A Comprehensive Assessment of Vascular and Nonvascular Risk Factors Associated with Migraine

Introduction: Migraine is a chronic disabling neurological disease, with an estimated expense of $15-20 million/year. Several studies with a small number of patients have studied risk factors for migraine such as cardiovascular disorders, stroke, smoking, demographic, and genetic factors but this is the first comprehensive study for evaluation of vascular and nonvascular risk factors. It is important to evaluate all the risk factors that help to prevent the healthcare burden related to migraine. Methodology: We performed a retrospective cross-sectional analysis of the Nationwide Inpatient Sample (NIS) (years 2013-2014) in adult (>18-years old) hospitalizations in the United States. Migraine patients were identified using ICD-9-CM code to determine the demographic characteristics, vascular, and nonvascular risk factors. Univariate analysis was performed using the chi-square test and a multivariate survey logistic regression analysis was performed to identify the prevalence of the risk factors and evaluate the odds of prevalence of risk factors amongst migraine patients compared to nonmigraine patients, respectively. Results: On weighted analysis, after removing missing data of age, gender and race, from years 2013 to 2014, of the total 983,065 (1.74%) migraine patients were identified. We found that younger (median age 48-years vs. 60-years), female (82.1% vs. 58.5%; p<0.0001), white population (76.8% vs. 70.5%; p<0.0001), and privately insured (41.1% vs. 27.4%; p<0.0001) patients were more likely to have migraine than others. Cerebral atherosclerosis, diabetes mellitus, ischemic heart disease, atrial fibrillation, and alcohol abuse were not significantly associated with migraine. Migraineurs had higher odds of having hypertension [odds ratio (OR): 1.44; 95% confidence interval (CI): 1.43-1.46; 44.49% vs. 52.84%], recent transient ischemic attack (TIA) (OR: 3.13; 95%CI: 3.02-3.25; 1.74% vs. 0.67%), ischemic stroke (OR: 1.40; 95%CI: 1.35-1.45; 2.06% vs. 1.97%), hemorrhagic stroke (OR: 1.11; 95%CI: 1.04-1.19; 0.49% vs. 0.46%), obesity (OR: 1.46; 95%CI: 1.44-1.48; 19.20% vs. 13.56%), hypercholesterolemia (OR: 1.33; 95%CI: 1.30-1.36; 5.75% vs. 5.54%), substance abuse (OR: 1.51; 95%CI: 1.48-1.54; 7.88% vs. 4.88%), past or current consumption of tobacco (OR: 1.40; 95%CI: 1.38-1.41; 31.02% vs. 27.39%), AIDS (OR: 1.13; 95%CI: 1.04-1.24; 0.33% vs. 0.41%), hypocalcemia (OR: 1.09; 95%CI: 1.03-1.14; 0.77% vs. 0.89%), and vitamin D deficiency (OR: 1.93; 95%CI: 1.88-1.99; 2.47% vs. 1.37%) than patients without migraine. Female patients were at a higher risk of migraine (OR: 3.02; 95%CI: 2.98-3.05) than male. Conclusion: In this study, we have identified significant risk factors for migraine hospitalizations. Early identification of these risk factors may improve the risk stratification in migraine patients.


Introduction
Migraine is a primary headache disorder that impacts about one-fifth of the general population of the United States, and females are three to four times more frequently affected than males [1][2][3]. Estimates of the incidence of chronic migraine worldwide vary from 0% to 5.1%, with most general population surveys reporting a prevalence of 1.4%-2.2% [4]. Migraine impacts 12% of adults in the United States, making this disease a predominant concern for many patients and their physicians [5]. Thus, the disability caused by diseases such as migraine has become a major cause of the global burden of public health and socioeconomic expenses.
Migraine often affects people during their most productive years of life, exerts substantive individual and societal costs, causing a considerable degree of disability and is associated with several risk factors and comorbidities [6]. The largest incidence is among females between the ages of 18 and 49 [4]. Treatment includes avoidance of migraine triggers and risk factor modification, as well as drug and nondrug-based treatments to prevent and abort migraine attacks. The recognition and management of disorders and risk factors that are comorbid with chronic migraine optimizes patient outcomes [7].
It is important to identify the full spectrum of risk factors to help prevent migraine hospitalizations. Several studies have looked at risk factors such as cardiovascular disorders, stroke, fibromyalgia, smoking, vitamin D deficiency, demographic and genetic factors [7][8][9][10][11], but a comprehensive study like this has not been done before. Due to the high prevalence of migraine, any association between migraine and vascular and nonvascular disease could have a substantial effect on public health.

