Sage Journals: Discover world-class research

Abstract

Objective

To investigate whether cluster headache (CH) was a risk factor for depression in a nationwide population-based follow-up study.

Background

There are few studies about the relationship between CH and depression, and prior research has been limited by cross-sectional studies or small sample sizes.

Methods

We identified 673 CH patients from the Taiwan National Health Insurance database between 2005 and 2009. The two comparison cohorts included age-, sex- and Charlson’s score-matched migraine patients (n = 2692) and controls (patients free from migraine or CH, n = 2692). The cumulative incidence of depression was compared among these three cohorts until the end of 2009. We also calculated predictors of depression in the CH cohort.

Results

After the median 2.5-year follow-up duration, the CH cohort had a greater risk for developing depression compared to the control cohort (adjusted hazard ratio; aHR = 5.6, 95% CI 3.0–10.6, p < 0.001) but not the migraine cohort (aHR = 1.1, 95% CI 0.7–1.7, p = 0.77). Of the CH patients, the number of cluster bout periods per year was a risk factor for depression (aHR = 3.8, 95% CI 2.6–5.4, p < 0.001).

Conclusion

Our results showed that CH is associated with an increased risk for depression. The strength of this association is similar to that of migraine.

Keywords

Cluster headache depression incidence migraine risk Taiwan

Introduction

Cluster headache (CH), once described as “suicide headache,” is an uncommon primary headache associated with specific pathophysiology and severe pain. The lifetime prevalence of CH is approximately 124/100,000 people (1). In addition to head discomfort, recent literature suggests that CH may be associated with psychiatric and non-psychiatric comorbidities (2 –4).

Depression is a common psychiatric comorbidity in headache disorders (5 –7). For example, migraine patients are at an increased risk for depression (7 –11), and a bidirectional relationship between migraine and depression suggests that these two diseases share similar mechanisms (11). Recent advances also show shared genetics between migraine and depression patients (12,13). Furthermore, migraine patients with depression are at a risk for migraine chronification and poor treatment outcomes (14,15).

In contrast to the well-researched relationship between migraine and depression, the association between CH and depression is not well studied. In addition, results are equivocal (2 –5,16 –21). Although most studies reveal that CH patients experience more depressive symptoms (3 –5,16 –18), some research shows only a few CH patients developing depression (2,19 –21). However, self-administered instruments, not psychiatrist-based interviews, were used for depression diagnoses in these studies, which may account for some of the inconsistent results. Regardless of diagnostic tools, much of the previous research in this area has relied on cross-sectional, short-term, or small-sized samples (20). Therefore, we conducted a nationwide population-based cohort follow-up study to explore the association between CH and depression.

Patients and methods

Data source

The current study used data from the National Health Insurance Research Database (NHIRD) of Taiwan. In 1995, a national health insurance program was launched in Taiwan; the enrollment is mandatory, and the coverage was approximately 98% by the end of 2002 and 99.6% by the end of 2011. In 1999, the Bureau of National Health Insurance began to release patient data in electronic form under the NHIRD project. Consequently, various extracted datasets are available to researchers, and up to now, hundreds of published papers have used the NHIRD for their studies. The diagnoses of several diseases, such as diabetes mellitus, acute coronary syndrome and stroke, as well as the accuracy of the comorbidity by Charlson’s score in the NHIRD, have been validated (22 –26). The NHIRD consists of de-identified secondary data released for research purposes; therefore, the present study was exempt from full review by the Institutional Review Board. The diseases were coded by the International Classification of Diseases (ICD)-9-CM codes, 2001 edition.

Study design and patient selection

We conducted a nationwide retrospective cohort study featuring three cohorts: CH, migraine, and control. Any patient diagnosed with an antecedent psychiatric disorder (ICD-9-CM codes 290.x to 319.x; equivalent ICD-10 codes F30–F39) was excluded from analyses.

Figure 1 shows subject and cohort selection. Briefly, patients diagnosed with CH (ICD-9-CM, code 346.2; equivalent ICD-10 code G44.0) from January 1, 2005, to December 31, 2009, comprised the CH cohort. CH diagnosis was further validated if it was coded by neurologists and presented with the standard CH prescription prophylactic drug, verapamil (27,28). The index date referred to the initial CH diagnosis given within the study period. In this study, clinical visits for CH that occurred within three months were considered as the same cluster period. According to the criteria from the International Classification of Headache Disorders, second edition (ICHD-2), chronic CH was diagnosed when the attacks of CH occur for more than a year without a pain-free remission of at least one month (29). Therefore, in our study, CH patients having consecutive clinical visits and medication refills for more than one year, without a minimum one-month gap interval based on the outpatient department (OPD) records, were given a chronic CH diagnosis (29).

