MEDICINE

Fibromyalgia and Myositis Linked to Higher Burden and Disabilityin Patients with Migraine

Urvish K. Patel1 & Preeti Malik2 & Richa Sheth3& Princy Malhi4 & Amita Kapoor4 & Bakhtiar M. Rasul2 & Saleha Saiyed2

&

Tapan Kavi5 & Ashish Kapoor6

Accepted: 6 August 2019# Springer Nature Switzerland AG 2019

AbstractThe increase in migraine frequency—with shift towards chronicity—promotes an enhancement of the central hyper-sensitivity which has been linked with fibromyalgia and myositis (FM). Migraine patients with FM have been knownto have higher burden of headaches across studies, but this is the first population-based study to evaluate disabilityand morbidity in migraine patients with concurrent FM. We performed a retrospective cross-sectional analysis ofmigraine hospitalizations using the nationwide database to determine cost, hospital stay, disability, and dischargedisposition. This analysis was performed in migraine patients with and without FM using ICD-9-CM codes. Weperformed weighted analyses using chi-square and t test. Among year 2014 hospitalizations, we identified FMpatients and regression analysis was performed to evaluate whether migraine or other headache disorders werepredictors of FM hospitalization. Between years 2003 and 2014, of the total 446,446 migraine hospitalizations,22,735(5.09%) patients had concurrent FM. Migraine patients with FM had higher prevalence of loss of function (8.4% vs.6.5%, p<0.0001) and transfer to rehabilitation facilities (5.5% vs. 4.5%, p<0.0001) compared to those without FM. Migrainepatients with FM also had higher hospitalization stay and cost. Through regression analysis, we found that migraine (aOR, 3.03;p<0.0001), cluster headache (aOR, 1.71; p=0.0124), and tension headache (aOR, 1.87; p<0.0001) were highly associated withFM hospitalization. FM was associated with significant increase in disability, morbidity, hospitalization stay, and cost in patientsadmitted with migraine. On the basis of this study finding, it would be reasonable to screen migraine patients with depression,anxiety, or other psychiatric disorders for symptoms of FM to mitigate the burden.

Keywords Migraine . Fibromyalgia . Myositis . Disability . Cluster headache . Tension headache . Nationwide inpatient sample

Tapan Kavi and Ashish Kapoor contributed equally to this work.

This article is part of the Topical Collection on Medicine

Electronic supplementary material The online version of this article(https://doi.org/10.1007/s42399-019-00129-7) contains supplementarymaterial, which is available to authorized users.

* Urvish K. Pateldr.urvish.patel@gmail.com

1 Department of Neurology & Public Health, Icahn School ofMedicine at Mount Sinai, 1 Gustave L. Levy Pl, NewYork, NY 10029, USA

2 Department of Public Health, Icahn School of Medicine at MountSinai, New York, NY, USA

3 Department of Internal Medicine, Albert Einstein College ofMedicine, New York, NY, USA

4 Department of Internal Medicine, CarePoint Health, BayonneMedical Center, Bayonne, NJ, USA

5 Department of Neurology, Cooper Medical School of RowanUniversity, Camden, NJ, USA

6 Department of Neurology, Jersey City Medical Center-RWJBarnabas Health & Bayonne, Medical Center-CarePoint Health,Bayonne, NJ, USA

https://doi.org/10.1007/s42399-019-00129-7SN Comprehensive Clinical Medicine (2019) 1:882–890

/Published online: 2 September 2019

Introduction

Migraines are a type of headache disorder, affecting 14% ofthe population, associated with debilitating pain, recurrent ep-isodes, nausea, and sensitivity to light and sound [1]. FM ischronic musculoskeletal disorders characterized by wide-spread pain, tenderness, and fatigue, as well as cognitive dys-function. The reported prevalence of fibromyalgia is 2.0 to5.7% [1]. While the cause of FM remains unknown, it ishypothesized that central sensitization causes the persistenceof musculoskeletal pain. Central sensitization involves a nox-ious stimulus that leads to permanent damage to nociceptivepathways [2]. Pain persistence in migraine may be due tocentral sensitization, indicating a possible cause of comorbid-ity between fibromyalgia and migraine [2]. In a study byCentonze et al., it is suggested that fibromyalgia and episodicmigraine may be a continuum of the same disease [3]. FM hasbeen shown to be associated with more severe headaches andpoorer quality of life in migraine patients [4–6]. In this retro-spective cross-sectional study, we hypothesized that concur-rent FM among migraineurs is associated with higher disabil-ity and morbidity.

Primary objective of this study was to investigate if FMpredicts disability and morbidity in the subset of large inpa-tient sample with 446,446 migraine hospitalizations.Secondary objective was to study whether migraine or otherheadache disorders predict FM hospitalizations.

