European Journal of Internal Medicine 26 (2015) 232–236
Contents lists available at ScienceDirect
European Journal of Internal Medicine journal homepage: www.elsevier.com/locate/ejim
Original Article
Higher risk of developing a subsequent migraine in adults with nonapnea sleep disorders: A nationwide population-based cohort study Tomor Harnod a,b, Yu-Chiao Wang c,d, Chia-Hung Kao e,f,⁎ a
Department of Neurosurgery, Hualien Tzu Chi General Hospital, Buddhist Tzu Chi Medical Foundation, Hualien, Taiwan College of Medicine, Tzu Chi University, Hualien, Taiwan Management Office for Health Data, China Medical University Hospital, Taichung, Taiwan d College of Medicine, China Medical University, Taichung, Taiwan e Graduate Institute of Clinical Medical Science and School of Medicine, College of Medicine, China Medical University, Taichung, Taiwan f Department of Nuclear Medicine and PET Center, China Medical University Hospital, Taichung, Taiwan b c
a r t i c l e
i n f o
Article history: Received 5 February 2015 Received in revised form 25 February 2015 Accepted 2 March 2015 Available online 19 March 2015 Keywords: Cohort study Insomnia Migraine National Health Insurance (NHI)
a b s t r a c t Objective: This nationwide population-based cohort study evaluated the effect of nonapnea sleep disorders (NSDs) on the subsequent development of a migraine. Methods: We identified 46,777 patients aged 20 years and older who were diagnosed with an NSD (ICD-9-CM: 307.4 or 780.5) and without coding for apnea-sleep disorders (ICD-9-CM: 780.51, 780.53, or 780.57) between 2000 and 2002 as the sleep disorder (SD) cohort. A comparison cohort of 93,552 people was enrolled. We calculated the adjusted hazard ratio (aHR) for developing a migraine (ICD-9-CM: 346) after adjusting for age, sex, comorbidity, and drug use. A Kaplan–Meier analysis was used to measure the cumulative incidence of a migraine between 2 curves until the end of 2011. Results: The cumulative incidence of a migraine was significantly higher in the SD cohort. The aHR for developing a migraine in the SD cohort was 3.52 (95% CI = 3.28–3.79). The risk of developing a migraine with an NSD was higher in men (aHR = 4.31) than in women (aHR = 3.35). The age-stratified effect of an NSD on developing a migraine was highest among patients aged 55 years and younger. Higher risks of developing a migraine were observed among the participants without any comorbidity and without any drug treatment for their insomnia. Conclusion: The findings of this population-based cohort study indicate a higher risk of developing a subsequent migraine in patients with an NSD, which could be considered an independent, predisposing factor for developing subsequent a migraine in adulthood. © 2015 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.
1. Introduction Both headaches and insomnia are common complaints of patients that affect the quality of their daily lives. In Asia, migraines were reported as the most prevalent type of headache diagnosed at neurological clinics with a range from 50.9% to 85.8% across various countries [1]. The estimated annual worldwide prevalence of migraines is approximately 15–20% in women and 6–10% in men respectively [2–4]. Migraines have been considered a neurovascular disease for decades and known to have higher associated risks of developing vascular diseases in both sexes [5,6]. Many studies reported the interacting relationships or comorbidities between sleep disorders, anxiety, depression, and migraines [7–9]. In migraineurs, fatigue, sleep deprivation, or simply a change of sleep pattern is frequently reported as the precipitating factor ⁎ Corresponding author at: Graduate Institute of Clinical Medical Science and School of Medicine, College of Medicine, China Medical University, No. 2, Yuh-Der Road, Taichung 404, Taiwan. Tel.: +886 4 22052121x7412; fax: +886 4 22336174. E-mail address: [email protected] (C.-H. Kao).
for headache attacks, with sleep itself being a relieving treatment for the headache [10,11]. Sleep is regulated by the circadian rhythm and is vital to life. Sleep dysfunction can have profound consequences on physiology, behavior, and daily abilities during waking hours [10,12]. Symptoms of dyssomnia or insomnia may be a temporary inconvenience or may chronically affect a person, owing to the many available types of sleep disorder (SDs). Among them, apnea-sleep disorders can result in hypoxic damage to multiple organs and are considered to pose a higher threat to health and for developing subsequent diseases. However, a population-based study analyzed data in 1906 Brazilian children with age of 5–12 years to survey their migraine or tension-type headache using validated questionnaires by interviewing the parents. Interestingly, in the children with migraine, some kinds of parasomnias, such as sleep talking, somnambulism, and bruxism were independently associated with the risk of their migraine headaches [13]. We have sought to determine whether NSDs could be considered independent, predisposing factors for developing a subsequent migraine in adulthood. To answer this question of causeand-effect, we conducted this study to determine the risk of developing
http://dx.doi.org/10.1016/j.ejim.2015.03.002 0953-6205/© 2015 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.
T. Harnod et al. / European Journal of Internal Medicine 26 (2015) 232–236
a migraine in a cohort with an NSD by using the Taiwan nationwide population-based database. 2. Methods and materials 2.1. Study design In this population-based cohort study, we used the Longitudinal Health Insurance Database (LHID) to establish the study cohort. The LHID database is one part of Taiwan's National Health Insurance Research Database (NHIRD), which contains insurance claim data for one million insured people randomly selected from Taiwan's entire 23-million population and covered by the National Health Insurance (NHI) program. The NHIRD is maintained by the National Health Research Institutes. Taiwan's NHI program has been a universal insurance system since 1995 for the residents of Taiwan, with more than 99% of the residents covered by the end of 2000. NHI-participant information that has been recorded in the claims data includes demographic characteristics and details regarding outpatient and inpatient services received from 1996 to 2011. We identified participants with specific diseases by using diagnostic codes from the International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM), which were diagnosed by physicians, either a general physician or a neurologist. The patient identification numbers were scrambled to safeguard patient privacy and confidentiality when linking to the data files. This study was granted ethical approval by the institutional review board of China Medical University (CMU-REC-101-012). 2.2. Selection of study participants Our study population was purposely focused on LHID patients aged 20 years and older with the diagnosis of an NSD (SD cohort). The definition of an NSD was sleep disorders (ICD-9-CM: 307.4 or 780.5) without coding for apnea SD (ICD-9-CM: 780.51, 780.53, or 780.57). We identified a first time diagnosis of an NSD as the index date between the years of 2000 and 2002, and we excluded the ones who had already developed a migraine (ICD-9-CM: 346) before the index date. The comparison cohort was randomly selected from LHID patients without sleep disorders. Each NSD patient was matched to two controls by frequency matching for age, sex, and index year. We analyzed the risk of developing a subsequent migraine between the SD and comparison cohorts. All study participants were followed up from the index date to the occurrence of a migraine, withdrawal from the LHID database, or until December 31, 2011. We also investigated all participants who had ever used zolpidem or benzodiazepines (BZD) before the end date and categorized them into four groups according to treatment for their dyssomnia or insomnia: without drug use, zolpidem use only, BZD use only, or use of both drugs. Several diseases, such as hypertension (ICD-9-CM: 401–405), hyperlipidemia (ICD-9-CM: 272), diabetes (ICD-9-CM: 250), stroke (ICD-9CM: 430–438), chronic obstructive pulmonary disease (COPD; ICD-9CM: 490–496), depression (ICD-9-CM: 2962, 2963, 29682, 3004, 3090, 3091, 30928, and 311), anxiety (ICD-9-CM: 3000, 3002, 3003, 3083, and 30981), were defined as comorbidities if one or more of them were diagnosed before the index date or during the follow-up period. 2.3. Data analysis A chi-square test and Student's t test were applied for a baseline demographic factors, comorbidity and drug-use analysis between the SD and comparison cohorts. The chi-square test was used to analyze the categorical variables, and the Student's t test was used to assess the continuous variables. In both cohorts, we calculated the incidence rate (per 1000 person-years) of a migraine with stratifying by age, sex, comorbidity (no or yes), and drug use (no or yes).
