The Prevalence and Incidence of Cluster Headache Authors: Einar Naveen Møen, Marte-Helene Bjørk, Nils Erik Gilhus, Dagrun Slettebø Daltveit, Daniela Contreras, Knut Hagen, Mahendran Jayaraman, Andrej Netland Khanevski, Tore Wergeland, Anker Stubberud, Erling Tronvik, Jannicke Igland Published: 2026 ABSTRACT Background and ObjectivesData on time trends in cluster headache epidemiology are sparse. The aim of this study was to report trends in prevalence and incidence of cluster headache in Norway over a 14-year period.MethodsWe conducted a registry-based study using linked data from the Norwegian Registry for Primary Health Care, the Norwegian Control and Payment of Health Reimbursements Database, the Norwegian Patient Registry, the Norwegian Prescribed Drug Registry, and Statistics Norway from 2009 to 2022. Data included diagnostic codes, prescriptions, and education. Adults (age ≥18 years) were included. Cluster headache prevalence was defined as ≥2 contacts (clinical consults or prescriptions) for cluster headache in a 365-day period. Age-standardized trends in prevalence and incidence by sex and year, and interactions between education and year, were analyzed with negative binomial regression. We estimated prevalence rate ratio (PRR) and incidence rate ratio per calendar year with 95% CIs.ResultsThe number of patients with cluster headache increased from 1,029 in 2009 (median age 44 years; 39.7% women) to 1,833 patients in 2022 (median age 47 years; 50.1% women). The annual age-standardized prevalence rate increased from 27.0 to 42.5 per 100,000 in the same period. Women had a 3-fold higher annual increase of 6% (PRR 1.06, 95% CI 1.05–1.07) compared with 2% (PRR 1.02, 95% CI 1.02–1.03) in men. The prevalence rate was higher in women than in men by 2022 (43.4 vs 41.7 per 100,000). The annual prevalence of chronic cluster headache and refractory chronic cluster headache varied between 6%–7% and 1%–2% of all cluster headache cases, respectively. The annual age-standardized incidence rate of cluster headache increased in women, from 10.1 to 14.6 per 100,000 from 2012 to 2022 and decreased in men, from 13.5 to 11.0 per 100,000. Incidence and prevalence rates were higher among individuals with lower education.DiscussionPrevalence increased over 14 years, possibly reflecting improved diagnostic practices and awareness. These findings challenge previous reports of cluster headache predominantly affecting men, illustrating distinct shifts and trends in disease epidemiology. A limitation was the lack of clinical validation of cluster headache diagnostic codes in primary health care. Introduction Cluster headache is a rare and debilitating primary headache disorder characterized by intense unilateral pain around the eye or temple, accompanied by cranial autonomic symptoms, restlessness, or both. Cluster headache attacks occur every other day up to 8 times a day and last 15–180 minutes. Attack periods last from weeks to months.1 Episodic cluster headache is characterized by pain-free intervals of more than 3 months in a 12-month period.1 Chronic cluster headache is characterized by continuous pain-free intervals of less than 3 months.1 People with chronic cluster headache, who do not respond adequately to traditional preventive treatments, have refractory chronic cluster headache.2 Prevalence estimates of cluster headache vary notably, ranging from 3 to 150 per 100,000, possibly because of geographical differences, timing of data collection, and methodological variation.3 Longitudinal data on cluster headache prevalence and incidence are sparse.4,5 Estimates suggest that 10%–20% of patients with cluster headache have the chronic variant, but large-scale studies are lacking.3 The prevalence of refractory chronic cluster headache is unknown.6 Recent studies indicate a shift in demographics for cluster headache. One study reported an increased prevalence among women, whereas another observed a rising overall prevalence.4,7 These findings suggest demographic and temporal changes in the cluster headache population. Therefore, our aim was to determine changes in the prevalence and incidence of cluster headache from 2009 to 2022 in Norway using nationwide registry data and to compare different subgroups defined by sex, age, and educational levels. This has implications for health care planning and resource allocation. Methods Study Design This study is a registry-based nationwide study including all individuals ≥18 years of age who were registered as residents in Norway at any time point during 2008–2022. Setting Health care in Norway is a tax-funded system providing social security and universal services. This enables detailed registries of routinely collected data on diagnoses, prescriptions, and demographics. Individuals with chronic diseases, defined as needing treatment for at least 3 months per year, receive reimbursement for their medications once they have reached a modest annual out-of-pocket payment cap. Primary care physicians report diagnoses