Methods

Data was obtained from the Agency for HealthcareResearch and Quality’s Healthcare Cost and UtilizationProject (HCUP) NIS files between January 2003 andDecember 2014. The NIS is the largest publicly availableall-payer inpatient care database in the USA and containsdischarge-level data provided by states that participate inthe HCUP (including a total of 46 in 2011). This admin-istrative dataset contains data on approximately 8 millionhospitalizations in 1000 hospitals that were chosen to ap-proximate a 20% stratified sample of all US communityhospitals, representing more than 95% of the nationalpopulation. Criteria used for stratified sampling of hospi-tals into the NIS include hospital ownership, patient vol-ume, teaching status, urban or rural location, and geo-graphic region. Discharge weights are provided for eachpatient discharge record, which allow extrapolation to ob-tain national estimates. Each hospitalization is treated asan individual entry in the database and is coded with oneprincipal diagnosis, up to 24 secondary diagnoses, and 15procedural diagnoses associated with that stay. Detailedinformation on NIS is available at http://www.hcup-us.ahrq.gov/db/nation/nis/nisdde.jsp

Study Population

We used the 9th revision of the International Classification ofDiseases, clinical modification code (ICD-9-CM) to identifyadult patients admitted to hospital with a primary diagnosis ofmigraine (ICD-9-CM code 346). Similarly, patients with fi-bromyalgia and myositis were identified as secondary diagno-sis associated with migraine using ICD-9-CM code 729.1.Age <18 years and admissions with missing data for age,gender, and race were excluded. The sample size was basedon the available data. We used ICD-9-CM codes to identifypatients with Cluster Headache 339.0 and Tension TypeHeadache 339.1 or 307.81.

Patient and Hospital Characteristics

Patient characteristics of interest were age, gender, race, insur-ance status, and concomitant diagnoses as defined above.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’smodification of the Charlson’s Comorbidity Index(Supplementary Table 1). Facilities were considered to beteaching hospitals if they have an American MedicalAssociation-approved residency program, are a member ofthe Council of Teaching Hospitals, or have a full-time equiv-alent interns and residents to patient’s ratio of ≥0.25. HCUPNIS contains data on total charges for each hospital in thedatabases, which represents the amount that hospitals billedfor services.

Outcomes

We tried to find out disability, morbidity, length of stay(LoS), and cost of hospitalization associated with FMamong migraine hospitalizations (years 2003–2014).The comparison of disability/loss of function was inves-tigated by All Patient Refined Diagnosis Related Groups(APR-DRGs) severity between patients with FM andpatients without FM. APR-DRGs were assigned usingsoftware developed by 3M Health InformationSystems, where score 1 indicates minor loss of function,2 moderate, 3 major, and 4 extreme loss of function.Morbidity was defined as length of stay ≥7 days (≥95percentile or +1.5 SD) and discharge other than home(short-term hospital, skilled nursing facility, intermediatecare facility).

Our secondary outcome of interest was to evaluatewhether headache disorders worsen FM which leads toFM-related hospitalization among January 2014–December 2014 US hospitalizations. The reason to

SN Compr. Clin. Med. (2019) 1:882–890 883

choose only year 2014 data for secondary outcome wasthe large number of US hospitalizations (more than 20million) each year to evaluate patients with and withoutFM and headache disorders.

Statistical Analysis

All statistical analyses were performed using the weight-ed survey methods in SAS (version 9.4). Weightedvalues of patient-level observations were generated toproduce a nationally representative estimate of the entireUS population of hospitalized patients. A p value of<0.05 was considered significant. Univariate analysisof differences between categorical variables was testedusing the chi-square test and analysis of differences be-tween continuous variables (length of stay and cost ofhospitalization) was tested using paired Student’s t test.Mixed-effects survey logistic regression models withweighted analysis were used for the categorical depen-dent variables, including migraine and outcomes of in-terest, in order to estimate odds ratio (OR) and 95%confidence interval for the association between headachedisorders (migraine, cluster headache, tension headache)and migraine in year 2014 cohort.

We included demographics (age, gender, race),patient-level hospitalization variables (admission day,primary payer, admission type, Median HouseholdIncome Category), hospital-level variables (hospital re-gion, teaching versus non-teaching hospital, hospital bedsize), comorbidities like hypertension, diabetes mellitus,hypercholesterolemia, obesity, smoking status, drugabuse, alcohol abuse, depression, generalized anxietydisorders, other psychiatric disorders, and Charlson’sComorbidity Index (CCI).

For each model, C-index (a measure of goodness of fit forbinary outcomes in a logistic regression model) was calculated.All statistical tests used were two-sided, and p<0.05 was deemedstatistically significant. No statistical power calculation was con-ducted prior to the study.