233
The incidence rate ratio (IRR) and the 95% confidence interval (95% CI) of developing a migraine in the SD cohort were calculated by the Poisson regression model and compared with the controls. After we adjusted for confounding risk factors, we determined the adjusted hazard ratio (aHR) and a 95% CI for a migraine by using the multivariable Cox proportional hazards model. We also used the Cox proportional hazards model to estimate the interaction between an NSD and the type of comorbidity. Cumulative incidence analysis was conducted using the Kaplan–Meier method, and the differences between two curves were calculated with the two-tailed log-rank test. All of the analyses were performed using the SAS Version 9.3 statistical package (SAS Institute Inc., NC, USA), and the P value (P b .05) in the two-tailed tests was determined to be statistically significant. 3. Results Table 1 shows the demographic data of the SD and comparison cohorts at baseline; 46,777 patients were recorded as having a first time, primary diagnosis of an NSD between 2000 and 2002, and 93,552 participants were selected as the controls by frequency matching for age, sex, and index year. After the frequency matching, the distributions of age and sex in the SD and comparison cohorts were similar; the proportion of women was 63.3%, and the mean age for the participants was 50.5 years. Comorbidities were more prevalent in the SD cohort than in the comparison cohort at baseline (P b .0001). Instances of drug use were also more common in the SD cohort (97.1%) than in the comparison cohort (75.9%). The cumulative incidence of developing a migraine was significantly higher in patients with an NSD than in the controls during the 12-year follow-up (log-rank test P b .0001; Fig. 1). Table 2 shows that the SD cohort had a 3.52-fold increased risk of developing a migraine compared with the controls, after adjustment for age, sex, comorbidities, and drug use (aHR = 3.52, 95% CI = 3.28–3.79). Regardless of sex, patients in the SD cohort had an increased risk of developing a migraine compared with those in the comparison cohort, with an even higher risk for the men (aHR = 3.35, 95% CI = 3.10–3.64 for women; aHR = 4.31, 95% CI = 3.65–5.09 for men). The age-specific analysis showed that Table 1 The characteristic of demographics, comorbidity and history of drug use in non-apnea sleep disorder cohort with comparing to controls. Non-apnea sleep disorder
Sex Women Men Age, year 21–35 36–45 46–55 56–65 66–75 N75 Mean (SD) Comorbidity Hypertension Hyperlipidemia Diabetes Stroke COPD Depression Anxiety Drug use Non Zolpidem BZD Both#
No (N = 93552)
Yes (N = 46777)
n
%
n
%
59,190 34,362
63.3 36.7
29,595 17,182
63.3 36.7
17,890 19,946 20,310 14,436 13,510 7460 50.5 (16.3)
19.1 21.3 21.7 15.4 14.4 7.97
8945 9973 10,155 7218 6755 3731 50.5 (16.2)
19.1 21.3 21.7 15.4 14.4 7.98
36,400 23,093 18,303 19,905 21,007 3243 7886
38.9 24.7 19.6 21.3 22.5 3.47 8.43
23,372 16,803 11,579 15,024 17,166 9965 17,426
50.0 35.9 24.8 32.1 36.7 21.3 37.3
22,520 695 57,259 13,078
24.1 0.74 61.2 14.0
1346 254 16,016 29,161
2.88 0.54 34.2 62.3
p-Value 0.99
0.99
SD: standard deviation; #Both: patient ever used zolpidem and BZD.
0.4551 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001
234
T. Harnod et al. / European Journal of Internal Medicine 26 (2015) 232–236
Table 3 shows the following statistically significant interactions of 7 comorbidities with NSD for developing a migraine: hypertension (interaction P = .0001), hyperlipidemia (interaction P = .0001), diabetes (interaction P = .0052), stroke (interaction P = .0005), COPD (interaction P b .0001), depression (interaction P = .0061), and anxiety (interaction P b .0001). In our study cohort, patients with an NSD and any comorbidity were demonstrated to have higher risks of developing a migraine compared with those of patients in the comparison group. However, only in the patients with depression or anxiety was the risk of developing a migraine further increased compared with patients with an NSD and without comorbid depression or anxiety (aHR = 4.07, with SD and depression; aHR = 3.90, with SD and no depression; aHR = 4.60, with SD and anxiety; aHR = 3.39, with SD and no anxiety). 4. Discussion Sleep is a highly organized and dynamic behavior that is regulated by the internal homeostatic and circadian rhythm. However, many environmental factors can reduce or disrupt sleep, and sleep disorders with dyssomnia or insomnia may be comorbidities or predisposing factors for many neurological and psychiatric diseases [10,14]. A migraine is a neurovascular disorder and is associated with an excitatory–inhibitory imbalance over dura, cortex, subcortical, and deep structures inside the brain [15,16]. In the brain of migraineurs, the cortex is hyperexcitable to either enhanced excitation or reduced inhibition, even during the pain-free intervals. This dysfunction in central-information processing can progressively increase the brain's susceptibility to a migraine triggers, such as sleep deprivation, fatigue, oversleeping, stress, hunger, or prolonged sensory stimulation, and generate a migraine attack [16,17]. The long-term triggers of an NSD with dyssomnia or insomnia to a relatively excitable brain may increase the risk of developing a hyperexcitable brain and, thus, a migraine. This development effectively explained why we observed an obviously high risk of developing a subsequent migraine in the SD cohort, even in participants without apnea hypoxic damage to the brain. Previous studies have found that the orexin system of the brain also plays a role in the association between sleep and developing a migraine [10,18]. In the lateral hypothalamus, orexin releasing neurons can fire during active waking and virtually cease firing during sleep. Meanwhile, orexin could promote trigeminovascular nociception and a migraine
Fig. 1. Cumulative incidence of developing migraine between the non-apnea sleep disorder cohort (dashed line) and comparison cohort (solid line).
incidences of developing a migraine decreased with age in both cohorts, particularly in the SD cohort. The age-stratified effect of an NSD on developing a migraine was highest in participants aged 55 years and younger, with the risk decreasing for participants aged 56 years and older. The association between an NSD and comorbidity for developing a migraine was analyzed, and a much higher risk was observed among participants without any comorbidity (aHR = 4.66, 95% CI = 4.11–5.28), while a less high risk of developing a migraine existed for patients with one or more comorbidity (aHR = 3.23, 95% CI = 2.97–3.50). When the aHR of developing a migraine in the SD cohort was stratified by drug use or not, the aHR was 7.04 (95% CI = 5.56–8.92) in participants without drug use and 3.32 (95% CI = 3.08–3.58) in participants with drug use for their dyssomnia or insomnia.