and symptoms using the International Classification of Primary Care, second edition (ICPC-2) coding system. Most specialist care neurologists work in hospitals and, to a lesser extent, in private practice as contracted specialists. Diagnoses in specialist care are reported using the International Classification of Diseases, 10th Revision (ICD-10) coding system. Materials Data on diagnoses from all consultations in the primary and specialist health care services for the period 2008–2022 were obtained from the Norwegian Registry for Primary Health Care, the Norwegian Control and Payment of Health Reimbursements Database, and the Norwegian Patient Registry.8-10 Data on prescriptions filled from pharmacies, including Anatomical Therapeutic Chemical (ATC) classification code for the prescribed drug, amount, date of prescription fills, and ICD-10 codes and ICPC-2 codes for indication were obtained from the Norwegian Prescribed Drug Registry for the period 2008–2022.11 This registry includes data on all filled prescriptions from all Norwegian pharmacies but does not provide information on over-the-counter purchases or medications administered in nursing homes or hospitalizations. Data on age, death, emigration, educational attainment, and sex were obtained from Statistics Norway.12 We acknowledge that gender identity may not always align with terms in clinical research. In this study, we used registry-recorded sex in the absence of gender data. Patients in Norway may change their sex from the sex assigned at birth in the population registry. When referring more generally, we use the term patients with cluster headache, recognizing that diversity exists among patients with the disorder. All registries contained person-level information and were subsequently linked using national identity numbers that were replaced with a unique individual code to ensure anonymity. Outcome Definitions A contact for cluster headache was defined as either (1) a consultation in primary health care with a registered diagnosis of cluster headache (ICPC-2: N90); (2) an outpatient visit or hospitalization in specialist health care with a diagnosis of cluster headache (ICD-10: G44.0); or (3) a filled prescription for cluster headache-specific triptans (defined as sumatriptan [ATC: N02CC01] or zolmitriptan [ATC: N02CC03], with a reimbursement code for cluster headache). For each calendar year, an individual was defined as a prevalent case of cluster headache if they had 2 or more contacts for cluster headache within the past 365 days. The prevalence date was defined as the first date when the criteria for prevalence were fulfilled (eFigure 1). Individuals had to fulfill the diagnostic criteria again each calendar year to be classified as prevalent cases for that specific year. We identified prevalent cases from 2009 to 2022 to ensure that each individual had at least 365 days of available observation time, because registry data were available from 2008. Triptans for cluster headache are predominantly administered subcutaneously or intranasally in Norway, and only prescriptions for these formulations are reimbursed. For each calendar year, an incident case of cluster headache was defined as a case fulfilling the abovementioned criteria, but without being classified as a cluster headache case in the previous 4 years. To ensure that all cases had a 4-year washout period available and that only new cases of cluster headache were included, we defined incident cases from and including 2012. The accuracy of a diagnostic code for cluster headache was found to be 79% (95% CI 74%–85%) in 4 departments of neurology in central Norway.13 Diagnoses from the Norwegian Registry for Primary Health Care have not been validated. A prevalent case of cluster headache was classified as either chronic or episodic based on the total amount of defined daily dosages for filled prescriptions of sumatriptan or zolmitriptan over a year. A defined daily dose is defined as the assumed average maintenance dose per day for a drug used for its main indication in adults.14 This was used as a proxy variable for the number of attacks. Individuals filling prescriptions of 137 or more defined daily doses of cluster headache triptans in the year before the prevalence date were classified as having chronic cluster headache.14 This definition translates to the clinical criteria for chronic cluster headache of having ≥0.5 attacks per day in a period of at least 9 months ([365/12] × 9 × 0.5).1 Those not fulfilling this criterion were classified as having episodic cluster headache. To identify persons with refractory chronic cluster headache, we used the definition in the consensus statement from the European Headache Federation of having ≥3 attacks per week despite trials of ≥3 preventive treatments.2 This definition was operationalized by examining the number of prescribed defined daily doses of cluster headache triptans in the 365-day period before the prevalence date of each year. This was used to estimate