Results

Disease Hospitalizations

We found a total of 596,231 hospitalizations (446,446: weighted-after removing missing data for age, gender and ethnicity) due tomigraine from year 2003 to 2014 after excluding patients with age<18 years and admissions with missing data for age, gender, andethnicity (Fig. 1). Out of 446,446 migraine hospitalizations,22,735 (5.09%) had FM and 423,711 (94.91%) had no FM.

Prevalence Trends

We analyzed trends of FM in migraine hospitalizations. Asshown in Fig. 2, trends of FM were increasing from 2003 to2014. (FM, 3.02% in 2003 to 6.91% in 2014; P-Trend<0.0001).

Demographics, Patient and Hospital Characteristics,and Comorbidities

FM was more in common in 35–65 years of age group.Migraine hospitalizations with FM were more likely to be infemale (96.27% vs. 79.68%, p<0.0001), white (82.02% vs.73.23%, p< 0.0001), had private insurance (49.47% vs.54.76%, p<0.0001), and elective admissions (12.45% vs.9.71%, p<0.0001) than those with absent FM. Comorbiditieslike depression, generalized anxiety disorder, other psychiatricdisorders and neurological disorders, arthritis, diabetes, hyper-tension, obesity, drug abuse, and current or past tobacco con-sumption were higher among FM patients than those withoutFM. Overall, hospitalizations with FM also have a high per-centage of CCI (Deyo’s Charlson’s Comorbidity) Index(Table 1).

The Outcomes

Table 2 has mentioned outcomes of FM among migrainehospitalizations. Outcomes were disability/loss of func-tion, morbidity (length of stay ≥7 days [≥ 95 percentile]and discharge other than home), discharge disposition(home vs. non-home [short-term hospital, skilled nursing/intermediate care facility, home health care]), cost of hos-pitalization, and length of stay.

The prevalence of moderate, major, and severe disabilitywas higher among FM patients. An overall prevalence ofmajor/severe loss of function was 8.37% in FM compared topatients without FM (6.47%) among migraineurs (p<0.0001).The morbidity was higher in FM patients (0.97% vs 0.52%,p<0.0001) than patients without FM. A total of 94.5% of FMpatients had been discharged to home compared to 95.54% ofpatients with no FM (p<0.0001). Overall, patients with FMhad higher prevalence of discharge other than home dis-charges (short-term hospital, skilled nursing/intermediate carefacility, home health care) compared to those with no FM(5.50% vs. 4.46%, p<0.0001). Mean length of stay (3.4 daysvs. 2.79 days, p<0.001) and total cost of hospitalization($20,174 vs. $18,092, p<0.001) were higher in FM patients(Table 2).

Regression Model Derivation

Among year 2014 of total 28,212,820 hospitalizations,we had considered 26,614,100 patients after excludingage <18 years and admissions with missing data for

SN Compr. Clin. Med. (2019) 1:882–890884

age, gender, and race. Out of which 411,835 (1.55%)patients had FM. Among this population, we looked

for concurrent headache disorders and other comorbidi-ties which could predict the FM hospitalizations.

Fig. 1 Flowchart detailing cohortselection and analysis modeling

Fig. 2 Yearly prevalence of fibromyalgia and myositis among migraine patients

SN Compr. Clin. Med. (2019) 1:882–890 885

Table 1 Characteristics of fibromyalgia and myositis (FM) patients in migraine hospitalizations

FM Non-FM Total p value

Migraine weighted (%) 22,735 (5.09) 423,711 (94.91) 446,446 (100) <0.0001Demographics of patientsAge group (years) <0.000118–34 3654 (16.07) 113,726 (26.84) 117,379 (26.29)35–49 10,379 (45.65) 173,678 (40.99) 184,057 (41.23)50–64 7365 (32.4) 100,318 (23.68) 107,683 (24.12)65–79 1229 (5.41) 28,787 (6.79) 30,016 (6.72)≥80 108 (0.47) 7202 (1.70) 7310 (1.64)Gender (%) <0.0001Male 849 (3.73) 86,095 (20.32) 86,943 (19.47)Female 21,886 (96.27) 337,612 (79.68) 359,497 (80.53)

Race (%) <0.0001White 18,235 (82.02) 301,256 (73.23) 319,491 (73.69)African American 2290 (10.30) 60,837 (14.79) 63,127 (14.56)Hispanic 1461 (6.57) 42,271 (10.28) 43,732 (10.09)Asian or Pacific Islander 101 (0.45) 4848 (1.48) 4949 (1.14)Native American 145 (0.65) 2145 (0.52) 2290 (0.53)

Characteristics of patientsMedian household income category forpatient’s zip code (%)

<0.0001

0–25th percentile 5368 (24.08) 107,876 (26.05) 113,244 (25.95)26–50th percentile 6161 (27.63) 102,939 (24.86) 109,100 (25)51–75th percentile 5887 (26.4) 105,658 (25.51) 111,545 (25.56)76–100th percentile 4880 (21.89) 97,630 (23.58) 102,510 (23.49)