Figure 1.

Patient selection.

Patients diagnosed with migraine (ICD-9-CM codes 346.x except for 346.2; equivalent ICD-10 codes G43.X), as coded by neurologists from January 1, 2005, to December 31, 2009, comprised the migraine cohort.

The control cohort was selected from a dataset containing 1,000,000 beneficiaries who were randomly sampled from the original NHIRD. There were no significant differences in age and gender distribution of patients in this sample of 1,000,000 beneficiaries and the original NHIRD. Patients diagnosed with either migraine or CH (ICD-9-CM codes 346.x) were excluded from the control cohort.

Due to significant differences in age, sex, and disease comorbidity among the three cohorts, we refined the criteria for our migraine and control cohorts by randomly selecting age (±two years), sex, index date, and Charlson’s score (± one score)-matched patients (four migraine and four control patients for each CH patient) based on the diagnostic index date. Each patient in the three cohorts was followed from the index date until December 31, 2009, death, or any depression diagnosis (ICD-9-CM codes 296.2X, 296.3X, 300.4, and 311; equivalent ICD-10 codes F32.X, 33.X and 34.2). Depression diagnosis was included only when coded by psychiatrists.

Statistical analysis

Normally distributed continuous data were expressed as means ± standard deviations. Numeric data with non-normal distributions were expressed as medians and interquartile ranges (IQRs). Pearson χ² tests were conducted on categorical variables; the independent t-test and Mann-Whitney U test were used for parametric and nonparametric continuous variables, respectively. The Poisson test examined depression risks between the different cohorts, and the Kaplan Meier method calculated the cumulative incidence of depression. Depression risks were compared between the cohorts by the log-rank test. In subgroup analyses, Cox proportional hazards models computed the hazard ratios (HRs) of potential risk factors accompanying 95% confidence intervals (CIs) after adjusting for confounding variables among CH patients. All probabilities were two tailed, and statistical significance was determined by p < 0.05. The SPSS statistical software program (version 20.0; IBM) conducted all analyses.

Results

Study population

Figure 1 and Table 1 illustrate subject sampling and cohort characteristics. There were 673 CH patients, including 19 chronic CH patients (2.8%), 2692 migraine patients, and 2692 control patients finally enrolled for analysis. Study subjects were predominantly male (79.9%) and the median age was 35.8 years (IQR, 29.4–43.8 years). The frequencies of pre-existing depression were similar between migraine and CH patients (9711/180,060; 5.4% vs. 39/777; 5.0%, p = 0.645).

Table 1.

Baseline patient characteristics.

	All patients with CH	Matched patients with migraine		Matched control patients
	n = 673	n = 2692	p value^a	n = 2692	p value^a
Age (IQR)	35.8 (29.4–43.8)	36.0 (29.6–44.1)	0.983	35.8 (29.3–43.7)	0.943
Follow-up, days	815 (341–1313)	889 (462–1304)	0.951	792 (341–1252)	0.381
Number of bouts of CH per year	0.63 (0.31–1.35)	—		—
Male	79.9% (n = 538)	79.9% (n = 2152)	1.000	79.9% (n = 2152)	1.000
Charlson’s score			0.993		0.932
Score 0	67.2% (n = 452)	67.2% (n = 1808)		67.1% (n = 1806)
Score 1–2	27.0% (n = 182)	27.2% (n = 731)		26.7% (n = 720)
Score≧3	5.8% (n = 39)	5.7% (n = 153)		6.2% (n = 166)

CH: cluster headache; IQR: interquartile range.

Compared with CH cohort.

Antidepressant use

We also screened the prescriptions for antidepressants within 30 days before the index date in all three cohorts. In our study, 10 patients (five migraine, five CH, 0 control) were prescribed antidepressants with a cumulative defined daily dosage of more than 10. Of these patients, tricyclic antidepressants were used in six patients (four migraine, two CH), trazodone in two (one migraine, one CH), selective serotonin reuptake inhibitor (SSRI), i.e. fluoxetine in one patient (migraine) and mirtazapine in one patient (CH).