the annual number of attacks, where ≥157 defined daily dosages were considered ≥3 attacks per week ([3/7] × 365). In our assessment of the number of attempted preventive treatments, we only counted treatments having a “consensus statement” level of evidence in the European Academy of Neurology guideline for cluster headache: verapamil (ATC: C08DA01), lithium (ATC: N05AN01), topiramate (ATC: N03AX11), ergotamine tartrate (ATC: N02CA52), frovatriptan (ATC: N02CC07), valproic acid (ATC: N03AG01), and melatonin (ATC: N05CH01).15 Individuals with cluster headache in Norway are currently not reimbursed for their preventive treatments because cluster headache is an off-label indication for these medications. Because clinicians may prescribe these medications under alternative approved indications to ensure patient reimbursement, we counted prescriptions of these medications regardless of the registered reimbursement code. One prescription in a fixed 5-year period was counted as 1 medication trial. This assessment of previous medication trials was only used to identify potential patients with refractory chronic cluster headache. To ensure that all cases had 5 years of available data, we defined refractory cases from and including 2013. Statistical Analysis Age-standardized prevalence rates and incidence rates were calculated using the total Norwegian population of 2022 in 1-year age groups as a standard population. Because of overdispersion and a poor goodness-of-fit with Poisson regression, we used negative binomial regression to investigate change in prevalence and incidence over time. A natural logarithm link function with counts of individuals in the population in each sex and 1-year age group was used as an offset variable in each analysis to model rates. Sex, age groups (<40, 40–59, and ≥60 years), and educational attainment were used as stratification variables to identify differences in incidence and prevalence and potential difference in patterns over time. The interaction between year and educational attainment was analyzed to quantify changes over time. Educational attainment was based on the Norwegian Standard Classification of Education, which is a Norwegian version of the International Standard Classification for Education.16 Low education was defined as only compulsory education (10 years). Medium education was defined as high school or preuniversity education (11–14 years). Higher education was defined as college, university, or postgraduate education (15 or more years). Given that educational attainment is typically completed by age 25, analyses involving education were restricted to individuals aged 25 years and older to ensure stable and meaningful classification. Age and year of diagnosis were included as covariates to adjust for differences in age distribution within each age stratum and to evaluate time trends in prevalence and incidence, respectively. Prevalence rate ratios (PRRs) and incidence rate ratios (IRRs) were calculated by exponentiating the coefficients for year, representing the relative change in prevalence or incidence proportion per 1-year increment, assuming a log-linear trend. All analyses were conducted in R version 4.3.0 and STATA version 18.17-32 No imputation methods were used because of minimal missing data. Sensitivity Analyses We performed multiple sensitivity analyses to assess potential sources of bias. We first explored how the total prevalence changed when we used different data sources and definitions of cluster headache prevalence. We calculated 3 separate prevalence estimates using (1) ≥2 diagnostic codes from primary care, (2) ≥2 diagnostic codes from specialist health care data, and (3) ≥2 reimbursement codes for cluster headache-specific triptans. Second, we assessed the number of patients with ≥2 diagnostic codes in primary care who had no such diagnosis in specialist care. Additional analyses were conducted to characterize this population: (1) the number of people who never received a cluster headache diagnosis in specialist care, (2) the number of people who had 3 or more cluster headache diagnosis codes in primary care, (3) the number of people who had received prescriptions for triptans with reimbursement for cluster headache, (4) the number of people who had received preventive drug therapies (steroids not included) for cluster headache, and (5) the number of people with at least 1 migraine diagnosis code. Ad Hoc Analyses We performed an ad hoc analysis to investigate possible underlying mechanisms of our findings. Using the same criteria as for cluster headache prevalence, we investigated the prevalence of (1) migraine and (2) trigeminal neuralgia during the same recording period. Standard Protocol Approvals, Registrations, and Patient Consents Ethical approval was granted by the Regional Ethical Committee in Western Norway (REK-ID: 198928). Data Availability The data analyzed in this study contain individual patient information and cannot, therefore, be shared. Similar data sets