Primary payer (%) <0.0001Medicare 6318 (27.83) 74,074 (17.51) 80,392 (18.04)Medicaid 3310 (14.58) 63,888 (15.1) 67,198 (15.08)Private insurance 11,229 (49.47) 231,675 (54.76) 242,904 (54.49)Other/self-pay/no charge 1843 (8.12) 53,399 (12.62) 55,242 (12.39)

Admission type (%) <0.0001Non-elective 19,833 (87.55) 381,596 (90.29) 401,429 (90.15)Elective 2820 (12.45) 41,041 (9.71) 43,861 (9.85)

Admission day (%) 0.2202Weekday 18,115 (79.68) 336,177 (79.34) 354,291 (79.36)Weekend 4620 (20.32) 87,534 (20.66) 92,154 (20.64)

Characteristics of hospitalsBed size of hospital (%)* <0.0001Small 2855 (12.65) 46,919 (11.13) 49,775 (11.2)Medium 5739 (25.43) 103,335 (24.51) 109,073 (24.55)Large 13,971 (61.91) 271,425 (64.37) 285,397 (64.24)

Hospital location and teaching status (%) 0.0001Rural 2086 (9.24) 37,213 (8.82) 39,299 (8.85)Urban non-teaching 9577 (42.44) 174,699 (41.43) 184,275 (41.48)Urban teaching 10,903 (48.32) 209,768 (49.75) 220,670 (49.67)

Hospital region (%) <0.0001Northeast 3874 (17.04) 96,850 (22.86) 100,724 (22.56)Midwest 5747 (25.28) 79,810 (18.84) 85,557 (19.16)South 9423 (41.45) 184,610 (43.57) 194,032 (43.46)West 3691 (16.23) 62,442 (14.74) 66,132 (14.81)

Comorbidities of patients (%)Arthritis 3085 (13.64) 12,422 (2.95) 15,507 (3.5) <0.0001Depression 7960 (35.19) 81,744 (19.42) 89,704 (20.23) <0.0001Generalized anxiety disorder 380 (1.67) 3393 (0.80) 3773 (0.85) <0.0001Psychiatric disorder 3290 (14.54) 30,019 (7.13) 33,309 (7.51) <0.0001Other neurological disorder ≤10 (0.04) 143 (0.03) 153 (0.03) 0.4746Diabetes 3192 (14.04) 54,125 (12.77) 57,317 (12.84) <0.0001Hypertension 8758 (38.52) 153,821 (36.30) 162,579 (36.42) <0.0001Obesity 3424 (15.06) 45,480 (10.73) 48,904 (10.95) <0.0001Hypercholesterolemia 1127 (4.95) 21,236 (5.01) 22,363 (5.01) 0.6997Drug abuse 1408 (6.19) 17,240 (4.07) 18,648 (4.18) <0.0001Alcohol abuse 136 (0.6) 5520 (1.3) 5656 (1.27) <0.0001Current or past smoker 4972 (21.87) 86,819 (20.49) 91,791 (20.56) <0.0001Acquired immunodeficiencysyndrome (AIDS)

49 (0.21) 1703 (0.4) 1752 (0.39) <0.0001

Deyo’s Charlson’s ComorbidityIndex (CCI)

<0.0001

0 12,517 (55.06) 280,395 (66.18) 292,912 (65.61)1 6630 (29.16) 92,655 (21.87) 99,285 (22.24)2 2252 (9.9) 31,895 (7.53) 34,147 (7.65)3 852 (3.75) 10,536 (2.49) 11,387 (2.55)4 301 (1.32) 3611 (0.85) 3912 (0.88)≥5 183 (0.81) 4619 (1.09) 4801 (1.08)

Percentage in brackets are column % indicates direct comparison between FM vs. non-FM among migraineurs

*Bed size of hospital indicates number of hospital beds which varies depending on hospital location (rural/urban), teaching status (teaching/non-teaching) and region (northeast/midwest/southern/western)

SN Compr. Clin. Med. (2019) 1:882–890886

In the multivariate regression analysis, after adjustingfor basic demographic with patient and hospital-levelvariables, comorbidities, concurrent conditions, andCCI, patients with migraine (adjusted OR, 3.03; 95%CI, 2.95–3.12; p<0.0001), cluster headache (adjustedOR, 1.71; 95% CI, 1.12–2.59; p=0.0124), and tensionheadache (adjusted OR, 1.87; 95% CI, 1.21–2.89;p<0.0001) were at higher risk of having hospitalizationsdue to FM than non-headache disorders (Table 3).

