Sage Journals: Discover world-class research

Abstract

Background

There is conflicting evidence for the association between migraine and increased mortality risk. The aim of this study was to investigate the relationship between migraine and non-migrainous headache, and all-cause mortality and cardiovascular mortality.

Methods

In this prospective population-based cohort study from Norway, we used baseline data from the second Nord-Trøndelag Health Survey (HUNT2), performed between 1995 and 1997 in the County of Nord-Trøndelag. These data were linked with a comprehensive mortality database with follow-up through the year 2011. A total of 51,853 (56% of invited) people were categorized based on their answers to the headache questions in HUNT2 (headache free, migraine or non-migrainous headache). Hazard ratios (HRs) of mortality during a mean of 14.1 years of follow-up were estimated using Cox regression.

Results

During the follow-up period 9408 died, 4321 of these from cardiovascular causes. There was no difference in all-cause mortality between individuals with migraine and non-migrainous headache compared to those without headache or between headache status and mortality by cardiovascular disease. There was, however, among men with migraine without aura a reduced risk of death by cardiovascular diseases (HR 0.72, 95% confidence interval 0.56–0.93). This relationship was not evident in women.

Conclusion

In this large, prospective cohort study there was no evidence for a higher all-cause mortality or cardiovascular mortality among individuals with migraine.

Keywords

Headache migraine mortality general population

Introduction

With a global prevalence of 15% (1 –3), migraine is ranked as the third most common disease in the world, and recognized as the seventh highest among specific causes of disability globally (3). While migraine, in particular migraine with aura, has been associated with increased risk of cardiovascular diseases (e.g. (4 –9)), the relationship between migraine and mortality remains unclear. Some studies have found increased all-cause mortality (10) and mortality from cardiovascular disease in migraine (10,11), whereas other studies have reported reduced all-cause mortality (12,13) or lack of statistically significant findings (14,15). Methodological differences make it difficult to compare studies and may explain some of the variation in results. A meta-analysis from 2011 reported that there was no evidence to suggest that any migraine alters the risk of all-cause mortality, cardiovascular disease-mortality or coronary heart disease-mortality (16), but underlined the need for further research.

The purpose of this large population-based follow-up study was to examine the relationship between migraine and non-migrainous headache and mortality.

Methods

This prospective population-based cohort study used baseline data from the second Nord-Trøndelag Health Survey (HUNT2) performed between August 1995 and June 1997. The HUNT data were linked to the Norwegian National Cause of Death register, and our study is based on mortality follow-up through the year 2011.

Participants

In HUNT2 all inhabitants ≥20 years old of Nord-Trøndelag were invited to participate. Each participant had to complete two extensive questionnaires, and participated in a brief medical examination including standardized measurements of height, weight and blood pressure (BP). Blood samples were also taken. Among 92,936 invited individuals, 64,787 (70%) participated, of whom 51,853 (56%) answered the headache questions in the second questionnaire and are the people included in this study (Figure 1). Details of the non-responders have been reported previously (17). Participating individuals tended to be younger, were more likely to be women and had higher socioeconomic status than non-responders (17).

Figure 1.

Flowchart of the study population.

Headache diagnoses

Headache questions were designed mainly to determine whether the participants had headache complaints, assess the frequency of attacks and diagnose migraine by applying modified migraine criteria of the International Classification of Headache Disorders, first edition (ICHD-I) (18). Participants were classified as having any headache when they answered “yes” to the screening question “Have you suffered from headache during the last 12 months” and otherwise classified as headache free. Headache that did not fulfill the criteria of migraine was classified as non-migrainous. Headache diagnoses (migraine and non-migrainous headache) were mutually exclusive, meaning that each person could get only one headache diagnosis. Participants were classified as migraine if they reported having migraine in the questionnaire or fulfilled the following three criteria: (1) headache-attacks lasting four to 72 hours (<4 hours were accepted for those who reported commonly occurring visual disturbance before headache); (2) headache with at least one of the following three characteristics: pulsating quality, unilateral location or aggravation by physical activity; (3) during headache, at least one of the following occurred: nausea, photophobia or phonophobia. Migraine commonly preceded by visual disturbances prior to headache was classified as migraine with aura.