are available from the registry holders, following appropriate regulatory approvals and registry data applications.33 Results Prevalence The Norwegian population aged ≥18 years increased from 3,694,473 in 2009 to 4,316,662 in 2022. The number of patients with cluster headache increased from 1,029 in 2009 to 1,833 in 2022. The median age of patients with cluster headache was 44 years (interquartile range [IQR] 35–55) in 2009, which increased to 47 years (IQR 35–57) in 2022. The percentage of women increased from 39.7% in 2009 to 50.1% in 2022. Correspondingly, the overall age-standardized prevalence rate of cluster headache increased from 27.0 to 42.5 per 100,000 from 2009 to 2022 (Figure 1; Table 1). Women and men demonstrated different trends in prevalence rates. For men, the prevalence rate increased from 32.4 to 41.7 per 100,000 during the study period, with an annual average increase of 2% (PRR 1.02, 95% CI 1.02–1.03). The prevalence rate in women increased from 21.6 to 43.4 per 100,000 during the study period, with an average annual increase of 6% (PRR 1.06, 95% CI 1.05–1.07) (eTable 1). Women had a higher prevalence rate than men for the first time in 2021 (43.5 compared with 42.4 per 100,000). The prevalence was highest in persons aged 40–59 years in both sexes. Men in this age group had the highest prevalence rate early in the study period (Figure 2). This group had only a moderate average annual increase of 1% (PRR 1.01, 95% CI 1.00–1.02). Women aged 40–59 years surpassed them in 2019, demonstrating a 7% annual prevalence increase (PRR 1.07, 95% CI 1.06–1.07) (Figure 2; eTable 2). Age-Standardized Prevalence and Incidence Rates of Cluster Headache by Sex, Norway 2009–2022 Age-standardized prevalence rates (upper panel, 2009–2022) and incidence rates (lower panel, 2012–2022) of cluster headache per 100,000 population by sex. Rates were standardized to the Norwegian population in 2022. Table 1Annual Age-Standardized Prevalence and Incidence Rates of Cluster Headache Subtypes by Sex, Norway 2009–2022YearTotal prevalence rateEpisodic prevalence rateChronic prevalence rateRefractory prevalence rateTotal incidence rateMenWomenTotalMenWomenTotalMenWomenTotalMenWomenTotalMenWomenTotal200932.421.627.029.920.125.12.51.42.0 (7.2)NANANANANANA201035.223.029.233.121.627.42.11.41.8 (6.0)NANANANANANA201134.323.629.032.322.027.22.01.51.8 (6.1)NANANANANANA201235.123.229.333.021.727.52.11.61.8 (6.3)NANANA13.510.111.8201336.927.532.334.625.930.42.21.61.9 (6.0)0.30.40.3 (1.1%)14.012.713.4201437.729.633.735.627.731.72.11.92.0 (5.9)0.30.40.4 (1.0%)12.413.813.1201537.830.934.435.129.032.12.71.92.3 (6.7)0.20.50.4 (1.1%)12.113.012.5201641.232.837.038.530.934.72.71.92.3 (6.2)0.30.60.4 (1.2%)13.513.413.4201741.135.838.538.633.436.02.52.42.4 (6.3)0.20.60.4 (1.0%)12.414.813.6201844.836.840.841.834.238.03.02.62.8 (6.9)0.51.00.7 (1.8%)14.714.214.5201944.043.143.541.140.740.92.92.42.6 (6.1)0.41.20.8 (1.8%)13.318.215.7202042.841.542.140.138.639.32.82.92.8 (6.7)0.50.90.7 (1.7%)13.115.114.1202142.443.542.839.841.240.42.62.32.4 (5.7)0.50.70.6 (1.4%)12.615.814.2202241.743.442.539.240.940.02.52.62.5 (5.9)0.50.60.5 (1.2%)11.014.612.8Age-standardized rates per 100,000 people for total, episodic, chronic, and refractory chronic cluster headache prevalence, and total cluster headache incidence. Percentages in parentheses represent the proportion of chronic and refractory cases among all cluster headache cases. Annual Age-Standardized Prevalence and Incidence Rates of Cluster Headache Subtypes by Sex, Norway 2009–2022 Age-standardized rates per 100,000 people for total, episodic, chronic, and refractory chronic cluster headache prevalence, and total cluster headache incidence. Percentages in parentheses represent the proportion of chronic and refractory cases among all cluster headache cases. Age-Standardized Prevalence and Incidence Rates of Cluster Headache by Sex and Age Group, Norway 2009–2022 Age-standardized prevalence rates (upper panels, 2009–2022) and incidence rates (lower panels, 2012–2022) of cluster headache per 100,000 stratified by sex (men, left; women, right) and age groups (<40 years, blue; 40–59 years, red; ≥60 years, green). Rates were standardized to the Norwegian population from 2022. When investigating prevalence rates by educational attainment, we found the highest prevalence rates in patients with low education in both sexes (Figure 3). Men with low education had 2.08 times higher prevalence of cluster headache than men with high education (PRR 2.08; 95% CI 1.95–2.22). Women with low education had 1.64 times higher prevalence of cluster headache than women with high education (PRR 1.64, 95% CI 1.53–1.76) (eTable 3). Men with low or medium education had an average annual increase of 2% (PRR 1.02, 95% CI 1.01–1.03), whereas those with high education demonstrated an average annual increase of 3% (PRR 1.03, 95% CI 1.02–1.04). Annual increases in prevalence rates were notably higher in women. Women with low education had an annual increase in