Table 3 also lists multivariate analysis of other predictors ofFM. Comorbidities like arthritis (adjusted OR, 4.72; 95% CI,4.61–4.82; p<0.0001), depression (adjusted OR, 2.31; 95%CI, 2.38–2.35; p<0.0001), generalized anxiety disorder (ad-justed OR, 1.54; 95% CI, 1.46–1.63; p<0.0001), other psychi-atric disorders (adjusted OR, 1.65; 95% CI, 1.60–1.69;p<0.0001), and other neurologic disorders (adjusted OR,1.31; 95% CI, 1.28–1.34; p<0.0001) were significant predic-tors of FM hospitalizations.

Concurrent conditions like drug abuse/dependence (adjust-ed OR, 1.67; 95%CI, 1.62–1.72; p<0.0001), obesity (adjustedOR, 1.54; 95% CI, 1.52–1.57; p<0.0001), hypercholesterol-emia (adjusted OR, 1.12; 95% CI, 1.08–1.15; p<0.0001), ac-quired immunodeficiency syndrome (adjusted OR, 1.41; 95%CI, 1.20–1.65; p<0.0001), current or past smoker (adjustedOR, 1.30; 95% CI, 1.27–1.32; p<0.0001), diabetes (adjustedOR, 1.15; 95% CI, 1.12–1.17; p<0.0001), and hypertension(adjusted OR, 1.23; 95% CI, 1.20–1.25; p<0.0001) were alsosignificantly associated with FM-related hospitalization.

Accuracy of the Model

c-statistic was 0.799 which is used to validate the accuracy ofthe regressions. Adjusted model has c-index >0.7, which in-dicates a good model.

Discussion

The major finding from our study is high prevalence of FM inpatients with migraine hospitalizations. Despite of differentorigin of pain, both FM and migraine are comorbidities dueto central sensitization phenomenon [7]. Response to somatichyperalgesia was enhanced by concurrent fibromyalgia andmigraine than one condition alone. Higher migraine frequencyand/or chronicity provokes hyperalgesia and fibromyalgiapain, which is reversed by effective migraine prophylaxis ac-cording to Giamberardino et al. [8]. Several studies have re-ported high prevalence of fibromyalgia in migraineurs [9–13].A retrospective cohort study, fibromyalgia in migraine, byWhealy et al., found that patients with comorbid FM andmigraine reported higher average headache intensity and se-verity than their age and sex-matched controls, who had mi-graine alone [4]. This reflects the findings of other studies, thatin patients with FM, headaches are more likely to be rated asincapacitating, as compared to patients without FM [5]. Whilethe Marcus et al. study found that there is not a significantdifference in pain and cognitive distress between FM patients

Table 2 Univariate analysis of outcomes fibromyalgia and myositis (FM) among migraine hospitalizations

FM Non-FM Total p value

APR-DRG severity or disability/loss of function (%) <0.0001

Minor loss of function 9133 (40.38) 210,664 (50.07) 219,797 (49.58)

Moderate loss of function 1591 (51.25) 182,861 (43.46) 194,452 (43.86)

Major loss of function 1810 (8) 26,257 (6.24) 28,067 (6.33)

Severe loss of function 84 (0.37) 958 (0.23) 1042 (0.23)

Total major/severe loss of function (%) 1894 (8.37) 27,215 (6.47) 29,109 (6.56)

Morbidity* (%) 222 (0.97) 2199 (0.52) 2420 (0.54) <0.0001

Discharge disposition (%) <0.0001

Routine/home 21,216 (94.5) 399,178 (95.54) 420,395 (95.49)

Transfer to short-term hospital 109 (0.48) 2255 (0.54) 2364 (0.54)

Transfer to SNF/ICF/another type of facility 305 (1.36) 5805 (1.39) 6110 (1.39)

Home health care 821 (3.66) 10,573 (2.53) 11,393 (2.59)

Total discharge other than home (%) 1235 (5.50) 18,633 (4.46) 19,868 (4.52)

Length of stay ± SE (days) 3.4±0.043 2.8±0.008 <0.0001

Cost of hospitalization ± SE ($) 20,174±273.2 18,092±55.9 <0.0001

Percentage in brackets are column % indicates direct comparison between FM vs. non-FM among migraineurs

APR-DRG All Patients Refined Diagnosis-Related Groups, SNF skilled nursing facility, ICF intermediate care facility, SE standard error

*Morbidity: length of stay ≥7 days (≥95 percentile or +1.5 SD) and discharge other than home

SN Compr. Clin. Med. (2019) 1:882–890 887

Table 3 Multivariate logistic regression analysis to predict the fibromyalgia and myositis hospitalization due to headache disorders

Odds ratio (OR) Confidence interval (CI) p value

LL UL

No-Headache ReferenceMigraine 3.03 2.95 3.12 <0.0001Cluster headache 1.71 1.12 2.59 0.0124Tension headache 1.87 1.21 2.89 0.0048