Table 2 Incidence and adjusted hazard ratio of developing migraine stratified by sex, age, comorbidity and drug use between non-apnea sleep disorder cohort and comparison cohort. Non-apnea sleep disorders
Compared with control
No
Yes
Variables
Event
PY
Rate
Event
PY
Rate
IRR (95% CI)
Adjusted HR (95% CI)
Overall Sex Women Men Age (years) 21–35 36–45 46–55 56–65 66–75 N75 Comorbidity No Yes Drug use No Yes
1412
869,136
1.62
2713
432,029
6.28
3.87(3.74–4.00)⁎⁎⁎
3.52(3.28–3.79)⁎⁎⁎
1154 2096
558,703 310,433
2.07 0.83
258 617
277,701 154,328
7.55 4.00
3.65(3.51–3.81)⁎⁎⁎ 4.81(4.53–5.11)⁎⁎⁎
3.35(3.10–3.64)⁎⁎⁎ 4.31(3.65–5.09)⁎⁎⁎
290 383 305 214 164 56
170,314 196,006 200,154 139,041 116,305 47,315
1.70 1.95 1.52 1.54 1.41 1.18
588 775 627 392 261 70
85,563 95,412 97,663 68,067 59,333 25,991
6.87 8.12 6.42 5.76 4.40 2.69
4.04(3.74–4.35)⁎⁎⁎ 4.16(3.87–4.46)⁎⁎⁎ 4.21(3.92–4.53)⁎⁎⁎ 3.74(3.43–4.08)⁎⁎⁎ 3.12(2.85–3.41)⁎⁎⁎ 2.28(2.01–2.58)⁎⁎⁎
3.85(3.28–4.52)⁎⁎⁎ 3.57(3.11–4.10)⁎⁎⁎ 3.88(3.33–4.53)⁎⁎⁎ 3.28(2.73–3.95)⁎⁎⁎ 3.28(2.65–4.06)⁎⁎⁎ 2.25(1.54–3.28)⁎⁎⁎
556 856
369,423 499,713
1.51 1.71
565 2148
77,475 354,554
7.29 6.06
4.85(4.58–5.13)⁎⁎⁎ 3.54(3.39–3.69)⁎⁎⁎
4.66(4.11–5.28)⁎⁎⁎ 3.23(2.97–3.50)⁎⁎⁎
244 1168
208,304 660832
1.17 1.77
104 2609
12,316 419713
8.44 6.22
7.21(6.58–7.90)⁎⁎⁎ 3.52(3.39-3.65)⁎⁎⁎
7.04(5.56–8.92)⁎⁎⁎ 3.32(3.08-3.58)⁎⁎⁎
PY: person-year; rate: incidence rate (per 1000 person-years); IRR: incidence rate ratio; adjusted HR: multiple analysis including age, sex, comorbidities and drug use. ⁎⁎⁎ p b .001.
T. Harnod et al. / European Journal of Internal Medicine 26 (2015) 232–236 Table 3 The adjusted hazard ratios of migraine associated non-apnea sleep disorders and interaction with comorbidities. Variable Sleep disorders No No Yes Yes Sleep disorders No No Yes Yes Sleep disorders No No Yes Yes Sleep disorders No No Yes Yes Sleep disorders No No Yes Yes Sleep disorders No No Yes Yes Sleep disorders No No Yes Yes
N
Event
Adjusted HR (95% CI)
Hypertension
p-Value# 0.0001
No Yes No Yes Hyperlipidemia
57,152 946 36,400 466 23,405 1705 23,372 1008
1.00 0.90(0.80–1.01) 4.36(4.03–4.72)⁎⁎⁎ 2.99(2.72–3.30)⁎⁎⁎
No Yes No Yes Diabetes
70,459 1086 23,093 326 29,974 1964 16,803 749
No Yes No Yes Stroke
75,249 1207 18,303 205 35,198 2298 11,579 415
No Yes No Yes COPD
73,647 1137 19,905 275 31,753 2027 15,024 686
No Yes No Yes Depression
72,545 1056 1.00 21,007 356 1.45(1.28–1.63)⁎⁎⁎ 29,611 1868 | 4.31(4.00–4.65)⁎⁎⁎ 17,166 845 3.97(3.62–4.36)⁎⁎⁎
No Yes No Yes Anxiety
90,309 1341 3243 71 36,812 2091 9965 622
No Yes No Yes
85,666 1194 7886 218 29,351 1563 17,426 1150
0.0001 1.00 0.95(0.84–1.08) 4.26(3.96–4.59)⁎⁎⁎ 3.01(2.73–3.31)⁎⁎⁎ 0.0052 1.00 0.81(0.69–0.94)⁎⁎ 4.09(3.81–4.38)⁎⁎⁎ 2.54(2.27–2.85)⁎⁎⁎ 0.0005 1.00 1.19(1.04–1.37)⁎ 4.11(3.82–4.42)⁎⁎⁎ 3.70(3.35–4.08)⁎⁎⁎ b.0001
0.0061 1.00 1.49(1.18–1.90)⁎⁎⁎ 3.90(3.64–4.18)⁎⁎⁎ 4.07(3.70–4.48)⁎⁎⁎
235
that the link between an NSD and developing a migraine is stronger in men than in women. Our results also show that all of the 7 comorbidities (hypertension, hyperlipidemia, diabetes, stroke, COPD, depression, and anxiety) exhibited significant interactions with NSDs to develop a migraine. However, the comorbidities did not have independent effects on the hazard ratio; the patients without any comorbidity had a higher risk of developing a migraine (aHR = 4.66 vs 3.23). The risk of developing a migraine was much lower for patients without drug use than for patients taking zolpidem or BZD for their NSD (aHR = 3.32 vs 7.04). Therefore, we suggest that an NSD could be considered an independent, predisposing factor for developing a subsequent migraine in adulthood. The strengths of this study are its nationwide population-based design and the representativeness of the cohort. However, this study has several limitations. First, information on migraine frequency, the presence or absence of auras, smoking habits, alcohol consumption, body mass index, socioeconomic status, and family history were not available in the NHIRD; any of these characteristics might have been confounding factors for the study. Second, the evidence derived from a cohort study and a cohort-study design is subject to biases related to adjusting for confounders. Despite our meticulous study design with adequate controls for confounding factors, a key limitation was that bias would remain if unmeasured or unknown confounders were present. Finally, the diagnoses in the NHI claims data are primarily used for administrative billing and do not undergo verification for scientific purposes. We could not approach the patients directly to obtain the details of their medical histories or their medication use because of the anonymity of their identification numbers. However, several studies have reported the high accuracy and validity of diagnoses made using ICD-9 codes in the NHIRD and of similar study designs [23,24], indicating that the results of the present study are valuable for understanding the association between an NSD and developing a migraine. 5. Conclusion
b.0001 1.00 1.94(1.68–2.24)⁎⁎⁎ 3.93(3.64–4.23)⁎⁎⁎ 4.60(4.24–4.99)⁎⁎⁎
Model adjusted for age and sex. # p-value for interaction. ⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001.