The validity of these questionnaire-based diagnoses has been reported previously in a study comparing diagnoses made through clinical interview with the questionnaire-based diagnoses (19). For any headache, the sensitivity were 85% and specificity 83% (kappa value 0.57); the corresponding figures for migraine were 69% and 89% (kappa value 0.59), for migraine with aura, 50% and 100% (kappa 0.64); and for non-migrainous headache, 61% and 81% (kappa 0.43). In our validation study, 80% of individuals with non-migrainous headache suffered from tension-type headache (19).

Causes of death

The National Cause of Death Registry in Norway, on the basis of the unique 11-digit person identification number, allowed a complete follow-up. Data on mortality were obtained from the database and were available from date of attendance to HUNT2 until December 2011 with a mean follow-up of 14.1 years (SD 3.4). In this study mortality was categorized into three main groups: all-cause death, cardiovascular death (International Statistical Classification of Diseases and Related Health Problems, 10th revision (ICD-10) code I00-99), and non-vascular death. The term cardiovascular death was used both when the immediate cause and the underlying cause of death were of cardiovascular nature, the latter increasing the number of cardiovascular deaths in the analysis from 3788 to 4231. Individuals with death caused by diseases of a cardiovascular nature were also subdivided into three groups: death by cerebrovascular disease (ICD-10 code I60-I69), death by ischemic heart disease (IHD) (ICD-10 code I20-I25), and deaths by other diseases of a cardiovascular nature (ICD-10 code I00-I15, I26-I52 and I70-I99).

Potential confounders

Education was categorized according to duration: ≤9 years, 10–12 years or ≥13 years. Responses to questions on physical activity were categorized according to duration and intensity of exercise per week: high (three or more hours of hard physical activity per week (i.e. being out of breath and/or sweating)), medium (one to two hours of hard physical activity per week), and low (less than one hour of hard physical activity per week or any form of light physical activity) (20). Cigarette smoking was categorized as “current daily smoking,” “previous daily smoking” or “never daily smoking.” Alcohol use was categorized in two groups based on CAGE, as “no positive CAGE-questions” or “1≥positive CAGE-questions” (21). Body mass index (BMI) was calculated and categorized as <25 kg/m², 25–29.9 kg/m² or ≥30 kg/m². Anxiety and depression were assessed by the Hospital Anxiety and Depression Scale (HADS) (22). Total cholesterol, triglycerides, and blood pressure were also considered among the potential confounders.

Ethics

HUNT2 was approved by the National Data Inspectorate and the study was also approved by the regional committee for ethics in medical research.

Statistical analysis

Cox’s proportional hazards regression model was used to estimate the hazard ratios (HRs) of mortality comparing headache sufferers with those without headache. Headache status was the major explanatory variable, and for each category of headache (the categories were “migraine” and “non-migrainous headache”), the risk of mortality was estimated by comparing those who were headache free as the reference category. The HR of mortality due to cardiovascular causes was then evaluated in more detail, by stratifying the analysis for gender, headache subtype (migraine with aura, migraine without aura and non-migrainous headache) and headache frequency (<7 days/month and ≥7 days/month). The cutoff of seven days/month was preferred rather than ≥15 days/month because very few headache sufferers reported chronic headache. Analyses were initially adjusted for age (continuous variable) and gender, and subsequently for other potential confounders; education (three categories), self-reported cardiovascular disease (yes/no), BMI (three categories), physical activity (three categories), smoking (three categories), alcohol use (two categories), total HADS (three categories), metabolic syndrome (yes/no), total cholesterol (continuous variable), triglycerides (continuous variable), and systolic BP (continuous variable). Individuals with incomplete data for one or several variables (number stated in Table 1) were nevertheless included (as a separate missing category) in all analyses to reduce the impact of response bias. Cox regression modeling was also used to test for possible interactions of age, gender, and smoking habits with headache categories in the prediction of mortality. Statistical significance was set at p < 0.05. The statistical software used for analysis was IBM SPSS, version 21.