Age (every 10 years) 0.99 0.99 0.99 <0.0001GenderFemale ReferenceMale 0.21 0.20 0.21 <0.0001

RaceWhite ReferenceAfrican American 0.58 0.57 0.60 <0.0001Hispanic 0.63 0.61 0.64 <0.0001Asian or Pacific Islander 0.33 0.30 0.36 <0.0001Native American 0.81 0.74 0.89 <0.0001

Median household income categoryfor patient’s zip code0–25th percentile Reference26-50th percentile 1.01 0.99 1.03 0.447351–75th percentile 0.96 0.94 0.98 <0.000176–100th percentile 0.88 0.86 0.90 <0.0001

Primary payerMedicare ReferenceMedicaid 0.69 0.68 0.71 <0.0001Private insurance 0.83 0.81 0.85 <0.0001Other/self-pay/no charge 0.74 0.71 0.76 <0.0001

Admission typeNon-elective ReferenceElective 0.92 0.91 0.94 <0.0001

Admission dayWeekday ReferenceWeekend 0.94 0.93 0.96 <0.0001

Bed size of hospitalSmall ReferenceMedium 0.97 0.95 0.99 0.0035

Large 0.97 0.95 0.99 0.0007Hospital location and teaching statusRural ReferenceUrban non-teaching 1.06 1.03 1.09 <0.0001Urban teaching 1.07 1.04 1.10 <0.0001

Hospital regionNortheast ReferenceMidwest 1.32 1.29 1.35 <0.0001South 1.21 1.18 1.23 <0.0001West 1.28 1.25 1.32 <0.0001

Comorbidities of patientsArthritis 4.72 4.61 4.82 <0.0001Depression 2.31 2.28 2.35 <0.0001Generalized anxiety disorder 1.54 1.46 1.63 <0.0001Other psychiatric disorder 1.65 1.60 1.69 <0.0001Other neurological disorder 1.31 1.28 1.34 <0.0001Diabetes 1.15 1.12 1.17 <0.0001Hypertension 1.23 1.20 1.25 <0.0001Obesity 1.54 1.52 1.57 <0.0001Hypercholesterolemia 1.12 1.08 1.15 <0.0001Drug abuse 1.67 1.62 1.72 <0.0001Alcohol abuse 0.74 0.71 0.77 <0.0001Current or past smoker 1.30 1.27 1.32 <0.0001Acquired immunodeficiency syndrome (AIDS) 1.41 1.20 1.65 <0.0001

Deyo’s Charlson’s Comorbidity Index (CCI)1 Reference0 1.24 1.21 1.27 <0.00012 1.11 1.08 1.15 <0.00013 1.00 0.96 1.03 0.86414 0.86 0.82 0.90 <0.0001≥5 0.74 0.71 0.78 <0.0001

Area under the ROC curve/c-index 0.799

The model is adjusted for basic demographic with patient-level variables, comorbidities, CCI, concurrent conditions, and hospital-level variables such ashospital region, teaching status, and bed size

UL upper limit, LL lower limit

SN Compr. Clin. Med. (2019) 1:882–890888

with migraine and FM patients without migraine, 76% of theFM patients seeking treatment suffered from chronic head-ache, indicating that headaches should be screened for inFM patients [14].

FM is associated with low quality of life in patients withmigraine [9, 10]. A study by Beyazal et al. found that inmigraine patients, FM comorbidity showed a significant im-pact on the patients’ quality of life. Those with FM had morefrequent migraines, significantly higher mean widespreadpain scores, and lower quality of life scores [6]. Our studycorroborates previous findings by showing high prevalenceof major/severe loss of function in migraine patient with FMas compared to those without FM (p<0.0001). We also foundthat migraineurs with FM had higher rates of discharge dispo-sitions other than home, higher length of stay, and higher costof hospitalizations than migraineurs without FM.

Similar to previous studies we also found a significantlyhigher prevalence of migraine and FM in females as comparedto males [12, 15, 16]. Total 96.27% patients were females formigraine with FM and 79.68% were females for migrainewithout FM. In 2006, Aloisi et al. reported the different ex-pression of pain between sexes is due to interaction betweensex hormones, brain functions, and processing of pain [17].We found increased prevalence of depression, generalizedanxiety disorder, and other psychiatric disorders in migrainepatient compared to those without FM similar to prior studies[9, 10, 12, 13]. Decreased pain habituation is common inmigraine and FM and may lead to central sensitization andmyofascial pain persistence in the presence of other favorablesituations, such as depression, sleep disturbances, and anxiety[13]. Mongini et al. [18] have reported that the presence ofanxiety alone or combined with depression significantly in-crease the level of muscle tenderness in the head and more inthe neck might facilitate into chronic headache forms such asmigraine. Similarly, anxiety may also increase FM and diffusemyofascial pain comorbidity in headache patients who presentwith increased pericranial muscle tenderness. Whealy et al.concluded that concurrent migraine and fibromyalgia patientshave severe depressive symptoms, higher headache intensity,and severe headache-related disability; thus, migraineurshould be evaluated for fibromyalgia especially with depres-sive symptoms, high headache intensity, or high headache-related disability [4]. According to Costantini et al., visceralpain due to comorbid irritable bowel syndrome, primary dys-menorrhea, endometriosis, and colon diverticulosis triggersfibromyalgia pain and hyperalgesia in female patients proba-bly due to enhancing the level of central sensitization whichdecreases significantly after treatment. In such patients, anassessment and treatment of visceral pain comorbiditiesshould be a part of management strategy [19].