headache could be triggered by the stimuli that activate the hypothalamus and orexin system, such as stress, fatigue, sleep deprivation, or poor sleep hygiene [19,20]. Furthermore, in the autonomic nervous system, there is a relative shift from sympathetic to parasympathetic tone with decreases of blood pressure and muscle tone during sleep. An NSD with repetitive dyssomnia or insomnia can cause a higher sympathetic tone and then result in a higher risk of developing a subsequent migraine [17,21]. The prevalence of migraines is well known to be higher from young adulthood to middle age, with a higher prevalence in women than men, and then to decrease among elderly persons [2–4]. In this study, our data showed a similar age distribution that higher risk of developing a subsequent migraine was noted in SD patients aged 21 to 55 years and the risk decreased more in patients aged 56 to 75 years and older. However, unexpectedly, the men had a higher risk of developing a migraine in later life than the women did (aHR = 4.31 vs 3.35). The prevalence of migraines can decrease with advancing age and may also improve in some women after menopause [22]. A consideration of previous epidemiological data with our own might suggest that the link between migraines and sex hormones is stronger in women and
This population-based cohort study indicates a higher risk of developing a subsequent migraine in patients with an NSD and suggests that NSDs be considered independent, predisposing factors for developing a subsequent migraine in adulthood. Collaterally, in consideration of the high prevalence of these two disorders, our study may suggest a cumulative impact of combining NSD and migraine in terms of disease burden. Additional large, unbiased, population-based studies are necessary before any confirmatory conclusion can be drawn. Clinical or public health relevance 1. The cumulative incidence of a migraine was significantly higher in the sleep disorder cohort. 2. The risk of developing a migraine was higher in men, aged 55 years, and younger. 3. Higher risks of migraine were observed among the participants without any comorbidity and drug treatment. 4. Insomnia could be considered an independent, predisposing factor for developing subsequent a migraine.
Author contributions Conception/Design: Tomor Harnod, Chia-Hung Kao Provision of study material and patients: Chia-Hung Kao Collection and assembly of data: All authors Data analysis and interpretation: All authors Manuscript writing: All authors Final approval of manuscript: All authors
236
T. Harnod et al. / European Journal of Internal Medicine 26 (2015) 232–236
Conflict of interest All of the authors declare no conflict of interest. Acknowledgments This study is supported in part by the Taiwan Ministry of Health and Welfare Clinical Trial and Research Center of Excellence (MOHW104TDU-B-212-113002); China Medical University Hospital, Academia Sinica Taiwan Biobank, Stroke Biosignature Project (BM104010092); NRPB Stroke Clinical Trial Consortium (MOST 103-2325-B-039-006); Tseng-Lien Lin Foundation, Taichung, Taiwan; Taiwan Brain Disease Foundation, Taipei, Taiwan; Katsuzo and Kiyo Aoshima Memorial Funds, Japan; and Health, and Welfare Surcharge of Tobacco Products, China Medical University Hospital Cancer Research Center of Excellence (MOHW104-TDU-B-212-124-002, Taiwan). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. No additional external funding received for this study. References [1] Wang SJ, Chung CS, Chankrachang S, Ravishankar K, Merican JS, Salazar G, et al. Migraine disability awareness campaign in Asia: migraine assessment for prophylaxis. Headache 2008;48:1356–65. [2] Cheung RT. Prevalence of migraine, tension-type headache, and other headaches in Hong Kong. Headache 2000;40:473–9. [3] Lipton RB, Stewart WF, Diamond S, Diamond ML, Reed M. Prevalence and burden of migraine in the United States: data from the American Migraine Study II. Headache 2001;41:646–57. [4] Stovner LJ, Zwart JA, Hagen K, Terwindt GM, Pascual J. Epidemiology of headache in Europe. Eur J Neurol 2006;13:333–45. [5] Kurth T, Gaziano JM, Cook NR, Logroscino G, Diener HC, Buring JE. Migraine and risk of cardiovascular disease in women. JAMA 2006;296:283–91. [6] Kurth T, Gaziano JM, Cook NR, Logroscino G, Diener HC, Buring JE. Migraine and risk of cardiovascular disease in men. Arch Intern Med 2007;167:795–801.
[7] Jensen R, Stovner LJ. Epidemiology and comorbidity of headache. Lancet Neurol 2008;7:354–61. [8] Kelman L, Rains JC. Headache and sleep: examination of sleep patterns and complaints in a large clinical sample of migraineurs. Headache 2005;45:904–10. [9] Pompili M, Cosimo DD, Innamorati M, Lester D, Tatarelli R, Martelletti P. Psychiatric comorbidity in patients with chronic daily headache and migraine: a selective overview including personality traits and suicidal risk. J Headache Pain 2009;10: 283–90. [10] Bigal ME, Hargreaves RJ. Why does sleep stop migraine? Curr Pain Headache Rep 2013;17:369. http://dx.doi.org/10.1007/s11916-013-0369-0. [11] Haque B, Rahman KM, Hoque A, Hasan AT, Chowdhury RN, Khan SU, et al. Precipitating and relieving factors of migraine versus tension type headache. BMC Neurol 2012;12:82. http://dx.doi.org/10.1186/1471-2377-12-82. [12] Ohayon MM. Observation of the natural evolution of insomnia in the American general population cohort. Sleep Med Clin 2009;4:87–92. [13] Arruda MA, Guidetti V, Galli F, Albuquerque RC, Bigal ME. Childhood periodic syndromes: a population-based study. Pediatr Neurol 2010;43:420–4. [14] Roth T, Roehrs T. Sleep organization and regulation. Neurology 2000;54:S2–7. [15] Murinova N, Krashin DL, Lucas S. Vascular risk in migraineurs: interaction of endothelial and cortical excitability factors. Headache 2014;54:583–90. [16] Vecchia D, Pietrobon D. Migraine: a disorder of brain excitatory-inhibitory balance? Trends Neurosci Aug 2012;35(8):507–20. [17] Chabriat H, Danchot J, Michel P, Joire JE, Henry P. Precipitating factors of headache. A prospective study in a national control-matched survey in migraineurs and nonmigraineurs. Headache 1999;39:335–8. [18] Holland PR. Headache and sleep: shared pathophysiological mechanisms. Cephalalgia 2014;34:725–44. [19] Buijs RM, van Eden CG. The integration of stress by the hypothalamus, amygdala and prefrontal cortex: balance between the autonomic nervous system and the neuroendocrine system. Prog Brain Res 2000;126:117–32. [20] Overeem S, van Vliet JA, Lammers GJ, Zitman FG, Swaab DF, Ferrari MD. The hypothalamus in episodic brain disorders. Lancet Neurol 2002;1:437–44. [21] De Lecea L. Hypocretins and the neurobiology of sleep–wake mechanisms. Prog Brain Res 2012;198:15–23. [22] Ashkenazi A, Silberstein SD. Hormone-related headache: pathophysiology and treatment. CNS Drugs 2006;20:125–41. [23] Cheng CL, Kao YH, Lin SJ, Lee CH, Lai ML. Validation of the National Health Insurance Research Database with ischemic stroke cases in Taiwan. Pharmacoepidemiol Drug Saf 2011;20:236–42. [24] Yu YB, Gau JP, Liu CY, Yan MH, Chiang SC, Hsu HC, et al. A nation-wide analysis of venous thromboembolism in 497,180 cancer patients with the development and validation of a risk-stratification scoring system. Thromb Haemost 2012;108: 225–35.