Table 1.

Demographic and clinical characteristics related to headache status in HUNT2 (headache free, migraine and non-migrainous headache).

Variables	Headache free (n = 31,737)	Migraine (n = 6831)	Non-migrainous headache (n = 13,285)
Mean age, years (SD)	52.6 (17.6)	44.0 (13.9)	46.2 (15.4)
Gender, female (%)	46.8	72.4	61.9
Education > 12 years (%) (missing^a= 2165)	19.4	22.6	20.6
Self-reported CVD (%)	11.7	4.6	7.6
Mean BMI (SD)	26.4 (3.9)	26.1 (4.3)	26.3 (4.2)
Physical activity, low (%) (missing^a= 4694)	64.4	69.4	68.1
Daily smoking (%) (missing^a= 906)	25.4	30.7	30.6
Alcohol (one positive CAGE-question) (%) (missing^a= 12,760)	14.3	12.6	15.3
HADS anxiety, severe (%) (missing^a= 7867)	2.5	8.5	6.6
HADS depression, severe (%) (missing^a= 4079)	2.1	4.3	3.5
Metabolic syndrome (%)	18.8	16.3	17.7
Mean systolic BP (SD)	140.1 (22.1)	132.0 (19.2)	134.1 (20.2)

Missing values are given as numbers of individuals when not zero. SD: standard deviation; CVD: cardiovascular disease; BMI: body mass index, CAGE: acronym for the four questions used to screen for alcohol abuse; HADS: Hospital Anxiety and Depression Scale: BP: blood pressure.

Results

The distribution of clinical and demographic baseline characteristics of the participants by headache status is presented in Table 1. During a mean follow-up of 14.1 years, a total of 9408 individuals died (5031 men, 4377 women). Of these, 4321 had died from cardiovascular diseases (2339 men, 1982 women). As shown in Table 2, there was no difference in all-cause mortality, cardiovascular mortality or non-cardiovascular mortality between individuals with migraine or non-migrainous headache compared to the group of headache-free controls. No major differences in HRs were found between age/gender- adjusted analyses, and analyses adjusting for all important confounders (Table 2).

Table 2.

Age-gender adjusted and complete^a adjustment hazard ratios (HR) (95% confidence intervals) for mortality from all causes, from cardiovascular disease and from all other causes among 51,853 participants in HUNT2 (1995–1997).

		All causes of mortality		Cardiovascular mortality		Mortality from all other causes
	Total	n	HR (95% CI)	n	HR (95% CI)	n	HR (95% CI)
Headache free	31,737	7255	1.00 Reference	3423	1.00 Reference	3832	1.00 Reference
Migraine, age-gender adjusted	6831	560	0.96 (0.86–1.04)	204	0.89 (0.77–1.02)	356	1.01 (0.90–1.13)
complete adjustment^a			0.94 (0.86–1.02)		0.88 (0.76–1.01)		0.98 (0.88–1.10)
Migraine with aura, age-gender adjusted	920	75	1.24 (0.98–1.56)	27	1.20 (0.82–1.76)	48	1.28 (0.96–1.70)
complete adjustment^a			1.18 (0.93–1.48)		1.17 (0.80–1.72)		1.21 (0.91–1.61)
Migraine without aura, age-gender adjusted	5911	485	0.92 (0.84–1.01)	177	0.85 (0.73–0.99)	308	0.98 (0.87–1.10)
complete adjustment^a			0.91 (0.83–1.00)		0.85 (0.73–0.97)		0.95 (0.85–1.07)
Non-migrainous headache, age-gender adjusted	13,285	1593	0.99 (0.94–1.05)	694	1.01 (0.93–1.10)	899	0.99 (0.92–1.07)
complete adjustment^a			0.95 (0.90–1.01)		0.95 (0.87–1.03)		0.96 (0.89–1.03)

Complete adjustment: Adjusted for age, gender, education, self-reported cardiovascular disease. BMI, physical activity, smoking, alcohol use, total HADS, metabolic syndrome and systolic BP.

BMI: body mass index; HADS: Hospital Anxiety and Depression Scale; BP: blood pressure; CI: confidence interval.

