Prevalence of Migraine and Other Types of Primary Headache in China: A Systematic Review and Meta-Analysis Open Access

Primary headaches, such as migraine, are believed to impose significant financial burdens on societies globally. Migraine alone accounted for 45.1 million years lived with disability worldwide in the global burden of disease of 2016 and ranked among the top 10 most burdensome diseases [1, 2]. Tension-type headache (TTH), another form of primary headache, is even more prevalent than migraine. The troublesome symptoms and comorbidities associated with these primary headaches have a substantial impact on sufferers’ social and academic functioning [3]. In a systematic review, Leonardi et al. [4] determined that the global prevalence of migraine was 14.0%. Additionally, a study revealed that the economic burden of migraine was substantial, comparable to that of diabetes and exceeding that of asthma [2]. Despite the significant burden and prevalence of primary headaches, such as migraine, they remain under-recognized and inadequately treated [5].

Epidemiological surveys play a crucial role in estimating disease burden through measures such as incidence and prevalence, informing public health policies for prevention and treatment. In the last 3 decades, numerous cross-sectional epidemiological surveys have been carried out to investigate migraine and other primary headache disorders in the Chinese population [6‒9]. However, due to variations in study duration and methodology, the findings of these studies have been inconsistent. A recent review by Stark et al. [10] focused on the prevalence of chronic migraine in the Asian-Pacific region yet did not encompass all epidemiological surveys conducted in China. As a result, epidemiological data on migraine and other primary headache disorders in China remain incomplete. This study aimed to conduct a systematic review and meta-analysis to determine the prevalence of migraine and other primary headache disorders among the Chinese population.

We searched for literatures published from January 1988 to December 2023 in Chinese Wanfang, CNKI (Chinese National Knowledge Infrastructure), PubMed, and MEDLINE databases, using the following terms “primary headache”, “migraine”, “tension-type headache (TTH)”, “cluster headache”, “trigeminal autonomic cephalalgias” “prevalence”, “incidence”, and “epidemiology”. We also reviewed the additional references for potentially eligible researches. This systematic review and meta-analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [11].

Two authors (Y.Z. and C.L.) independently reviewed the title, abstract, and full texts of relevant studies to determine their eligibility. Eligible studies fulfilled the following inclusion criteria: (1) study data being general population of the Chinese communities; (2) providing the accurate study dates and sufficient data on prevalence or incidence; (3) diagnosis of any primary headache based on International Classification of Headache Disorders (ICHD) criteria (any edition); (4) the most detailed study of repeated studies in the same population. There was no limitation in language. The studies were excluded if they were conducted in a hospital-based population, or included participants with concomitant diseases (e.g., epilepsy) or specific occupations. Additionally, articles that did not provide accurate epidemiologic data, such as meta-analyses, case reports, reviews, brief or oral communications, etc., were also excluded.

Quality of all eligible studies was assessed and data were extracted independently by two reviewers (Y.Z. and C.L.). A third investigator (N.C.) was involved if there were any disagreements. We used STROBE standard guideline (Strengthening the Reporting of Observation Studies in Epidemiology) with 22 criteria [12] to assess the quality of literatures. Total scores ranged from 0 to 22 and satisfying one criterion counted one point. Overall scores greater than or equal to 16 points (≥16) were considered as high quality, 10∼15 points were considered as moderate quality, and less than 10 points were considered as low quality. Information was extracted as follows: first author, year of publication, diagnostic criteria, data collection dates, geographical region, sample size, proportion of males and females, and prevalence of primary headache.

We calculated prevalence estimate with the Freeman-Tukey variance-stabilizing double arcsine transformation. The pooled prevalence of migraine and TTH were calculated separately. I² and Q test were used to assess the heterogeneity between studies, and I² ≥ 50% or p ≤ 0.10 indicated significant heterogeneity [13]. We used random-effects models for summary of prevalence estimates because of significant heterogeneity (I² > 75%). Subgroup analysis was performed to understand the prevalence of primary headache in different sex, age, geographical regions, and years of publication and to explore the source of heterogeneity between studies. When metapro packages were utilized for subgroup analysis, the results of the heterogeneity test were not displayed if the number of subgroup studies was 3 or fewer. Egger’s test was employed to assess publication bias, with a significance level of p < 0.05 indicating the presence of publication bias. Meta-analysis was performed using Stata version 16.0, while differences in prevalence among groups were analyzed using the chi-square test in SPSS version 22.0.

The initial literature search yielded 4,340 papers, and 46 papers related to the topic were found through title and abstract screening. Finally, 18 papers (14 in English and 4 in Chinese) containing 26 studies (15 regarding migraine, 10 regarding TTH, and 1 for cluster headache) fulfilled our inclusion criteria after full text evaluation [6, 8, 14‒29]. The selection processes were shown in Figure 1.