Contents lists available at ScienceDirect
European Journal of Internal Medicine journal homepage: www.elsevier.com/locate/ejim
Original Article
Higher risk of developing a subsequent migraine in adults with nonapnea sleep disorders: A nationwide population-based cohort study Tomor Harnod a,b, Yu-Chiao Wang c,d, Chia-Hung Kao e,f,⁎ a
Department of Neurosurgery, Hualien Tzu Chi General Hospital, Buddhist Tzu Chi Medical Foundation, Hualien, Taiwan College of Medicine, Tzu Chi University, Hualien, Taiwan Management Office for Health Data, China Medical University Hospital, Taichung, Taiwan d College of Medicine, China Medical University, Taichung, Taiwan e Graduate Institute of Clinical Medical Science and School of Medicine, College of Medicine, China Medical University, Taichung, Taiwan f Department of Nuclear Medicine and PET Center, China Medical University Hospital, Taichung, Taiwan b c
a r t i c l e
i n f o
Article history: Received 5 February 2015 Received in revised form 25 February 2015 Accepted 2 March 2015 Available online 19 March 2015 Keywords: Cohort study Insomnia Migraine National Health Insurance (NHI)
a b s t r a c t Objective: This nationwide population-based cohort study evaluated the effect of nonapnea sleep disorders (NSDs) on the subsequent development of a migraine. Methods: We identified 46,777 patients aged 20 years and older who were diagnosed with an NSD (ICD-9-CM: 307.4 or 780.5) and without coding for apnea-sleep disorders (ICD-9-CM: 780.51, 780.53, or 780.57) between 2000 and 2002 as the sleep disorder (SD) cohort. A comparison cohort of 93,552 people was enrolled. We calculated the adjusted hazard ratio (aHR) for developing a migraine (ICD-9-CM: 346) after adjusting for age, sex, comorbidity, and drug use. A Kaplan–Meier analysis was used to measure the cumulative incidence of a migraine between 2 curves until the end of 2011. Results: The cumulative incidence of a migraine was significantly higher in the SD cohort. The aHR for developing a migraine in the SD cohort was 3.52 (95% CI = 3.28–3.79). The risk of developing a migraine with an NSD was higher in men (aHR = 4.31) than in women (aHR = 3.35). The age-stratified effect of an NSD on developing a migraine was highest among patients aged 55 years and younger. Higher risks of developing a migraine were observed among the participants without any comorbidity and without any drug treatment for their insomnia. Conclusion: The findings of this population-based cohort study indicate a higher risk of developing a subsequent migraine in patients with an NSD, which could be considered an independent, predisposing factor for developing subsequent a migraine in adulthood. © 2015 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.
1. Introduction Both headaches and insomnia are common complaints of patients that affect the quality of their daily lives. In Asia, migraines were reported as the most prevalent type of headache diagnosed at neurological clinics with a range from 50.9% to 85.8% across various countries [1]. The estimated annual worldwide prevalence of migraines is approximately 15–20% in women and 6–10% in men respectively [2–4]. Migraines have been considered a neurovascular disease for decades and known to have higher associated risks of developing vascular diseases in both sexes [5,6]. Many studies reported the interacting relationships or comorbidities between sleep disorders, anxiety, depression, and migraines [7–9]. In migraineurs, fatigue, sleep deprivation, or simply a change of sleep pattern is frequently reported as the precipitating factor ⁎ Corresponding author at: Graduate Institute of Clinical Medical Science and School of Medicine, College of Medicine, China Medical University, No. 2, Yuh-Der Road, Taichung 404, Taiwan. Tel.: +886 4 22052121x7412; fax: +886 4 22336174. E-mail address: [email protected] (C.-H. Kao).
for headache attacks, with sleep itself being a relieving treatment for the headache [10,11]. Sleep is regulated by the circadian rhythm and is vital to life. Sleep dysfunction can have profound consequences on physiology, behavior, and daily abilities during waking hours [10,12]. Symptoms of dyssomnia or insomnia may be a temporary inconvenience or may chronically affect a person, owing to the many available types of sleep disorder (SDs). Among them, apnea-sleep disorders can result in hypoxic damage to multiple organs and are considered to pose a higher threat to health and for developing subsequent diseases. However, a population-based study analyzed data in 1906 Brazilian children with age of 5–12 years to survey their migraine or tension-type headache using validated questionnaires by interviewing the parents. Interestingly, in the children with migraine, some kinds of parasomnias, such as sleep talking, somnambulism, and bruxism were independently associated with the risk of their migraine headaches [13]. We have sought to determine whether NSDs could be considered independent, predisposing factors for developing a subsequent migraine in adulthood. To answer this question of causeand-effect, we conducted this study to determine the risk of developing
http://dx.doi.org/10.1016/j.ejim.2015.03.002 0953-6205/© 2015 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.
T. Harnod et al. / European Journal of Internal Medicine 26 (2015) 232–236
a migraine in a cohort with an NSD by using the Taiwan nationwide population-based database. 2. Methods and materials 2.1. Study design In this population-based cohort study, we used the Longitudinal Health Insurance Database (LHID) to establish the study cohort. The LHID database is one part of Taiwan's National Health Insurance Research Database (NHIRD), which contains insurance claim data for one million insured people randomly selected from Taiwan's entire 23-million population and covered by the National Health Insurance (NHI) program. The NHIRD is maintained by the National Health Research Institutes. Taiwan's NHI program has been a universal insurance system since 1995 for the residents of Taiwan, with more than 99% of the residents covered by the end of 2000. NHI-participant information that has been recorded in the claims data includes demographic characteristics and details regarding outpatient and inpatient services received from 1996 to 2011. We identified participants with specific diseases by using diagnostic codes from the International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM), which were diagnosed by physicians, either a general physician or a neurologist. The patient identification numbers were scrambled to safeguard patient privacy and confidentiality when linking to the data files. This study was granted ethical approval by the institutional review board of China Medical University (CMU-REC-101-012). 2.2. Selection of study participants Our study population was purposely focused on LHID patients aged 20 years and older with the diagnosis of an NSD (SD cohort). The definition of an NSD was sleep disorders (ICD-9-CM: 307.4 or 780.5) without coding for apnea SD (ICD-9-CM: 780.51, 780.53, or 780.57). We identified a first time diagnosis of an NSD as the index date between the years of 2000 and 2002, and we excluded the ones who had already developed a migraine (ICD-9-CM: 346) before the index date. The comparison cohort was randomly selected from LHID patients without sleep disorders. Each NSD patient was matched to two controls by frequency matching for age, sex, and index year. We analyzed the risk of developing a subsequent migraine between the SD and comparison cohorts. All study participants were followed up from the index date to the occurrence of a migraine, withdrawal from the LHID database, or until December 31, 2011. We also investigated all participants who had ever used zolpidem or benzodiazepines (BZD) before the end date and categorized them into four groups according to treatment for their dyssomnia or insomnia: without drug use, zolpidem use only, BZD use only, or use of both drugs. Several diseases, such as hypertension (ICD-9-CM: 401–405), hyperlipidemia (ICD-9-CM: 272), diabetes (ICD-9-CM: 250), stroke (ICD-9CM: 430–438), chronic obstructive pulmonary disease (COPD; ICD-9CM: 490–496), depression (ICD-9-CM: 2962, 2963, 29682, 3004, 3090, 3091, 30928, and 311), anxiety (ICD-9-CM: 3000, 3002, 3003, 3083, and 30981), were defined as comorbidities if one or more of them were diagnosed before the index date or during the follow-up period. 2.3. Data analysis A chi-square test and Student's t test were applied for a baseline demographic factors, comorbidity and drug-use analysis between the SD and comparison cohorts. The chi-square test was used to analyze the categorical variables, and the Student's t test was used to assess the continuous variables. In both cohorts, we calculated the incidence rate (per 1000 person-years) of a migraine with stratifying by age, sex, comorbidity (no or yes), and drug use (no or yes).