The analyses with complete adjustment showed a reduced cardiovascular mortality rate among men with migraine without aura, evident for those with migraine less than seven days per month (HR 0.71, 95% confidence interval (CI) 0.52–0.96, Table 3). A reduced mortality was evident only for IHD (HR 0.56, 95% CI 0.38–0.84), not for cerebrovascular diseases (HR 0.93, 95% CI 0.58–1.49). There was no such relationship in men between cardiovascular mortality and migraine with aura. In addition, there was no significant association evident among women with migraine with or without aura.

Table 3.

Hazard ratio (HR) for cardiovascular death related to headache type among 51,853 participants in HUNT2 (1995–1997).

		Cardiovascular mortality
		Both genders		Men		Women
Variable	Total	n	HR (95% CI)	n	HR (95% CI)	n	HR (95%CI)
Adjusted for age and gender
Headache free	31737	3423	1.00 Reference	1971	1.00 Reference	1452	1.00 Reference
Migraine	6831	204	0.89 (0.77–1.02)	68	0.74 (0.58–0.94)	136	1.03 (0.86–1.23)
Migraine with aura	920	27	1.20 (0.82–1.76)	7	0.85 (0.41–1.79)	20	1.49 (0.96–2.32)
Migraine without aura	5911	177	0.85 (0.73–0.99)	61	0.73 (0.56–0.94)	116	0.98 (0.80–1.18)
Non-migrainous headache	13285	694	1.01 (0.93–1.10)	300	0.95 (0.84–1.07)	394	1.09 (0.98–1.22)
Complete adjustment ^a
Headache free	31,737	3423	1.00 Reference	1971	1.00 Reference	1452	1.00 Reference
Migraine with aura	920	27	1.17 (0.80–1.72)	7	0.88 (0.42–1.87)	20	1.39 (0.89–2.17)
<7 headache/days/month	680	17	1.07 (0.66–1.72)	5	0.81 (0.34–1.96)	12	1.26 (0.71–2.23)
≥7 headache/days/month	240	10	1.40 (0.75–2.62)	2	1.09 (0.27–4.41)	8	1.72 (0.86–3.47)
Migraine without aura	5911	177	0.85 (0.73–0.97)	61	0.72 (0.56–0.93)	116	0.97 (0.80–1.18)
<7 headache/days/month	4440	120	0.79 (0.66–0.95)	42	0.71 (0.52–0.96)	78	0.89 (0.71–1.12)
≥7 headache/days/month	1471	57	0.99 (0.76–1.29)	19	0.76 (0.48–1.20)	38	1.21 (0.87–1.67)
Non-migrainous headache	13,285	694	0.95 (0.87–1.03)	300	0.91 (0.80–1.03)	394	1.01 (0.90–1.14)
<7 headache/days/month	10,971	508	0.97 (0.88–1.07)	225	0.99 (0.86–1.14)	283	0.97 (0.86–1.11)
≥7 headache/days/month	2314	186	0.90 (0.78–1.05)	75	0.73 (0.58–0.92)	111	1.13 (0.93–1.38)

Complete adjustment: Adjusted for age, gender, education, self-reported cardiovascular disease. BMI: body mass index; HADS: Hospital Anxiety and Depression Scale; BP: blood pressure; CI: confidence interval.

Discussion

In the present study there was no significant association between migraine and all-cause or cardiovascular mortality when compared to those without headache. There was, however, moderately reduced risk for cardiovascular mortality in men with low headache frequency migraine without aura, evident only for IHD.