Depression incidence

The median follow-up period was 815 days (IQR, 341–1313 days). Across the study cohort, the risk of developing depression during the observational period was 2.1% (130/6057), and the corresponding risks for the CH, migraine, and control cohorts during the follow-up period were 3.6% (24/673), 3.3% (90/2692), and 0.6% (16/2692), respectively. Figure 2 illustrates the cumulative incidence rates of depression over time. The CH cohort had a higher incidence rate of depression than the control group (15.6 vs. 2.7 per 1000 person-years, p < 0.001), but not compared to the migraine cohort (15.6 vs. 13.6 per 1000 person-years, p = 0.275). In addition, the median duration between the last recorded clinical visit for CH and the first depression diagnosis was 297 days (IQR, 70–747). Although chronic CH patients appeared to have a higher incidence rate than episodic CH patients, this difference was not statistically significant (11.8% vs. 6.4%, p = 0.102).

Figure 2.

Cumulative rates for developing depression.

Predictors for depression

Table 2 reports the adjusted HRs for depression by cohort. The CH cohort had a greater risk for developing depression compared to the control cohort (adjusted HR; aHR = 5.6, 95% CI 3.0–10.6, p < 0.001) but not to the migraine cohort (aHR = 1.1, 95% CI 0.7–1.7, p = 0.77). The migraine cohort also carried a higher risk for depression compared to the control cohort (aHR = 5.4, 95% CI 3.2–9.2, p < 0.001). In subgroup analyses, compared to the control cohort, female CH subjects displayed a stronger association with depression (aHR 16.8, 95% CI 3.6–79.5, p < 0.001). However, compared to migraine patients, younger (age < 36 years) CH patients had a higher risk of developing depression, regardless of sex. Besides, CH patients still had a higher risk for developing depression even though only the diagnostic codes for major depression (296.2x, 296.3x) were counted for analysis (data not shown).

Table 2.

Crude and adjusted hazard ratio for depression for CH and comparison groups.

	CH group	Control group	Compared with control				Migraine group	Compared with migraine
	CH group	Control group	Crude		Adjusted^a		Migraine group	Crude		Adjusted^a
	Event/no. of patient	Event/no. of patient	HR (95% CI)	p value	HR (95% CI)	p value	Event/no. of patient	HR (95% CI)	p value	HR (95% CI)	p value
All patients	24/673	16/2692	5.7 (3.0–10.8)	<0.001	5.6 (3.0–10.6)	<0.001	90/2692	1.1 (0.7–1.8)	0.549	1.1 (0.7–1.7)	0.776
Male	16/538	14/2152	4.2 (2.1–8.6)	<0.001	4.0 (2.0–8.3)	<0.001	55/2152	1.3 (0.8–2.3)	0.328	1.3 (0.8–2.3)	0.323
Female	8/135	2/540	16.2 (3.4–76.1)	<0.001	16.8 (3.6–79.5)	<0.001	35/540	0.9 (0.4–1.9)	0.770	0.8 (0.4–1.7)	0.541
Age >36	12/335	8/1338	5.7 (2.3–14.0)	<0.001	5.8 (2.4–14.1)	<0.001	72/1357	0.7 (0.4–1.3)	0.284	0.6 (0.3–1.2)	0.152
Age <36	12/338	8/1354	5.7 (2.3–14.0)	<0.001	5.5 (2.2–13.5)	<0.001	18/1335	2.8 (1.4–5.9)	0.005	3.0 (1.4–6.2)	0.004

CH: cluster headache; HR: hazard ratio; CI: confidence interval.

Adjusted for age, gender, and Charlson’s score.

Within the CH cohort, Cox proportional regression hazard models showed that the annual number of cluster bout periods predicted a higher risk for depression (aHR 3.8, 95% CI 2.6–5.4) (Table 3).

Table 3.

Multivariate Cox regression of depression risk factors in CH patients.

	Univariate			Multivariate
	HR	95% CI	p value	HR	95% CI	p value
Age≧36	0.88	0.40–1.96	0.755
Male	0.55	0.24–1.29	0.171
Charlson’s score	0.49	0.21–1.14	0.097
Diabetes	0.05	0.00–920.06	0.544
Cancer	3.30	0.44–24.64	0.244
Hypertension	1.26	0.43–3.69	0.674
Chronic pulmonary disease	0.86	0.20–3.65	0.834
Cerebral vascular disease	0.80	0.11–5.96	0.832
Peptic ulcer disease	0.04	0.00–6.41	0.213
Liver disease	0.04	0.00–28.81	0.344
Number of CH attacks per year	3.71	2.56–5.38	<0.001	3.77	2.63–5.42	<0.001

CH: cluster headache; CI: confidence interval.