prevalence rate of 5% (PRR 1.05, 95% CI 1.04–1.06), whereas those with medium or high education had an annual increase of 7% (PRR 1.07, 95% CI 1.06–1.08), and 7% (95% CI 1.05–1.08) (Figure 3; eTable 4). Educational disparities within the cluster headache group remained stable over time for men but narrowed significantly over time for women with low compared with high education (p = 0.004) (eTable 5). Age-Standardized Prevalence and Incidence Rates of Cluster Headache by Sex and Educational Attainment, Norway 2009–2022 Age-standardized prevalence rates (upper panels, 2009–2022) and incidence rates (lower panels, 2012–2022) of cluster headache per 100,000 stratified by sex (men, left; women, right) and educational attainment (low education, blue; medium education, red; high education, green) restricted to individuals ≥25 years of age. Rates were standardized to the 2022 Norwegian population. Incidence The overall age-standardized incidence rate of cluster headache fluctuated from 11.8 per 100,000 per year in 2012, to 15.7 per 100,000 per year in 2019, and to 12.8 per 100,000 in 2022 (Figure 1; Table 1). Women and men demonstrated different patterns. The incidence in women increased from 10.1 per 100,000 per year in 2012 to 14.6 per 100,000 per year in 2022. The incidence decreased in men, from 13.5 per 100,000 per year in 2012, to 11.0 per 100,000 per year in 2022 (Table 1). From 2019 onward, the incidence rate for women was higher than for men (Figure 1; Table 1). The total incidence rates of cluster headache in men remained relatively stable during the observation period (IRR 0.99, 95% CI 0.98–1.00) (eTable 1). We found differences when examining age groups within each sex. The incidence in women increased significantly across all groups. The increase was highest among those <40 years and those ≥60 years, with an average annual increase in incidence of 4% (IRR 1.04, 95% CI 1.02–1.06 and 95% CI 1.01–1.07, respectively). By contrast, the incidence remained stable or slightly declining in men across the different age groups (eTable 6). Incidence rates were highest in the low educational attainment (Figure 3). The incidence decreased in men with low and medium educational attainment and increased in men with high educational attainment. Women with medium and high educational attainment had increasing incidence rates during the entire study period. Incidence in women with low educational attainment increased until 2018 and decreased from 2019 onward. By 2020, women with medium education had surpassed those with low education, and by 2022, women with high education also exceeded those with low education (Figure 3). Interaction testing confirmed that these patterns and revealed significant changes in educational disparities over time (eTable 7). Narrowing of differences in incidence rates were observed between low and high educational attainment in both sexes (p = 0.040 and <0.001, respectively) and between medium and high educational attainment in men (p = 0.011) (eTable 7). Chronic and Refractory Cluster Headache The age-standardized prevalence rate of chronic cluster headache increased from 2.0 per 100,000 in 2009 to 2.5 per 100,000 in 2022. However, the proportion of chronic cluster headache within the cluster headache population decreased from 7.2% to 5.9% (Table 1). The average annual increase in prevalence of chronic cluster headache was largest in the oldest age groups. Men ≥60 years demonstrated an average annual increase of 8% (PRR 1.08, 95% CI 1.05–1.12), whereas women ≥60 years had an average annual increase of 9% (PRR 1.09, 95% CI 1.04–1.15) (eTable 8). The age-standardized prevalence rate of refractory chronic cluster headache increased from 0.3 per 100,000 (1.1% of all cluster headache cases) in 2013 to 0.5 per 100,000 (1.2%) in 2022 and was largest in women, especially among those between 40 and 59 years where refractory cluster headache had an average annual increase of 12% (PRR 1.12, 95% CI 1.02–1.23) (eTable 9). Sensitivity Analyses We found differences in prevalence when applying different definitions. However, the prevalence increased over time with all definitions. We found an increase from 5.5 to 10.2 per 100,000 when we only counted prescriptions from 2009 to 2022. Similarly, we found an increase from 18.9 to 24.6 per 100,000 when we only counted primary care contacts and an increase from 3.6 to 10.3 per 100,000 when we only counted specialist care contacts (eTable 10; eFigure 2). A total of 5,802 patients had ≥2 cluster headache codes from primary care within a <365-day period during the study period. In this population, 3,232 patients (55.7%) never received a cluster headache diagnosis in specialist care. Among these 3,232, a total of 2,266 (70.1%) had ≥3 cluster headache diagnosis codes in primary care. Furthermore, 528 (16%) of the 3,232 had received a prescription for triptan with reimbursement for cluster headache, and 338 (10%) had received a prescription for cluster headache preventive therapy (steroids not