Clinical Implications of the Association between Respiratory and Gastrointestinal Disorders in Migraine and Non-Migraine Headache Patients

Headaches, particularly migraine, are associated with gastrointestinal (GI) disorders. In addition to the gut–brain axis, the lung–brain axis is suspected to be involved in the relationship between pulmonary microbes and brain disorders. Therefore, we investigated possible associations of migraine and non-migraine headaches (nMH) with respiratory and GI disorders using the clinical data warehouse over 11 years. We compared data regarding GI and respiratory disorders, including asthma, bronchitis, and COPD, among patients with migraine, patients with nMH, and controls. In total, 22,444 patients with migraine, 117,956 patients with nMH, and 289,785 controls were identified. After adjustment for covariates and propensity score matching, the odds ratios (ORs) for asthma (1.35), gastroesophageal reflux disorder (1.55), gastritis (1.90), functional GI disorder (1.35), and irritable bowel syndrome (1.76) were significantly higher in patients with migraine than in controls (p = 0.000). The ORs for asthma (1.16) and bronchitis (1.33) were also significantly higher in patients with nMH than in controls (p = 0.0002). When the migraine group was compared with the nMH group, only the OR for GI disorders was statistically significant. Our findings suggest that migraine and nMH are associated with increased risks of GI and respiratory disorders.


Introduction
Headaches, particularly migraine, are associated with various gastrointestinal (GI) symptoms and several GI disorders. The gut-brain axis facilitates bidirectional communication between the brain and gut through neural, endocrine, and immune pathways [1][2][3].
There may be a connection between headache and the gut. GI symptoms, such as nausea and vomiting, often occur with migraine attacks. Previous studies have reported that GI disorders associated with primary headaches include dyspepsia, gastroesophageal reflux disease (GERD), constipation, functional abdominal pain, inflammatory bowel syndrome (IBS), inflammatory bowel disorders, celiac disease, and Helicobacter pylori infection [4]. In particular, an increased frequency of GI disorders has been demonstrated in patients with migraine compared with the general population. H. pylori infection, IBS, gastroparesis, hepatobiliary disorders, celiac disease, and changes in the microbiota have been linked with the occurrence of migraine, although the precise mechanisms and pathways related to the gut-brain axis in patients with migraine need to be fully elucidated [5].
In addition to the gut-brain axis, the lung-brain axis is suspected to be involved in the relationship between the central nervous system and lungs through neuroanatomical, endocrine, immune, metabolic, microbial, and gas-mediated pathways [6]. A recent report

Subjects
We analyzed clinical data from the smart CDW of the Hallym University Medical Center between January 2012 and November 2022. The smart CDW is based on the QlikView Elite Solution (Qlik, Lund, Sweden) and is used at all five Hallym University Medical Center hospitals. It offers text data and integrated analysis of fixed data from electronic medical records. Patients with migraine were eligible for inclusion if they met the following criteria: age 20-80 years; diagnosis of migraine after assessment by board-certified neurologists; presence of International Classification of Diseases tenth revision (ICD-10) code G43 in medical records; and a history of more than two outpatient visits or at least one admission to the neurology department. Patients with nMH were eligible if they were aged 20-80 years and had two diagnosis ICD-10 codes (G44 and R51) in the database, indicating tension-type headache (TTH) or non-specific headaches. Subjects with a history of more than one visit for treatment of migraine were excluded from the nMH group. The control group included patients aged 20-80 years who had undergone general health checkups at a health promotion center. Patients with a history of headaches or migraine were assessed using a basic questionnaire completed before their health checkups; patients who visited our medical center for the treatment of headaches or migraine were excluded. The enrollment process is presented in Figure 1. In total, 22,444 patients with migraine, 117,956 patients with nMH, and 289,785 controls were enrolled. After excluding subjects based on the exclusion criteria, 19,629 migraine, 86,297 nMH, and 254,274 non-headache controls were included in the study (Figure 1). The Clinical Research Ethics Committee of Chuncheon Sacred Heart Hospital, Hallym University, approved the study protocol (IRB no. 2022-10-010). Because only de-identified data were used in this study, the requirement for informed consent was waived by the review board.
subjects based on the exclusion criteria, 19,629 migraine, 86,297 nMH, and 254,274 nonheadache controls were included in the study (Figure 1). The Clinical Research Ethics Committee of Chuncheon Sacred Heart Hospital, Hallym University, approved the study protocol (IRB no. 2022-10-010). Because only de-identified data were used in this study, the requirement for informed consent was waived by the review board.

