were recruited from the community, which can confound the results. Additionally, the people with migraine were recruited from the TSGH outpatient clinic, likely limiting generalizability. Therefore, a population-based research study is also warranted. Third, unhealthy lifestyle habits were evaluated using questionnaires, and therefore, it cannot be guaranteed that the patients always provided accurate information, and recall bias is important to consider. Finally, some diagnoses of comorbid diseases were based on self-reported questionnaires and medical records (peptic ulcer disease, mitral valve prolapse, asthma, dysmenorrhea, history of head injury, and glaucoma), which might not be sufficient to ensure reliability and validity.
Headache. 2021;61(5):715-726. © 2021 Blackwell Publishing