Discussion

This study reports the following key findings: About 3.6% CH patients developed depression during a mean follow-up period of 2.5 years, with the incidence rate of 15.6 per 1000 person-years. CH patients were 5.6 times more likely to develop depression compared to control patients; however, depression incidence rates were similar between CH and migraine cohorts. Among CH patients, the number of cluster periods per year predicted the development of depression.

A major strength of this study was its population-based design, which relied on a comprehensive medical utilization claim database that included all migraine and CH patients diagnosed 2005–2009 (NHIRD). In addition, the large sample size offered enough power for appropriate statistical analyses. Moreover, this study recruited CH patients coded by neurologists, and we included only patients who were prescribed verapamil, the standard preventive medication (27). In Taiwan, flunarizine is the most common calcium channel blocker for migraine prophylaxis, whereas verapamil is prescribed for CH patients (28). Another strength of the current research was that our study demographics (i.e. male predominance, similar mean age, low chronic CH prevalence) were similar to those reported in other Asian studies (30,31). In addition, we controlled Charlson’s scores in order to prevent potential confounding effects due to comorbid chronic diseases. Finally, we are confident in the validity of depression diagnosis because we included only psychiatrist-coded medical records and recruited all ICD-9-CM codes about depression, including mild dysthymia to major depression. In fact, if only major depression (296.2x, 296.3x) were analyzed, the results were similar.

It is well known that migraine is a risk factor for depression (7 –11); however, the precise underlying mechanisms of this relationship remain unknown. In addition to etiology and genetic theories, research also suggests that depression is a direct consequence of headache suffering and migraine chronification (6,32 –34). In our study, the depression incidence rates between CH patients and migraine subjects were similar. In fact, these two groups also displayed similar frequencies of pre-existing depression rates (migraine 5.4%; CH 5.0%). This comparable risk for depression in CH and migraine cohorts further supports our finding that CH was associated with a higher risk for depression. Of note, our results revealed that younger CH patients were at a greater risk for depression than age-matched migraine patients. The effect of age on depression in CH and migraine patients needs further evaluation.

Although the mechanism underlying the association between depression and CH is unclear, there are several hypotheses that may explain this comorbidity. First, one observational study attributed depression to a prodromal symptom of CH, which may be related to diffuse cerebral disturbance or the response to severe pain or sleep interference (35). Second, positron emission tomography (PET) data suggest that depression in CH individuals may be associated with insular cortex activation or prefrontal hypometabolism; both are structures commonly associated with pain and depression processing (18,36). A third possible mechanism may be hypothalamic dysfunction (37) because CH and depression share common chronobiological features, such as episodic or seasonal attacks and have similar responses to melatonin or melatonergic agonist treatment (25,38). A final explanation for the relationship may be that headache suffering and chronification facilitate depression in some CH patients. Similar to migraine patients, chronic CH patients have higher depression prevalence than episodic CH patients (2,16). In our cohort, the median duration of latency from the assumed last CH attack to depression diagnosis was approximately 10 months, suggesting that most CH patients developed depression outside of a headache incident. Besides, most of our CH patients were episodic rather than chronic. Our findings indicate that depression in CH cannot be completely explained by pain and physiologic change during the headache bout and/or chronification of headache.

Importantly, our results have several clinical implications. First, CH patients should be evaluated for depression both initially and at follow-up because they are at risk for developing the disorder. Furthermore, depression may occur beyond a specific CH attack. Second, clinicians should target episodic CH patients who experience frequent headache incidences for the diagnosis of depression. Finally, our results showed that female CH patients had a greater risk for developing depression. This finding is in line with a recent study (39) and suggests that additional research is needed to evaluate the influence of sex on the CH-depression relationship; however, clinicians should be aware of potential sex differences in depression development.