Statistical Analysis
Categorical data are presented as frequencies and percentages, whereas continuous data are presented as means and standard deviations. The chi-squared test was used to analyze categorical data, and t-tests were performed to compare continuous data among patients with migraine, patients with nMH, and controls. Because of the inability to randomize patients based on migraine or nMH status, we adjusted for covariates and selection biases using propensity scores. Propensity score matching (PSM) was performed among patients with migraine, patients with nMH, and controls using Python

Statistical Analysis
Categorical data are presented as frequencies and percentages, whereas continuous data are presented as means and standard deviations. The chi-squared test was used to analyze categorical data, and t-tests were performed to compare continuous data among patients with migraine, patients with nMH, and controls. Because of the inability to randomize patients based on migraine or nMH status, we adjusted for covariates and selection biases using propensity scores. Propensity score matching (PSM) was performed among patients with migraine, patients with nMH, and controls using Python (version 3.7, https://www.anaconda.com, accessed on 25 March 2023; Anaconda Inc., Austin, TX, USA) and Pymatch (version 0.3.4; https://github.com/benmiroglio/pymatch, accessed on 25 March 2023). Propensity scores ranged from 0.07 to 0.87; all matched cases had scores within 0.0001 of each other at a matching ratio of 1:1. This process resulted in 19,618 matched pairs of migraine patients and normal controls, 84,152 matched pairs of nMH patients and normal controls, and 19,595 matched pairs of migraine patients and nMH patients. Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated using logistic regression to measure the associations between exposure and outcome for respiratory disorders (such as asthma, bronchitis, and COPD) and GI disorders (such as GERD, gastritis, FGID, and IBS) compared with their respective controls. After PSM, adjusted ORs were calculated for each disorder among the three groups using covariates and propensity scores [19,20]. The Bonferroni correction was applied to correct for multiple testing, and p-values < 0.002 were considered significant. This value was derived by dividing the p-value threshold of 0.05 by the number of tests performed, i.e., 21 (seven outcome variables tested in three groups). All p-values were two sided. SPSS software (version 24.0; IBM Corp., Armonk, NY, USA) was used for statistical analyses.

Subject Characteristics
Among the 19,629 enrolled patients with migraine, 14,357 were women (73.1%), with a mean age of 44.5 ± 14.5 years. In total, 86,297 patients with nMH and 254,274 controls were identified from the same database. Additionally, 56,760 (58.8%) patients with nMH and 124,936 (49.1%) controls were women; their mean ages were 49.2 ± 14.6 and 45.7 ± 15.0 years, respectively. After PSM, all absolute standardized differences between migraine and control groups were < 0.1 (Table 1). Similarly, no significant differences were observed for any variable between the nMH and control groups ( Table 2). In a comparative analysis of migraine and nMH groups, no significant differences were observed for any variable after PSM (Table 3).    (Figure 2). Figure 3 shows the ORs for respiratory and GI disorders in nMH patients versus controls. The fully adjusted ORs were 1.16 (95% CI, 1.09-1.24) for asthma, 1.33 (95% CI, 1.24-1.43) for bronchitis, 1.17 (95% CI, 1.13-1.21) for GERD, and 1.38 (95% CI, 1.33-1.42) for gastritis (p < 0.001) (Figure 3). Additionally, we have added a table presenting the prevalence of the outcome variables among the patients with migraine, patients with nMH, and controls (Table A1) (Figure 3). Additionally, we have added a table presenting the prevalence of the outcome variables among the patients with migraine, patients with nMH, and controls (Table A1).

Discussion
In this study, we compared clinical data regarding respiratory and GI dis among patients with migraine, patients with nMH, and controls. After adjusting variates and PSM, the ORs for asthma and all GI disorders, including GERD, g FGID, and IBS, were significantly higher in patients with migraine than in contro ORs for asthma and bronchitis were significantly higher in patients with nMH controls. However, when comparing the migraine and nMH groups, there were no tically significant differences in respiratory disorders, but the ORs for all GI dis including GERD, gastritis, FGID, and IBS, were significant. Our findings suggest t graine and nMH patients have an increased risk of GI and respiratory disorders.