Screening for FM in patients with migraine hospitalizationsmay be helpful in identifying this under-recognized comorbid-ity. Treatment of FM with cognitive behavioral therapy,

exercise, and drug therapy may help improve outcomes inmigraine patients as well because of high impact of FM ondisability and quality of life in these patients. On basis ofcentral sensitization phenomenon, Yilmaz et al. had evaluatedthe efficacy of greater occipital nerve blockage in patients withconcurrent migraine and fibromyalgia and found it reducespain severity, headache frequency, duration, and increasesquality of life [20].

A major strength of the study was that findings were na-tionally representative for the USA. NIS data is a largest in-patient database, and our study has good statistical power.APR-DRG coding system used in this study to assess theseverity of illness is external validated. It is a reliable methodwith accurate and consistent results and is widely used byhospitals, consumers, payers, and regulators [21, 22].However, there are limitations to the study. Since this is aninpatient population-based study, there might beunderreporting of concurrent prevalence as all migraine andFM patients are not hospitalized and managed as outpatients.ICD-9-CM code recognizes fibromyalgia and myositis as acombined code so burden of individual diseases cannot beestablished. Data from clinical registries are obtained retro-spectively by chart abstractions based on the discharge diag-nosis codes, billing codes, etc. and hence susceptible to codingerrors. In such cross-sectional study, it is difficult to identifytemporal relationship between FM and headache disorder.

Conclusion

The conclusion derived from the study data is the understand-ing of the significant association of FM with migraine andother headache disorders and its role in increasing burden ofdisability, morbidity, LOS, and cost of hospitalization amongmigraineurs. The evaluation of FM comorbidity in differenttypes of chronic headache subtype patients may increase theknowledge about chronicization mechanism and central sen-sitization phenomenon. Our findings suggest the importanceof screening for FM in migraine and other headache disordersdue to significant impact of FM on quality of life. This willalso help in creating the optimal individual treatment plan. Onthe basis of this study finding, it would also be reasonable toscreen migraine patient with depression, anxiety, or other psy-chiatric disorders for symptoms of FM.

Authors’ Contributions U.P. and P.M. conceive of the idea and performedbiostatistics and analysis. R.S. and P.M. wrote the manuscript with sup-port from A.K., B.R., and S.S. who contributed to the literature review,tables, figure, and citation. A.K. and T.K. supervised the project.

Compliance with Ethical Standards

Conflict of Interest The authors declare that there is no conflict ofinterest.

SN Compr. Clin. Med. (2019) 1:882–890 889

Informed Consent The data has been taken from Nationwide InpatientSample, which is a deidentified database from “Health Care UtilizationProject (HCUP)” sponsored by the Agency for Healthcare Research andQuality, so informed consent or IRB approval was not needed for thestudy. The relevant ethical oversight and HCUP Data Use Agreement(HCUP-4Q28K90CU) were obtained for the study.

References

1. Wolfe F, Ross K, Anderson J, Russell IJ, Hebert L. The prevalenceand characteristics of fibromyalgia in the general population.Arthritis Rheum. 1995;38(1):19–28.

2. de Tommaso M. Prevalence, clinical features and potential thera-pies for fibromyalgia in primary headaches. Expert Rev Neurother.2012;12(3):287–95; quiz 296. https://doi.org/10.1586/ern.11.190.

3. Centonze V, Bassi A, Cassiano MA, Munno I, Dalfino L,Causarano V. Migraine, daily chronic headache and fibromyalgiain the same patient: an evolutive "continuum" of non organicchronic pain? About 100 clinical cases. Neurological sciences :official journal of the Italian Neurological Society and of theItalian Society of Clinical Neurophysiology. 2004;25(Suppl 3):S291–2. https://doi.org/10.1007/s10072-004-0314-4.

4. Whealy M, Nanda S, Vincent A, Mandrekar J, Cutrer FM.Fibromyalgia in migraine: a retrospective cohort study. JHeadache Pain. 2018;19(1):61. https://doi.org/10.1186/s10194-018-0892-9.