233
The incidence rate ratio (IRR) and the 95% confidence interval (95% CI) of developing a migraine in the SD cohort were calculated by the Poisson regression model and compared with the controls. After we adjusted for confounding risk factors, we determined the adjusted hazard ratio (aHR) and a 95% CI for a migraine by using the multivariable Cox proportional hazards model. We also used the Cox proportional hazards model to estimate the interaction between an NSD and the type of comorbidity. Cumulative incidence analysis was conducted using the Kaplan–Meier method, and the differences between two curves were calculated with the two-tailed log-rank test. All of the analyses were performed using the SAS Version 9.3 statistical package (SAS Institute Inc., NC, USA), and the P value (P b .05) in the two-tailed tests was determined to be statistically significant. 3. Results Table 1 shows the demographic data of the SD and comparison cohorts at baseline; 46,777 patients were recorded as having a first time, primary diagnosis of an NSD between 2000 and 2002, and 93,552 participants were selected as the controls by frequency matching for age, sex, and index year. After the frequency matching, the distributions of age and sex in the SD and comparison cohorts were similar; the proportion of women was 63.3%, and the mean age for the participants was 50.5 years. Comorbidities were more prevalent in the SD cohort than in the comparison cohort at baseline (P b .0001). Instances of drug use were also more common in the SD cohort (97.1%) than in the comparison cohort (75.9%). The cumulative incidence of developing a migraine was significantly higher in patients with an NSD than in the controls during the 12-year follow-up (log-rank test P b .0001; Fig. 1). Table 2 shows that the SD cohort had a 3.52-fold increased risk of developing a migraine compared with the controls, after adjustment for age, sex, comorbidities, and drug use (aHR = 3.52, 95% CI = 3.28–3.79). Regardless of sex, patients in the SD cohort had an increased risk of developing a migraine compared with those in the comparison cohort, with an even higher risk for the men (aHR = 3.35, 95% CI = 3.10–3.64 for women; aHR = 4.31, 95% CI = 3.65–5.09 for men). The age-specific analysis showed that Table 1 The characteristic of demographics, comorbidity and history of drug use in non-apnea sleep disorder cohort with comparing to controls. Non-apnea sleep disorder
Sex Women Men Age, year 21–35 36–45 46–55 56–65 66–75 N75 Mean (SD) Comorbidity Hypertension Hyperlipidemia Diabetes Stroke COPD Depression Anxiety Drug use Non Zolpidem BZD Both#
No (N = 93552)
Yes (N = 46777)
n
%
n
%
59,190 34,362
63.3 36.7
29,595 17,182
63.3 36.7
17,890 19,946 20,310 14,436 13,510 7460 50.5 (16.3)
19.1 21.3 21.7 15.4 14.4 7.97
8945 9973 10,155 7218 6755 3731 50.5 (16.2)
19.1 21.3 21.7 15.4 14.4 7.98
36,400 23,093 18,303 19,905 21,007 3243 7886
38.9 24.7 19.6 21.3 22.5 3.47 8.43
23,372 16,803 11,579 15,024 17,166 9965 17,426
50.0 35.9 24.8 32.1 36.7 21.3 37.3
22,520 695 57,259 13,078
24.1 0.74 61.2 14.0
1346 254 16,016 29,161
2.88 0.54 34.2 62.3
p-Value 0.99
0.99
SD: standard deviation; #Both: patient ever used zolpidem and BZD.
0.4551 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001
234
T. Harnod et al. / European Journal of Internal Medicine 26 (2015) 232–236
Table 3 shows the following statistically significant interactions of 7 comorbidities with NSD for developing a migraine: hypertension (interaction P = .0001), hyperlipidemia (interaction P = .0001), diabetes (interaction P = .0052), stroke (interaction P = .0005), COPD (interaction P b .0001), depression (interaction P = .0061), and anxiety (interaction P b .0001). In our study cohort, patients with an NSD and any comorbidity were demonstrated to have higher risks of developing a migraine compared with those of patients in the comparison group. However, only in the patients with depression or anxiety was the risk of developing a migraine further increased compared with patients with an NSD and without comorbid depression or anxiety (aHR = 4.07, with SD and depression; aHR = 3.90, with SD and no depression; aHR = 4.60, with SD and anxiety; aHR = 3.39, with SD and no anxiety). 4. Discussion Sleep is a highly organized and dynamic behavior that is regulated by the internal homeostatic and circadian rhythm. However, many environmental factors can reduce or disrupt sleep, and sleep disorders with dyssomnia or insomnia may be comorbidities or predisposing factors for many neurological and psychiatric diseases [10,14]. A migraine is a neurovascular disorder and is associated with an excitatory–inhibitory imbalance over dura, cortex, subcortical, and deep structures inside the brain [15,16]. In the brain of migraineurs, the cortex is hyperexcitable to either enhanced excitation or reduced inhibition, even during the pain-free intervals. This dysfunction in central-information processing can progressively increase the brain's susceptibility to a migraine triggers, such as sleep deprivation, fatigue, oversleeping, stress, hunger, or prolonged sensory stimulation, and generate a migraine attack [16,17]. The long-term triggers of an NSD with dyssomnia or insomnia to a relatively excitable brain may increase the risk of developing a hyperexcitable brain and, thus, a migraine. This development effectively explained why we observed an obviously high risk of developing a subsequent migraine in the SD cohort, even in participants without apnea hypoxic damage to the brain. Previous studies have found that the orexin system of the brain also plays a role in the association between sleep and developing a migraine [10,18]. In the lateral hypothalamus, orexin releasing neurons can fire during active waking and virtually cease firing during sleep. Meanwhile, orexin could promote trigeminovascular nociception and a migraine
Fig. 1. Cumulative incidence of developing migraine between the non-apnea sleep disorder cohort (dashed line) and comparison cohort (solid line).
incidences of developing a migraine decreased with age in both cohorts, particularly in the SD cohort. The age-stratified effect of an NSD on developing a migraine was highest in participants aged 55 years and younger, with the risk decreasing for participants aged 56 years and older. The association between an NSD and comorbidity for developing a migraine was analyzed, and a much higher risk was observed among participants without any comorbidity (aHR = 4.66, 95% CI = 4.11–5.28), while a less high risk of developing a migraine existed for patients with one or more comorbidity (aHR = 3.23, 95% CI = 2.97–3.50). When the aHR of developing a migraine in the SD cohort was stratified by drug use or not, the aHR was 7.04 (95% CI = 5.56–8.92) in participants without drug use and 3.32 (95% CI = 3.08–3.58) in participants with drug use for their dyssomnia or insomnia.