The present results are in accordance with the most recent meta-analysis, which did not find evidence for an effect of any migraine on mortality overall, or mortality from cardiovascular or coronary heart disease (16). These results contrast with the results from an Icelandic study that reported higher all-cause and cardiovascular mortality in people suffering from migraine with aura (10). However, the Icelandic study population was different from the present study in that the fraction of individuals with aura was 69% compared to only 13% in the HUNT2 population. Thus one cannot exclude the possibility that this is the reason for the discrepancy between these two studies. There are several studies that have linked migraine with aura, but not migraine without aura, to increased mortality (10,11,23). On the other hand, there are some studies that found significantly lower mortality in persons with migraine (12,13), while in some studies there was only a tendency that did not reach statistical significance (14,15). A study of migraine in men reported increased risk of major cardiovascular events and myocardial infarctions, although the risk of ischemic cardiovascular death was not increased (HR 1.07, CI 0.80–1.43) (24). Similar results were found in a study in women with insignificant changes in mortality (25). Our findings are more in line with these studies, as the point estimate of the HR is 0.94 for all-cause mortality, with the upper limit of the 95% CI barely above one.

We found that migraine without aura was associated with a reduced overall cardiovascular mortality, and mortality caused by IHD. This negative association was evident for men with low headache frequency. The reason is unclear, but we cannot rule out the possibility of positive statistical significant findings occurring by chance alone because multiple comparisons were performed. However, if the result is true, several explanations may be suggested. Firstly, if men with migraine used acetylsalicylic acid for pain relief, they may benefit from a cardioprotective effect from this medication. Secondly, men with migraine may be more likely to seek medical attention than men not suffering from headache, increasing the probability of an early cardiovascular diagnosis. Thirdly, men with infrequent migraine could be forced to adopt a more regular lifestyle (e.g. regular sleep pattern and less alcohol overuse (26)) than headache-free men, decreasing the mortality risk. The reason why this appears to be stronger for men than women is unclear.

The major strengths of this study were the population-based design with a mean follow-up time of 14 years, mortality data on all participants, the large sample size with a wide age range, and the use of validated headache diagnoses (19). Furthermore, the analysis was adjusted for several important confounding factors, although the possibility of residual confounding by an unrecognized factor cannot be ruled out.

However, some weaknesses should also be mentioned. Firstly, the relatively low number of events in each subgroup makes it difficult to make strong conclusions with regard to subgroup-specific effects, especially for migraine with aura and for those with headache ≥7 days/month. Secondly, only 56% of the invited individuals answered the headache questions and results should be generalized with caution. Nonparticipants had higher mortality than participants (27). They also had a higher prevalence of comorbid conditions, and probably also a higher prevalence of headache and migraine (27). However, it is difficult to predict whether a participation rate of 100% would have changed our results, and in which direction.

Conclusion

In this large, prospective study participants with migraine did not experience a higher all-cause mortality or cardiovascular mortality than individuals without headache. This knowledge should give reassurance to migraine sufferers who fear that their attacks could represent a health hazard in the long run.

Clinical implications

No evidence was found for a higher all-cause mortality or cardiovascular mortality among individuals with migraine.

This research may give reassurance to migraine sufferers who fear that their attacks could represent a health hazard in the long run.

Footnotes

Acknowledgments

The Nord-Trøndelag Health Study (The HUNT Study) is a collaboration between HUNT Research Centre (Faculty of Medicine, Norwegian University of Science and Technology NTNU), Nord-Trøndelag County Council, Central Norway Health Authority, and the Norwegian Institute of Public Health.

This study was conceived of by KH. ANÅ and KH performed the statistical analysis. ANÅ, LJS, JAZ, BSW, IH and KH all participated in drafting the manuscript. All authors have read and approved the final manuscript.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Conflict of interest

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Stovner

Hagen

Jensen

. The global burden of headache: A documentation of headache prevalence and disability worldwide. Cephalalgia 2007; 27: 193–210.

Linde

Stovner

Zwart

. Time trends in the prevalence of headache disorders. The Nord-Trondelag Health Studies (HUNT 2 and HUNT 3). Cephalalgia 2011; 31: 585–596.

Steiner

Stovner

Birbeck

. Migraine: The seventh disabler. J Headache Pain 2013; 14: 1–1.

Etminan

Takkouche

Isorna

. Risk of ischaemic stroke in people with migraine: Systematic review and meta-analysis of observational studies. BMJ 2005; 330: 63–63.

Scher

Terwindt

Picavet

. Cardiovascular risk factors and migraine: The GEM population-based study. Neurology 2005; 64: 614–620.