5. Peres MF, Young WB, Kaup AO, Zukerman E, Silberstein SD.Fibromyalgia is common in patients with transformed migraine.Neurology. 2001;57(7):1326–8. https://doi.org/10.1212/wnl.57.7.1326.

6. Beyazal MS, Tufekci A, Kirbas S, Topaloglu MS. The impact offibromyalgia on disability, anxiety, depression, sleep disturbance,and quality of life in patients with migraine. Noro Psikiyatri Arsivi.2018;55(2):140–5. https://doi.org/10.5152/npa.2016.12691.

7. FitzcharlesMA, YunusMB. The clinical concept of fibromyalgia asa changing paradigm in the past 20 years. Pain Res Treat.2012;2012:184835–8. https://doi.org/10.1155/2012/184835.

8. Giamberardino MA, Affaitati G, Martelletti P, Tana C, Negro A,Lapenna D, et al. Impact of migraine on fibromyalgia symptoms. JHeadache Pain. 2015;17:28. https://doi.org/10.1186/s10194-016-0619-8.

9. Ifergane G, Buskila D, Simiseshvely N, Zeev K, Cohen H.Prevalence of fibromyalgia syndrome in migraine patients.Cephalalgia : an international journal of headache. 2006;26(4):451–6. https://doi.org/10.1111/j.1468-2982.2005.01060.x.

10. Kucuksen S, Genc E, Yilmaz H, Salli A, Gezer IA, Karahan AY,et al. The prevalence of fibromyalgia and its relation with headachecharacteristics in episodic migraine. Clin Rheumatol. 2013;32(7):983–90. https://doi.org/10.1007/s10067-013-2218-2.

11. Marcus DA, Bhowmick A. Fibromyalgia comorbidity in a commu-nity sample of adults with migraine. Clin Rheumatol. 2013;32(10):1553–6. https://doi.org/10.1007/s10067-013-2310-7.

12. de Tommaso M, Sardaro M, Serpino C, Costantini F, Vecchio E,Prudenzano MP, et al. Fibromyalgia comorbidity in primary head-aches. Cephalalgia : an international journal of headache.2009;29(4):453–64. https://doi.org/10.1111/j.1468-2982.2008.01754.x.

13. de Tommaso M, Federici A, Serpino C, Vecchio E, Franco G,Sardaro M, et al. Clinical features of headache patients with fibro-myalgia comorbidity. J Headache Pain. 2011;12(6):629–38. https://doi.org/10.1007/s10194-011-0377-6.

14. Marcus DA, Bernstein C, Rudy TE. Fibromyalgia and headache: anepidemiological study supporting migraine as part of the fibromy-algia syndrome. Clin Rheumatol. 2005;24(6):595–601. https://doi.org/10.1007/s10067-005-1121-x.

15. Wolfe F. Fibromyalgia: the clinical syndrome. Rheum Dis Clin NAm. 1989;15(1):1–18.

16. MacGregor EA. Oestrogen and attacks of migraine with and with-out aura. Lancet Neurol. 2004;3(6):354–61. https://doi.org/10.1016/s1474-4422(04)00768-9.

17. Aloisi AM, Bonifazi M. Sex hormones, central nervous system andpain. Horm Behav. 2006;50(1):1–7. https://doi.org/10.1016/j.yhbeh.2005.12.002.

18. Mongini F, Deregibus A, Rota E. Psychiatric disorders and muscletenderness in episodic and chronic migraine. Expert Rev Neurother.2005;5(5):635–42. https://doi.org/10.1586/14737175.5.5.635.

19. Costantini R, Affaitati G, Wesselmann U, Czakanski P,Giamberardino MA. Visceral pain as a triggering factor for fibro-myalgia symptoms in comorbid patients. Pain. 2017;158(10):1925–37. https://doi.org/10.1097/j.pain.0000000000000992.

20. Yilmaz V, Aras B, Erturk FA, Cakci FA, Umay E. Migraine inpatients with fibromyalgia and outcomes of greater occipital nerveblockage. Clin Neurol Neurosurg. 2019;181:54–7. https://doi.org/10.1016/j.clineuro.2019.04.004.

21. McCormick PJ, Lin HM, Deiner SG, Levin MA. Validation of theall patient refined diagnosis related group (APR-DRG) risk of mor-tality and severity of illness modifiers as a measure of perioperativerisk. J Med Syst. 2018;42(5):81. https://doi.org/10.1007/s10916-018-0936-3.

22. Baram D, Daroowalla F, Garcia R, Zhang G, Chen JJ, Healy E,et al. Use of the all patient refined-diagnosis related group (APR-DRG) risk of mortality score as a severity adjustor in the medicalICU. Clin Med Circ Respirat Pulm Med. 2008;2:19–25.

Publisher’s Note Springer Nature remains neutral with regard to jurisdic-tional claims in published maps and institutional affiliations.

SN Compr. Clin. Med. (2019) 1:882–890890