Materials And Methods
Data were obtained from the Agency for Healthcare Research and Quality's Healthcare Cost and Utilization Project (HCUP) Nationwide Inpatient Sample (NIS) between January 2013 and December 2014. The NIS is the largest publicly available all-payer inpatient care database in the United States and contains discharge-level data provided by states that participate in the HCUP (including a total of 46 in 2011). This administrative dataset contains data on approximately eight million hospitalizations in 1,000 hospitals that were chosen to approximate a 20% stratified sample of all US community hospitals, representing more than 95% of the national population. Discharge weights are provided for each patient discharge record, which allows extrapolation to obtain national estimates. Each hospitalization is treated as an individual entry in the database and is coded with one principal diagnosis, up to 24 secondary diagnoses, and 15 procedural diagnoses associated with that stay. Detailed information on NIS is available at http://www.hcup-us.ahrq.gov/db/nation/nis/nisdde.jsp. The NIS is a de-identified database, so informed consent or Institutional Review Board (IRB) approval was not needed for the study. The HCUP Data Use Agreement (HCUP-348L73IZS) for the data utilized in this study was obtained.

Study population
We used the ninth revision of the International Classification of Diseases, clinical modification code (ICD-9-CM) to identify adult patients admitted to hospital with a primary or secondary diagnosis of migraine (ICD-9-CM code: 346) from all adult hospitalizations of the year 2013-2014. Similarly, patients with concurrent vascular and nonvascular risk factors were identified as secondary diagnosis using ICD-9-CM code. Age <18 years and admissions with missing data for age, gender, and race were excluded. The sample size was based on the available data. Data from NIS have been previously used to identify and analyze the trend, outcomes, healthcare cost, and disparity of care [7][8].

Patient and hospital characteristics
Patient characteristics of interest were age, gender, race, insurance status, and concomitant diagnoses as defined above. The race was defined by White (referent), African American, Hispanic, Asian or Pacific Islander, and Native American. Insurance status was defined by Medicare (referent), Medicaid, Private Insurance, and Other/Self-pay/No charge. We defined the severity of comorbid conditions using Deyo's modification of the Charlson comorbidity index ( Table 1).

Outcomes
The primary outcome of interest was to assess the vascular and nonvascular risk factors associated with migraine amongst patients admitted between January 2013 and December 2014 US hospitalizations.

Statistical analysis
All statistical analyses were performed using the weighted survey methods in SAS (version 9.4). Weighted values of patient-level observations were generated to produce a nationally representative estimate of the entire US population of hospitalized patients. A p-value of <0.05 was considered significant. Univariate analysis of differences between categorical variables was tested using the chi-square test and analysis of differences between the continuous variable (age) was tested using paired Student's t-test. Mixed-effects survey logistic regression models with weighted analysis were used for categorical dependent variables in order to estimate odds ratio (OR) and 95% confidence interval (CI) for prediction of vascular and nonvascular risk factors associated with migraine in the year 2013-2014. We included demographics (age, gender, race), patient-level hospitalization variables (admission day, primary payer, admission type, median household income category), hospital-level variables (hospital region, teaching versus nonteaching hospital, hospital bed size), and Charlson's comorbidity index (CCI).
For each model, c-index (a measure of goodness of fit for binary outcomes in a logistic regression model) was calculated. All statistical tests used were two-sided, and p<0.05 was deemed statistically significant. No statistical power calculation was conducted prior to the study.

Disease hospitalizations
From the year 2013 to 2014, after excluding patients with aged <18 years and admissions with missing data for age, gender, and race, we found a total of 56,499,788 nationwide inpatient adult hospitalizations. Out of them, a total of 983,065 (1.74%) patients had migraine.    The percentages here are column % that indicates a direct comparison of risk factors between migraineurs and nonmigraineurs amongst nationwide inpatient hospitalizations.

Regression model derivation
We performed the multivariable survey logistic regression model to predict the vascular and nonvascular risk factors associated with migraine after adjusting for basic demographic characteristics with patient and hospital-level variables, and CCI (

Accuracy of the model
Adjusted model to predict the vascular and nonvascular risk factors associated with migraine had the c-statistic of 0.729, which is >0.7 indicating a very good model.