Schurks

Rist

Bigal

. Migraine and cardiovascular disease: Systematic review and meta-analysis. BMJ 2009; 339: b3914–b3914.

Bigal

Kurth

Santanello

. Migraine and cardiovascular disease: A population-based study. Neurology 2010; 74: 628–635.

Winsvold

Hagen

Aamodt

. Headache, migraine and cardiovascular risk factors: The HUNT study. Eur J Neurol 2011; 18: 504–511.

Winsvold

Sandven

Hagen

. Migraine, headache and development of metabolic syndrome: An 11-year follow-up in the Nord-Trøndelag Health Study (HUNT). Pain 2013; 154: 1305–1311.

10.

Gudmundsson

Scher

Aspelund

. Migraine with aura and risk of cardiovascular and all cause mortality in men and women: Prospective cohort study. BMJ 2010; 341: c3966–c3966.

11.

Kurth

Gaziano

Cook

. Migraine and risk of cardiovascular disease in women. JAMA 2006; 296: 283–291.

12.

Hall

Brown

. Triptans in migraine: The risks of stroke, cardiovascular disease, and death in practice. Neurology 2004; 62: 563–568.

13.

Waters

Campbell

Elwood

. Migraine, headache, and survival in women. Br Med J (Clin Res Ed) 1983; 287: 1442–1443.

14.

Velentgas

Cole

. Severe vascular events in migraine patients. Headache 2004; 44: 642–651.

15.

Cook

Benseñor

Lotufo

. Migraine and coronary heart disease in women and men. Headache 2002; 42: 715–727.

16.

Schurks

Rist

Shapiro

. Migraine and mortality: A systematic review and meta-analysis. Cephalalgia 2011; 31: 1301–1314.

17.

Hagen

Zwart

Vatten

. Prevalence of migraine and non-migrainous headache—head-HUNT, a large population-based study. Cephalalgia 2000; 20: 900–906.

18.

Classification and diagnostic criteria for headache disorders, cranial neuralgia, and facial pain. Headache Classification Committee of the International Headache Society. Cephalalgia 1988; 8(Suppl 7): 1–96.

19.

Hagen

Zwart

Vatten

. Head-HUNT: Validity and reliability of a headache questionnaire in a large population-based study in Norway. Cephalalgia 2000; 20: 244–251.

20.

Varkey

Hagen

Zwart

. Physical activity and headache: Results from the Nord-Trøndelag Health Study (HUNT). Cephalalgia 2008; 28: 1292–1297.

21.

Skogen

Overland

Knudsen

. Concurrent validity of the CAGE questionnaire. The Nord-Trøndelag Health Study. Addict Behav 2011; 36: 302–307.

22.

Mykletun

Stordal

Dahl

. Hospital Anxiety and Depression (HAD) scale: Factor structure, item analyses and internal consistency in a large population. Br J Psychiatry 2001; 179: 540–544.

23.

Liew

Wang

Mitchell

. Migraine and coronary heart disease mortality: A prospective cohort study. Cephalalgia 2007; 27: 368–371.

24.

Kurth

Gaziano

Cook

. Migraine and risk of cardiovascular disease in men. Arch Intern Med 2007; 167: 795–801.

25.

Ahmed

Bairey Merz

McClure

. Migraines, angiographic coronary artery disease and cardiovascular outcomes in women. Am J Med 2006; 119: 670–675.

26.

Aamodt

Stovner

Hagen

. Headache prevalence related to smoking and alcohol use. The Head-HUNT Study. Eur J Neurol 2006; 13: 1233–1238.

27.

Langhammer

Krokstad

Romundstad

. The HUNT study: Participation is associated with survival and depends on socioeconomic status, diseases and symptoms. BMC Med Res Methodol 2012; 12: 143–143.

Migraine as a predictor of mortality: The HUNT study

Abstract

Background

Methods

Results

Conclusion

Keywords

Introduction

Methods

Participants

Headache diagnoses

Causes of death

Potential confounders

Ethics

Statistical analysis

Results

Discussion

Conclusion

Clinical implications

Footnotes

Acknowledgments

Funding

Conflict of interest

References