Discussion
In this study, we compared clinical data regarding respiratory and GI disorders among patients with migraine, patients with nMH, and controls. After adjusting for covariates and PSM, the ORs for asthma and all GI disorders, including GERD, gastritis, FGID, and IBS, were significantly higher in patients with migraine than in controls. The ORs for asthma and bronchitis were significantly higher in patients with nMH than in controls. However, when comparing the migraine and nMH groups, there were no statistically significant differences in respiratory disorders, but the ORs for all GI disorders, including GERD, gastritis, FGID, and IBS, were significant. Our findings suggest that migraine and nMH patients have an increased risk of GI and respiratory disorders.
Previous epidemiological studies revealed a high prevalence of asthma in patients with migraine [21,22]; migraine is also significantly more common in patients with asthma, particularly adolescents [23][24][25]. Migraine and asthma exhibit a bidirectional association, such that each condition is reciprocally associated with the other [26,27]. Additionally, asthma increases the risk of chronic migraine 1 year later among individuals with episodic migraine, potentially influencing the clinical course of migraine [28]. Migraine headaches may also be associated with poor asthma control [29]. Migraine and asthma are chronic disorders characterized by recurrent episodic attacks; they share underlying pathophysiological mechanisms involving inflammation and neurological processes. Both conditions are influenced by hormonal imbalances and environmental triggering factors [30][31][32]. However, only a few studies have investigated associations between migraine and respiratory disorders other than asthma or between respiratory disorders and primary headaches other than migraine.
In a comparative study of chronic and episodic migraine groups based on a large population sample, respiratory disorders (e.g., asthma, bronchitis, and COPD) were significantly more common in individuals with chronic migraine [16]. Among 288 outpatients with asthma diagnosed through early or late reversibility tests, 60.4% reported headaches, and 32.6% met the International Headache Society criteria for migraine [24]. Our study revealed an association between migraine and asthma that was similar to the findings in previous studies. Additionally, the ORs for asthma and bronchitis were significantly higher in the nMH group than in the control group. The nMH group in our study included patients with TTH or non-specific headache codes; it excluded patients with all types of migraines. Thus, our findings suggest an association between primary headaches and respiratory disorders.
A previous large, population-based, cross-sectional study showed that migraine and nMH were 1.5-fold more likely among patients with asthma, asthma-related symptoms, hay fever, and chronic bronchitis than among patients without these conditions [33]. Previous studies demonstrated a plausible relationship between primary headaches and immunological or autoimmune disorders [34,35]; immunological events are potentially involved in the pathophysiology of primary headaches. The prevalence of headaches is higher in patients with immunological or autoimmune disorders than in the general population, suggesting that headaches constitute a risk factor or clinical manifestation of these conditions [35]. Cytokine profiles of cluster headaches and TTH patients suggest immunological dysregulation in the pathophysiology of primary headaches, but the corresponding evidence is weaker than the evidence for migraine [34]. Although the presence of a causal relationship is uncertain, our results suggest that comorbid conditions (e.g., allergies and respiratory disease) should be considered in patients with frequent primary headaches.
Our results showed that the ORs for respiratory and GI disorders (e.g., asthma, GERD, gastritis, FGID, and IBS) were significantly higher in patients with migraine than in controls. Compared with the control group, the nMH group also had significantly greater risks of respiratory and GI disorders, although the ORs for these conditions, including asthma, GERD and gastritis, were lower in the nMH group than in the migraine group. However, only the ORs for GI disorders were statistically significant in the comparison between the migraine and nMH groups. The brain has many connections with various organs. Headaches, particularly migraine, are associated with GI disorders and are often accompanied by various GI symptoms [1][2][3]5,[36][37][38][39]. While nearly all previous observational studies focused on migraine, some studies investigated the relationship between GI symptoms and headache, including migraine. There was a statistically significant association between TTH and the prevalence of GI disorders [37,38]. All GI complaints were as common among subjects with nMH as in migraineurs, and the strength of the association between headache and GI complaints increased with increasing headache frequency [39]. Additionally, some studies reported an association between primary headache and GI disorders and demonstrated improvement of headache following treatment of the accompanying GI disorder [40][41][42][43][44]. Current evidence shows that the gut-brain axis may affect migraine, although the mechanism explaining this interaction is not entirely clear. Emerging evidence suggests that the gut microbiota is crucial in the pathogenesis of migraine [2]. Studies using animal models and human studies have demonstrated that the gut microbiome is altered in migraine sufferers compared to healthy controls [45]. Recent evidence suggests that the gut microbiota may also play a critical role in many other types of chronic pain, including headache [46]. Therefore, further epidemiological, clinical, and pathophysiological evidence is needed to identify associations between primary headaches other than migraine and the gut microbiome. In addition to the gut microbiome, the lung microbiome regulates the immune reactivity of the brain [6]. Experimental reports have described immunological findings in patients with migraine and other primary headaches [34]. Thus, more research is needed to determine the association between primary headache and the lung microbiome.
This study had some limitations. First, it was a retrospective study that used clinical data from individuals who visited a university medical center with five hospitals. Thus, it is difficult to generalize the results to the general population; there was also potential for selection bias and unmeasured confounding variables that were not considered. Second, patients were included according to diagnosis codes in the CDW, but detailed clinical information regarding primary headache or respiratory disease characteristics was not collected. Additionally, various psychiatric comorbidities are common in patients with primary headache. Of the psychiatric comorbidities, only depression and anxiety disorders, which are the most prevalent in headache patients, were selected and analyzed as covariates in this study. However, other psychiatric comorbidities should also be considered. To increase the accuracy of patient inclusion, data were collected for patients who had a history of at least two outpatient treatments or at least one hospitalization for each diagnostic code of migraine, GI, and respiratory disorders. Future prospective, population-based studies are needed to investigate the association between primary headaches and respiratory and GI disorders.

Conclusions
In addition to GI disorders, our findings indicate that migraine and nMH are associated with increased risks of respiratory disorders, such as asthma and bronchitis. However, comparison of the migraine and nMH groups revealed that there were no statistically significant differences in respiratory disorders, but significant differences were shown in all GI disorders. Our findings suggest that migraine and nMH are associated with increased risks of GI and respiratory disorders. Future prospective studies are needed to provide further evidence and fully investigate the associations between primary headache disorders and respiratory and GI disorders.