included) during the study period (eTable 11). In the subgroup of 3,232 patients with ≥2 cluster headache diagnostic codes from primary care and no confirmatory diagnostic code in specialist care, a total of 2,379 (74%) fulfilled at least one of the cluster headache indicators above (≥3 diagnosis codes, triptan prescription, or preventive prescription) (eTable 11). In this subgroup, 1,412 (43.7%) had received at least 1 diagnosis code for migraine during the study period. Ad Hoc Analyses When applying the same prevalence criteria for cluster headache to migraine and trigeminal neuralgia, we observed substantial increases in both conditions from 2009 to 2022. Migraine prevalence increased from 697.8 to 1401.9 per 100,000 (+100.9%). Trigeminal neuralgia increased from 45.9 to 65.5 per 100,000 (+42.7%). In comparison, cluster headache prevalence increased from 27.0 to 42.5 per 100,000 (+57.4%). Prevalence was higher in women than in men across all headache conditions (eTable 12, eFigure 3). Discussion The prevalence of cluster headache in Norway increased from 0.03% to 0.04% over 14 years. We observed a dramatic change in the demographic profile during this period. Although initially higher in men, the prevalence shifted to predominance in women during the study period. A higher prevalence was observed in patients with low educational attainment. Chronic cluster headache and refractory cluster headache made up 6%–7% and 1%–2%, respectively, of all cluster headache cases. We found a substantial increase over time in the prevalence of cluster headache, with a shift in gender distribution from predominance in men to near-equal representation. The incidence of cluster headache among women was particularly high around 2019. Across the full study period, the largest increases in incidence were observed in women <40 and ≥60 years. Similar patterns in women have been reported previously, raising the possibility of a hormonal component in cluster headache disease onset in women, such as menopause.34,35 Prevalence rates were higher in those with low education, indicating an association between education and cluster headache onset. This may be explained by lifestyle factors such as smoking, a known causative factor for cluster headache, and other factors associated with socioeconomic disadvantages.36 We believe that the increase in prevalence observed in our study reflects improved diagnostic practices, including new targeted treatments and increased awareness among both clinicians and patients. Ad hoc analyses demonstrated increases in both migraine and trigeminal neuralgia prevalence (eTable 12, eFigure 3). A similar trend across different headache types supports our hypothesis of increased diagnostics as the primary driver. However, minor contributions from unmeasured environmental factors cannot be entirely excluded. Some of the observed increase could also be attributed to reduced sex-based diagnostic stereotypes, similar to acute myocardial infarction.37 Onabotulinumtoxin A was approved for chronic migraine in Europe in 2011.38 Erenumab was released in 2018 in Norway as the first calcitonin gene-related peptide–related preventive therapy for migraine in the Norwegian market.39 Some of the observed increase in cluster headache prevalence may reflect increased referrals of women with presumed chronic migraine to specialist care for assessment of eligibility for new treatment options, where specialist assessment resulted in a diagnosis or codiagnosis of cluster headache. A meta-analysis from 2008 found a 1-year cluster headache prevalence of 53 per 100,000 (95% CI 26–95) and a higher prevalence in men.3 This finding is comparable with our observed prevalence of 42.5 per 100,000 at the end of the study period, although with predominance among women. There is a paucity of reports on changes in cluster headache prevalence in the same region over time. An Italian study from 2021 analyzing a large prescription database identified a minor increase of cluster headache prevalence from 2013 to 2017 of 6.4 to 6.7 per 100,000 (+4.7%). Increased diagnostic awareness, including awareness of pharmacologic management, was proposed as an explanation.4 When examining prevalence defined as prescriptions only in our material, we found a larger increase from 8.0 to 8.8 per 100,000 (+10.0%) in the same period (eTable 10). This demonstrates comparable trends in 2 European countries when using similar methods. However, prescription-based data represent a substantial underestimate of the true prevalence, as demonstrated in our main analyses. The prevalence of chronic cluster headache reflected the overall pattern of cluster headache prevalence—initially higher prevalence in men, but highest in women at the end of the study period. The proportion of chronic cluster headache among all cluster headache cases remained relatively stable at around 6%–7% of total cluster headache cases. A meta-analysis from 2008 of studies using questionnaire surveys and medical record reviews found