Some methodological issues should be addressed here. First, the diagnoses of depression and migraine CH in patients are often delayed, so the exact temporal or causal relationship should be interpreted cautiously. Therefore, our study results could suggest only an increased association of depression in patients with CH rather than a causal relationship. Second, we acknowledge that the NHIRD dataset is an administrative database that lacks thorough clinical data, such as disease severity and laboratory results. In addition, there may be deviations in the claim data, as physicians’ coding may purposefully correspond to their prescriptions or other factors. Despite these potential drawbacks, the data remain representative, as the Bureau of NHI routinely and randomly samples a fixed percentage of claims from every contracted medical institution. To further minimize information bias, we used a matched control cohort to balance any possible coding errors. Third, 10 patients within our cohorts were prescribed antidepressants 30 days before the index date. We could not completely exclude the possibility that some of them did have depression but did not receive a corresponding diagnostic coding by their physicians. However, certain antidepressants were also prescribed for insomnia, pain treatment or headache prevention. Fourth, we did not consider cohort sociodemographic differences (i.e. urbanization or monthly income), which may be related to depression occurrence. Finally, there was no corresponding ICD-9-CM code for CH subtypes until the 2009 edition. To address this issue, we divided CH patients into episodic and chronic subtypes based on their frequency of clinic visits to neurologists. However, it is possible that a patient with chronic CH might be misclassified to have episodic CH if he or she did not regularly visit the clinics or his or her follow-up duration was less than one year.

In conclusion, this population-based follow-up study showed an association between CH and increased depression risks. This risk was similar to that in migraine patients. Our findings have several clinical implications, and future longitudinal studies should be conducted to delineate mechanisms underlying comorbid depression.

Clinical implications

Cluster headache (CH) patients are associated with a 5.6 times greater risk for depression.

The strength of this association is similar to that of migraine.

Female CH patients have a greater risk for developing depression.

Footnotes

Funding

This study was supported in part by grants from the Taiwan National Science Council (98-2314-B-010-019-MY2, NSC 100-2628-E-010-002-MY3), Taipei Veterans General Hospital (VGHUST101-G7-1-1, V101C-106, V101E7-003), NSC support for Centre for Dynamical Biomarkers and Translational Medicine, National Central University, Taiwan (NSC100-2911-I-008-001), Brain Research Center, National Yang-Ming University, Tri-Service General Hospital (TSGH-C101-159) and a grant from the Ministry of Education, Aim for the Top University Plan.

Conflict of interest

None declared.

References

Fischera

Marziniak

Gralow

. The incidence and prevalence of cluster headache: A meta-analysis of population-based studies. Cephalalgia 2008; 28: 614–618.

Robbins

Bronheim

Lipton

. Depression and anxiety in episodic and chronic cluster headache: A pilot study. Headache 2012; 52: 600–611.

Rozen

Fishman

. Cluster headache in the United States of America: Demographics, clinical characteristics, triggers, suicidality, and personal burden. Headache 2012; 52: 99–113.

Donnet

Lanteri-Minet

Guegan-Massardier

. Chronic cluster headache: A French clinical descriptive study. J Neurol Neurosurg Psychiatry 2007; 78: 1354–1358.

Jelinski

Magnusson

Becker

. Factors associated with depression in patients referred to headache specialists. Neurology 2007; 68: 489–495.

Deligianni

Vikelis

Mitsikostas

. Depression in headaches: Chronification. Curr Opin Neurol 2012; 25: 277–283.

Breslau

Schultz

Stewart

. Headache and major depression: Is the association specific to migraine? Neurology 2000; 54: 308–313.

Breslau

Davis

Schultz

. Migraine and major depression: A longitudinal study. Headache 1994; 34: 387–393.

Jette

Patten

Williams

. Comorbidity of migraine and psychiatric disorders – a national population-based study. Headache 2008; 48: 501–516.

10.

Modgill

Jette

Wang

. A population-based longitudinal community study of major depression and migraine. Headache 2012; 52: 422–432.

11.

Breslau

Lipton

Stewart

. Comorbidity of migraine and depression: Investigating potential etiology and prognosis. Neurology 2003; 60: 1308–1312.

12.

Marino

Fanny

Lorenzi

. Genetic bases of comorbidity between mood disorders and migraine: Possible role of serotonin transporter gene. Neurol Sci 2010; 31: 387–391.

13.

Ligthart

Nyholt

Penninx

. The shared genetics of migraine and anxious depression. Headache 2010; 50: 1549–1560.

14.

Lanteri-Minet

Radat

Chautard

. Anxiety and depression associated with migraine: Influence on migraine subjects’ disability and quality of life, and acute migraine management. Pain 2005; 118: 319–326.

15.

Lipton

Diamond

. Migraine diagnosis and treatment: Results from the American Migraine Study II. Headache 2001; 41: 638–645.

16.

Jurgens

Gaul

Lindwurm

. Impairment in episodic and chronic cluster headache. Cephalalgia 2011; 31: 671–682.

17.