Discussion
Migraine is a highly burdensome disease and has great deleterious effect on patient's life, their families, and society. Migraine is the leading cause of outpatient and ED visits and is an important public health problem [3]. Americans annually (14.9% on average). This prevalence is close to the 11.7% migraine prevalence observed in the American Migraine Prevalence and Prevention (AMPP) study [5].
We also found higher prevalence of migraine in females (82.11% vs. 58.52%; p<0.0001) and White population (76.84% vs. 70.51%; p<0.0001). In support of our findings, four large US population-based migraine studies including AMS I, AMS II, AMPP, and CaMEO study have also reported a higher prevalence of migraine in females and whites [5,10,[13][14].
This study found various vascular and nonvascular risk factors associated with migraine. In our study results, migraineurs had 1.46 times higher odds of obesity compared to nonmigraineurs (p<0.0001). A meta-analysis study by Ornello et al. also reported that obese subjects had an increased risk of having chronic migraine [pooled adjusted effect estimate (PAEE): 1.75; p<0.001] [15]. Similar results are reported by Fava et al. and explained that insulin resistance associated with obesity might increase the risk of chronic migraine [16]. However, the mechanism linking obesity with increased migraine attacks is not clearly understood.
We also found that migraineurs have higher odds of having hypertension (OR: 1.44; p<0.0001). The Gardener et al. study also reported that hypertension is associated with migraine (OR: 1.76, p<0.0001) [17]. Common environmental and lifestyle factors that may be associated with both migraine and hypertension include diet (salty foods), physical inactivity, chronic stress; both are linked to the metabolic syndrome [18]. Stress-related occupation has been linked to migraine, high blood pressure, and more recently ischemic stroke [19].
Important finding of our study is that migraineurs had a higher risk of cerebrovascular disease (CVD) including transient ischemic attack (OR: 3.13; p<0.0001), acute ischemic stroke (OR: 1.40; p<0.0001), and hemorrhagic stroke (OR: 1.14; p=0.0017). Our study results were supported by a 2017 meta-analysis of 11 prospective studies and 2,221,888 patients; they found a relative risk (RR) of 1.64 for ischemic stroke and RR of 1.15 for hemorrhagic stroke among migraineurs compared to nonmigraineurs [20]. The pathogenic mechanisms explaining the association between migraine and stroke are not clearly understood, but a few hypotheses exist. Some suggested hypothetical theories to explain the migraine stroke relation are cortical spreading depression [21], vasoconstriction, endovascular dysfunction [22], increased prevalence of vascular risk factors, shared genetic defects [23], neurogenic inflammation [24], and hypercoagulability [22,25].
We found that migraineurs are at 1.40 times higher risk of smoking (current/past). Some studies have indicated the positive correlation between cigarette smoking and migraine [9]. Smoking among migraineurs may be related to the development of cranial autonomic symptoms, but clear evidence showing tobacco use increasing the migraine attack is lacking [11].
We found, another major risk associated with a migraine that is, hypocalcemia (OR: 1.09; p=0.00250), and vitamin D deficiency (OR: 1.93; p<0.0001). A study by Patel et al. similarly reported that the odds of having migraine attacks were higher among vitamin D deficiency (OR: 1.97%; p<0.0001) and hypocalcemia (OR: 1.11; p=0.0061) patients [8]. The possible explanation of this association between migraine and Vitamin D deficiency and hypocalcemia can be neurogenic inflammation involved with the migraine cascade of events [26,27] and the role of vitamin D in increasing the inflammation and the observed phenomenon of neuronal hyperexcitability in hypocalcemia leading to increased severity of the migraine attacks [28].

Strength and limitations
This is the largest population-based study which comprehensively evaluated the vascular and nonvascular risk factors associated with migraine. Despite being a large study with good statistical power, this study has limitations. Data from clinical registries or administrative databases are obtained retrospectively by chart abstractions based on the discharge diagnosis codes, billing codes, etc., and hence susceptible to coding errors. We evaluated the risk factors in migraine patients who were admitted primarily due to migraine-related hospitalization or secondary to other conditions, so these risk factors may or may not be responsible for increasing severity of migraine or status migrainosus.

Conclusions
In this study, we identified risk factors for migraine hospitalizations. Identification and management of these risk factors are very important to prevent migraine hospitalizations and to reduce the enormous healthcare burden resulting from it.