that chronic cluster headache account for 14.3% of all cluster headache cases.3 Three studies from East Asia from 2019 to 2024 report prevalence between 2.33% and 4.8%.40-42 Although our results are closer to East-Asian estimates, they are likely on the conservative side because we did not have data on oxygen therapy or opioid usage. A Swedish study from 2023 provided important data on gender differences in persons with cluster headache. Women with cluster headache reported longer bouts and more frequent usage of prophylactic treatment and had a higher proportion of chronic cluster headaches compared with men.35 The prevalence of refractory chronic cluster headache in our study was numerically higher in women compared with men throughout the entire study period (0.6 vs 0.5 per 100,000 in 2022). This supports the notion that the disease burden of cluster headache differs in men and women. The major strength of our study is the use of mandatory data from 5 nationwide Norwegian registries. This enables us to capture most cases of cluster headache across the entire Norwegian health care system. The most important weakness of our study is the uncertainty on the specificity of our cluster headache definition. Sensitivity analyses revealed variations in prevalence estimates depending on the applied criterion. This could indicate an overestimation of prevalence because half of the patients diagnosed in primary care never received a cluster headache diagnosis in specialist care, and there is some uncertainty regarding diagnostic validity in primary care. On the contrary, 74% in this subgroup had either ≥3 diagnostic codes or drug use indicating cluster headache. This illustrates the necessity of including multiple data sources, such as primary care, to avoid underestimating the prevalence of cluster headache. Although the definitions used can affect overall frequency estimates, it is unlikely that this would influence the sex-specific patterns and overall trends we observed. We did not have data on oxygen therapy use. This limitation is unlikely to have affected our ability to identify prevalent cases, as access to oxygen therapy requires diagnostic codes to which we had access. It may, however, have affected our analysis of cluster headache subgroup prevalence. Triptan prescriptions do not capture all attacks, as oxygen therapy is a commonly used acute treatment either alone or in combination.15 This could have resulted in an underestimation of both chronic cluster headache and refractory chronic cluster headache prevalence. Using triptan prescriptions as an attack proxy implies a risk of missing the distinction between episodic and chronic cluster headache. Patients with episodic cluster headache and a very high triptan consumption during attack periods with a high frequency of attacks per day may have been misclassified as having chronic cluster headache. Patients with cluster headache in Norway are not reimbursed for their preventive treatments.43 When defining refractory cluster headache, we, therefore, included all prescriptions of the included drugs, regardless of reimbursement codes. This could have resulted in an overestimation of refractory chronic cluster headache prevalence if some individuals with a cluster headache diagnosis used the preventive drugs for other indications. Because individuals in our study population may have tried different preventive treatments without a net positive effect before the study period began, the actual prevalence of refractory disease could be higher than found by us. Nationwide real-world mandatory data reduce selection bias and attrition, thereby enhancing the generalizability of our findings. However, because our case identification was based on diagnostic codes from a publicly funded health care system, our findings might be less representative of different health care settings, demographics, ethnicities, and latitudes.44 In addition, regional variations in established risk factors such as smoking may limit generalizability across populations. Because smoking is an established risk factor of cluster headache, our findings from Norway, where smoking is increasingly infrequent (less than 10% in 2021), probably represent conservative estimates compared with populations with higher smoking prevalence.36 This nationwide registry-based cohort study revealed a substantial increase in cluster headache prevalence from 2009 to 2022, accompanied by a notable shift from predominance in men to predominance in women. Incidence increased in women and remained stable in men. Both prevalence and incidence varied substantially by age group and educational level, with higher rates in individuals aged 40–59 years and among those with lower educational attainment. Chronic cluster headache and refractory chronic cluster headache accounted for 6%–7% and 1%–2% of all cluster headache cases, respectively. The observed increase may be related to increased diagnostic awareness and reduced sex-based diagnostic stereotypes.