Robbins

. The prevalence of anxiety and depression in cluster headache patients (abstract). Headache 2004; 44: 507–507.

18.

Seifert

Valet

Pfaffenrath

. Neurometabolic correlates of depression and disability in episodic cluster headache. J Neurol 2011; 258: 123–131.

19.

Ertsey

Manhalter

Bozsik

. Health-related and condition-specific quality of life in episodic cluster headache. Cephalalgia 2004; 24: 188–196.

20.

Jorge

Leston

Arndt

. Cluster headaches: Association with anxiety disorders and memory deficits. Neurology 1999; 53: 543–547.

21.

Gesztelyi

Bereczki

. Disability is the major determinant of the severity of depressive symptoms in primary headaches but not in low back pain. Cephalalgia 2005; 25: 598–604.

22.

Lin

Lai

Syu

. Accuracy of diabetes diagnosis in health insurance claims data in Taiwan. J Formos Med Assoc 2005; 104: 157–163.

23.

Chan

. Histamine-2-receptor antagonists are an alternative to proton pump inhibitor in patients receiving clopidogrel. Gastroenterology 2010; 139: 1165–1171.

24.

Cheng

Kao

Yang

. Validation of claims record on acute myocardial infarction and ischemic stroke in the National Health Insurance Research Database in Taiwan. In: ISPE third Asian Meeting, Seoul, Korea, 3–5 November 2008, oral presentation. Bethesda, MD, USA: ISPE.

25.

Cheng

Kao

Lin

. Validation of the National Health Insurance Research Database with ischemic stroke cases in Taiwan. Pharmacoepidemiol Drug Saf 2011; 20: 236–242.

26.

Chang

. A study of validation on comorbidity derived from claims data. Master thesis, National Yang-Ming University, Taiwan, 2004.

27.

Francis

Becker

Pringsheim

. Acute and preventive pharmacologic treatment of cluster headache. Neurology 2010; 75: 463–473.

28.

Treatment Guideline Subcommittee of the Taiwan Headache Society. Treatment guidelines for acute and preventive treatment of cluster headache. Acta Neurol Taiwan 2011; 20: 213–227.

29.

Headache Classification Subcommittee of the International Headache Society. The international classification of headache disorders, 2nd edition. Cephalalgia 2004; 24(Suppl 1): 45–45.

30.

Lin

Wang

Fuh

. Cluster headache in the Taiwanese – a clinic-based study. Cephalalgia 2004; 24: 631–638.

31.

Imai

Yagi

Kuroda

. Clinical profile of cluster headaches in Japan: Low prevalence of chronic cluster headache, and uncoupling of sense and behaviour of restlessness. Cephalalgia 2011; 31: 628–633.

32.

Antonaci

Nappi

Galli

. Migraine and psychiatric comorbidity: A review of clinical findings. J Headache Pain 2011; 12: 115–125.

33.

Buse

Manack

Serrano

. Sociodemographic and comorbidity profiles of chronic migraine and episodic migraine sufferers. J Neurol Neurosurg Psychiatry 2010; 81: 428–432.

34.

Gentili

Panicucci

Guazzelli

. Psychiatric comorbidity and chronicisation in primary headache. J Headache Pain 2005; 6: 338–340.

35.

Blau

Engel

. Premonitory and prodromal symptoms in cluster headache. Cephalalgia 1998; 18: 91–93.

36.

Sprenger

Ruether

Boecker

. Altered metabolism in frontal brain circuits in cluster headache. Cephalalgia 2007; 27: 1033–1042.

37.

Costa

Leston

Cavallini

. Cluster headache and periodic affective illness: Common chronobiological features. Funct Neurol 1998; 13: 263–272.

38.

Olié

Kasper

. Efficacy of agomelatine, a MT1/MT2 receptor agonist with 5-HT2C antagonistic properties, in major depressive disorder. Int J Neuropsychopharmacol 2007; 5: 661–673.

39.

Rozen

Fishman

. Female cluster headache in the United States of America: What are the gender differences? Results from the United States Cluster Headache Survey. J Neurol Sci 2012; 317: 17–28.

Cluster headache is associated with an increased risk of depression: A nationwide population-based cohort study

Abstract

Objective

Background

Methods

Results

Conclusion

Keywords

Introduction

Patients and methods

Data source

Study design and patient selection

Statistical analysis

Results

Study population

Antidepressant use

Depression incidence

Predictors for depression

Discussion

Clinical implications

Footnotes

Funding

Conflict of interest

References