Table 2 Incidence and adjusted hazard ratio of developing migraine stratified by sex, age, comorbidity and drug use between non-apnea sleep disorder cohort and comparison cohort. Non-apnea sleep disorders
Compared with control
No
Yes
Variables
Event
PY
Rate
Event
PY
Rate
IRR (95% CI)
Adjusted HR (95% CI)
Overall Sex Women Men Age (years) 21–35 36–45 46–55 56–65 66–75 N75 Comorbidity No Yes Drug use No Yes
1412
869,136
1.62
2713
432,029
6.28
3.87(3.74–4.00)⁎⁎⁎
3.52(3.28–3.79)⁎⁎⁎
1154 2096
558,703 310,433
2.07 0.83
258 617
277,701 154,328
7.55 4.00
3.65(3.51–3.81)⁎⁎⁎ 4.81(4.53–5.11)⁎⁎⁎
3.35(3.10–3.64)⁎⁎⁎ 4.31(3.65–5.09)⁎⁎⁎
290 383 305 214 164 56
170,314 196,006 200,154 139,041 116,305 47,315
1.70 1.95 1.52 1.54 1.41 1.18
588 775 627 392 261 70
85,563 95,412 97,663 68,067 59,333 25,991
6.87 8.12 6.42 5.76 4.40 2.69
4.04(3.74–4.35)⁎⁎⁎ 4.16(3.87–4.46)⁎⁎⁎ 4.21(3.92–4.53)⁎⁎⁎ 3.74(3.43–4.08)⁎⁎⁎ 3.12(2.85–3.41)⁎⁎⁎ 2.28(2.01–2.58)⁎⁎⁎
3.85(3.28–4.52)⁎⁎⁎ 3.57(3.11–4.10)⁎⁎⁎ 3.88(3.33–4.53)⁎⁎⁎ 3.28(2.73–3.95)⁎⁎⁎ 3.28(2.65–4.06)⁎⁎⁎ 2.25(1.54–3.28)⁎⁎⁎
556 856
369,423 499,713
1.51 1.71
565 2148
77,475 354,554
7.29 6.06
4.85(4.58–5.13)⁎⁎⁎ 3.54(3.39–3.69)⁎⁎⁎
4.66(4.11–5.28)⁎⁎⁎ 3.23(2.97–3.50)⁎⁎⁎
244 1168
208,304 660832
1.17 1.77
104 2609
12,316 419713
8.44 6.22
7.21(6.58–7.90)⁎⁎⁎ 3.52(3.39-3.65)⁎⁎⁎
7.04(5.56–8.92)⁎⁎⁎ 3.32(3.08-3.58)⁎⁎⁎
PY: person-year; rate: incidence rate (per 1000 person-years); IRR: incidence rate ratio; adjusted HR: multiple analysis including age, sex, comorbidities and drug use. ⁎⁎⁎ p b .001.
T. Harnod et al. / European Journal of Internal Medicine 26 (2015) 232–236 Table 3 The adjusted hazard ratios of migraine associated non-apnea sleep disorders and interaction with comorbidities. Variable Sleep disorders No No Yes Yes Sleep disorders No No Yes Yes Sleep disorders No No Yes Yes Sleep disorders No No Yes Yes Sleep disorders No No Yes Yes Sleep disorders No No Yes Yes Sleep disorders No No Yes Yes
N
Event
Adjusted HR (95% CI)
Hypertension
p-Value# 0.0001
No Yes No Yes Hyperlipidemia
57,152 946 36,400 466 23,405 1705 23,372 1008
1.00 0.90(0.80–1.01) 4.36(4.03–4.72)⁎⁎⁎ 2.99(2.72–3.30)⁎⁎⁎
No Yes No Yes Diabetes
70,459 1086 23,093 326 29,974 1964 16,803 749
No Yes No Yes Stroke
75,249 1207 18,303 205 35,198 2298 11,579 415
No Yes No Yes COPD
73,647 1137 19,905 275 31,753 2027 15,024 686
No Yes No Yes Depression
72,545 1056 1.00 21,007 356 1.45(1.28–1.63)⁎⁎⁎ 29,611 1868 | 4.31(4.00–4.65)⁎⁎⁎ 17,166 845 3.97(3.62–4.36)⁎⁎⁎
No Yes No Yes Anxiety
90,309 1341 3243 71 36,812 2091 9965 622
No Yes No Yes
85,666 1194 7886 218 29,351 1563 17,426 1150
0.0001 1.00 0.95(0.84–1.08) 4.26(3.96–4.59)⁎⁎⁎ 3.01(2.73–3.31)⁎⁎⁎ 0.0052 1.00 0.81(0.69–0.94)⁎⁎ 4.09(3.81–4.38)⁎⁎⁎ 2.54(2.27–2.85)⁎⁎⁎ 0.0005 1.00 1.19(1.04–1.37)⁎ 4.11(3.82–4.42)⁎⁎⁎ 3.70(3.35–4.08)⁎⁎⁎ b.0001
0.0061 1.00 1.49(1.18–1.90)⁎⁎⁎ 3.90(3.64–4.18)⁎⁎⁎ 4.07(3.70–4.48)⁎⁎⁎
235
that the link between an NSD and developing a migraine is stronger in men than in women. Our results also show that all of the 7 comorbidities (hypertension, hyperlipidemia, diabetes, stroke, COPD, depression, and anxiety) exhibited significant interactions with NSDs to develop a migraine. However, the comorbidities did not have independent effects on the hazard ratio; the patients without any comorbidity had a higher risk of developing a migraine (aHR = 4.66 vs 3.23). The risk of developing a migraine was much lower for patients without drug use than for patients taking zolpidem or BZD for their NSD (aHR = 3.32 vs 7.04). Therefore, we suggest that an NSD could be considered an independent, predisposing factor for developing a subsequent migraine in adulthood. The strengths of this study are its nationwide population-based design and the representativeness of the cohort. However, this study has several limitations. First, information on migraine frequency, the presence or absence of auras, smoking habits, alcohol consumption, body mass index, socioeconomic status, and family history were not available in the NHIRD; any of these characteristics might have been confounding factors for the study. Second, the evidence derived from a cohort study and a cohort-study design is subject to biases related to adjusting for confounders. Despite our meticulous study design with adequate controls for confounding factors, a key limitation was that bias would remain if unmeasured or unknown confounders were present. Finally, the diagnoses in the NHI claims data are primarily used for administrative billing and do not undergo verification for scientific purposes. We could not approach the patients directly to obtain the details of their medical histories or their medication use because of the anonymity of their identification numbers. However, several studies have reported the high accuracy and validity of diagnoses made using ICD-9 codes in the NHIRD and of similar study designs [23,24], indicating that the results of the present study are valuable for understanding the association between an NSD and developing a migraine. 5. Conclusion
b.0001 1.00 1.94(1.68–2.24)⁎⁎⁎ 3.93(3.64–4.23)⁎⁎⁎ 4.60(4.24–4.99)⁎⁎⁎
Model adjusted for age and sex. # p-value for interaction. ⁎ p b .05. ⁎⁎ p b .01. ⁎⁎⁎ p b .001.
headache could be triggered by the stimuli that activate the hypothalamus and orexin system, such as stress, fatigue, sleep deprivation, or poor sleep hygiene [19,20]. Furthermore, in the autonomic nervous system, there is a relative shift from sympathetic to parasympathetic tone with decreases of blood pressure and muscle tone during sleep. An NSD with repetitive dyssomnia or insomnia can cause a higher sympathetic tone and then result in a higher risk of developing a subsequent migraine [17,21]. The prevalence of migraines is well known to be higher from young adulthood to middle age, with a higher prevalence in women than men, and then to decrease among elderly persons [2–4]. In this study, our data showed a similar age distribution that higher risk of developing a subsequent migraine was noted in SD patients aged 21 to 55 years and the risk decreased more in patients aged 56 to 75 years and older. However, unexpectedly, the men had a higher risk of developing a migraine in later life than the women did (aHR = 4.31 vs 3.35). The prevalence of migraines can decrease with advancing age and may also improve in some women after menopause [22]. A consideration of previous epidemiological data with our own might suggest that the link between migraines and sex hormones is stronger in women and
This population-based cohort study indicates a higher risk of developing a subsequent migraine in patients with an NSD and suggests that NSDs be considered independent, predisposing factors for developing a subsequent migraine in adulthood. Collaterally, in consideration of the high prevalence of these two disorders, our study may suggest a cumulative impact of combining NSD and migraine in terms of disease burden. Additional large, unbiased, population-based studies are necessary before any confirmatory conclusion can be drawn. Clinical or public health relevance 1. The cumulative incidence of a migraine was significantly higher in the sleep disorder cohort. 2. The risk of developing a migraine was higher in men, aged 55 years, and younger. 3. Higher risks of migraine were observed among the participants without any comorbidity and drug treatment. 4. Insomnia could be considered an independent, predisposing factor for developing subsequent a migraine.
Author contributions Conception/Design: Tomor Harnod, Chia-Hung Kao Provision of study material and patients: Chia-Hung Kao Collection and assembly of data: All authors Data analysis and interpretation: All authors Manuscript writing: All authors Final approval of manuscript: All authors
236
T. Harnod et al. / European Journal of Internal Medicine 26 (2015) 232–236
Conflict of interest All of the authors declare no conflict of interest. Acknowledgments This study is supported in part by the Taiwan Ministry of Health and Welfare Clinical Trial and Research Center of Excellence (MOHW104TDU-B-212-113002); China Medical University Hospital, Academia Sinica Taiwan Biobank, Stroke Biosignature Project (BM104010092); NRPB Stroke Clinical Trial Consortium (MOST 103-2325-B-039-006); Tseng-Lien Lin Foundation, Taichung, Taiwan; Taiwan Brain Disease Foundation, Taipei, Taiwan; Katsuzo and Kiyo Aoshima Memorial Funds, Japan; and Health, and Welfare Surcharge of Tobacco Products, China Medical University Hospital Cancer Research Center of Excellence (MOHW104-TDU-B-212-124-002, Taiwan). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. No additional external funding received for this study. References [1] Wang SJ, Chung CS, Chankrachang S, Ravishankar K, Merican JS, Salazar G, et al. Migraine disability awareness campaign in Asia: migraine assessment for prophylaxis. Headache 2008;48:1356–65. [2] Cheung RT. Prevalence of migraine, tension-type headache, and other headaches in Hong Kong. Headache 2000;40:473–9. [3] Lipton RB, Stewart WF, Diamond S, Diamond ML, Reed M. Prevalence and burden of migraine in the United States: data from the American Migraine Study II. Headache 2001;41:646–57. [4] Stovner LJ, Zwart JA, Hagen K, Terwindt GM, Pascual J. Epidemiology of headache in Europe. Eur J Neurol 2006;13:333–45. [5] Kurth T, Gaziano JM, Cook NR, Logroscino G, Diener HC, Buring JE. Migraine and risk of cardiovascular disease in women. JAMA 2006;296:283–91. [6] Kurth T, Gaziano JM, Cook NR, Logroscino G, Diener HC, Buring JE. Migraine and risk of cardiovascular disease in men. Arch Intern Med 2007;167:795–801.
[7] Jensen R, Stovner LJ. Epidemiology and comorbidity of headache. Lancet Neurol 2008;7:354–61. [8] Kelman L, Rains JC. Headache and sleep: examination of sleep patterns and complaints in a large clinical sample of migraineurs. Headache 2005;45:904–10. [9] Pompili M, Cosimo DD, Innamorati M, Lester D, Tatarelli R, Martelletti P. Psychiatric comorbidity in patients with chronic daily headache and migraine: a selective overview including personality traits and suicidal risk. J Headache Pain 2009;10: 283–90. [10] Bigal ME, Hargreaves RJ. Why does sleep stop migraine? Curr Pain Headache Rep 2013;17:369. http://dx.doi.org/10.1007/s11916-013-0369-0. [11] Haque B, Rahman KM, Hoque A, Hasan AT, Chowdhury RN, Khan SU, et al. Precipitating and relieving factors of migraine versus tension type headache. BMC Neurol 2012;12:82. http://dx.doi.org/10.1186/1471-2377-12-82. [12] Ohayon MM. Observation of the natural evolution of insomnia in the American general population cohort. Sleep Med Clin 2009;4:87–92. [13] Arruda MA, Guidetti V, Galli F, Albuquerque RC, Bigal ME. Childhood periodic syndromes: a population-based study. Pediatr Neurol 2010;43:420–4. [14] Roth T, Roehrs T. Sleep organization and regulation. Neurology 2000;54:S2–7. [15] Murinova N, Krashin DL, Lucas S. Vascular risk in migraineurs: interaction of endothelial and cortical excitability factors. Headache 2014;54:583–90. [16] Vecchia D, Pietrobon D. Migraine: a disorder of brain excitatory-inhibitory balance? Trends Neurosci Aug 2012;35(8):507–20. [17] Chabriat H, Danchot J, Michel P, Joire JE, Henry P. Precipitating factors of headache. A prospective study in a national control-matched survey in migraineurs and nonmigraineurs. Headache 1999;39:335–8. [18] Holland PR. Headache and sleep: shared pathophysiological mechanisms. Cephalalgia 2014;34:725–44. [19] Buijs RM, van Eden CG. The integration of stress by the hypothalamus, amygdala and prefrontal cortex: balance between the autonomic nervous system and the neuroendocrine system. Prog Brain Res 2000;126:117–32. [20] Overeem S, van Vliet JA, Lammers GJ, Zitman FG, Swaab DF, Ferrari MD. The hypothalamus in episodic brain disorders. Lancet Neurol 2002;1:437–44. [21] De Lecea L. Hypocretins and the neurobiology of sleep–wake mechanisms. Prog Brain Res 2012;198:15–23. [22] Ashkenazi A, Silberstein SD. Hormone-related headache: pathophysiology and treatment. CNS Drugs 2006;20:125–41. [23] Cheng CL, Kao YH, Lin SJ, Lee CH, Lai ML. Validation of the National Health Insurance Research Database with ischemic stroke cases in Taiwan. Pharmacoepidemiol Drug Saf 2011;20:236–42. [24] Yu YB, Gau JP, Liu CY, Yan MH, Chiang SC, Hsu HC, et al. A nation-wide analysis of venous thromboembolism in 497,180 cancer patients with the development and validation of a risk-stratification scoring system. Thromb Haemost 2012;108: 225–35.