Table 1 shows the main characteristics of the included studies reporting the prevalence of primary headache. All studies had a cross-sectional design and investigated areas including all provinces of Chinese mainland, Hong Kong, and Taiwan (Taipei and Kinmen). Sample sizes varied widely, ranging from 197 to 3,457,170. Primary headache was diagnosed using 1988 IHS criteria (ICHD-I), ICHD-II and ICHD-III, respectively, depending on the year of data collection.

The evaluation of literature quality was conducted using the STROBE checklist. Total scores in all included studies were between 10 and 17 points (moderate-high quality, Table 1), indicating deficiencies in sample size, methodology, and statistical analysis.

The data were extracted from 15 studies, with 3,951 participants receiving a diagnosis of migraine [6, 8, 14‒26]. With the exception of five articles that did not specify the type of migraine [14, 15, 19, 20, 25], the remaining articles identified migraine without aura, migraine with aura, and headache that did not meet the above criteria. None of these articles clarified whether chronic migraine was included. The weighted-pooled prevalence was 6.1% (95% CI: 4.1–8.6%), with considerable high heterogeneity (I² = 99.3%, p < 0.001; shown in Fig. 2).

The results of subgroup analysis are presented in Table 2. The weighted-pooled prevalence of migraine exhibited an increase from 1.3% (95% CI: 1.1–1.5%) in the period spanning 1988–1999 to 7.0% (95% CI: 4.9–9.4%) in the years 2000–2023 (χ² = 211.5, p < 0.001). This prevalence was found to be 8.3% (95% CI: 5.2–11.9%) in mainland China, while in Hong Kong and Taiwan it was 4.1% (95% CI: 1.7–7.4%; χ² = 440.8, p < 0.001). According to geographical spatial classification, the prevalence of migraine in southern China is 6.3% (95% CI: 4.0–8.6%), while that in northern China is 8.6% (95% CI:1.1–16.0%). Analysis of data from ten studies [6, 15‒17, 20‒25] revealed a weighted-pooled prevalence of 7.9% (95% CI: 4.4–12.2%) in females and 3.1% (95% CI: 1.5–5.2%) in males (χ² = 453.9, p < 0.001). Furthermore, a pooled analysis of data from 12 studies [6, 15‒18, 20‒26] indicated that the highest prevalence of migraine was observed in individuals aged 40–49 (10.9%, 95% CI: 10.1–11.6%; χ² = 468.0, p < 0.001). Heterogeneity of the subgroup analyses was ranked as high.

Figures 3 and 4 represent a comparison of age-sex-specific rates between the periods of 1988–1999 and 2000–2023. It can be observed that across all age-groups in both time periods, the prevalence rate is higher among females than males. During the 1988–1999 period, the peak prevalence rates for both females and males occur in the 35–44 age-group. In the 2000–2023 period, the peak prevalence rate for females shifts to the 30–39 age group, while for males, it is the 50–59 age-group.

Data were extracted from ten studies with 3,176 subjects diagnosed with TTH [6, 14‒20, 27, 28]. The weighted-pooled prevalence was 13.4% (95% CI: 7.2–21.2%) and heterogeneity was ranked as high (I² = 99.7%, p < 0.001; shown in Fig. 5).

The results of subgroup analysis are shown in Table 3. The weighted-pooled prevalence of TTH in mainland China was 13.5% (95% CI: 6.8–21.9%), while in Hong Kong and Taiwan it was reported as 13.2% (95% CI: 2.9–29.4%; χ² = 22.1, p < 0.001). The prevalence of TTH was found to be 5.0% (95% CI: 4.5–5.4%) during the years 1988–1999 and 13.2% (95% CI: 7.1–20.8%) in the period of 2000–2023 (χ² = 71.9, p < 0.001). Data on TTH prevalence by gender were obtained from seven studies [6, 15‒17, 20, 27, 28]. The weighted-pooled prevalence of TTH in females was 16.0% (95% CI: 7.5–26.8%) and in males was 9.4% (95% CI: 4.2–16.3%; χ² = 300.9, p < 0.001). A combined analysis of the three studies [20, 27, 28] revealed that the highest prevalence of TTH was observed in individuals under 30 years old (16.8%, 95% CI: 15.6–18.0%; χ² = 570.9, p < 0.001). The heterogeneity observed in the subgroup analyses was significant (p < 0.001).

Only one study encompassing 3,457,170 participants provided epidemiological data on cluster headache [29]. The prevalence of cluster headache was 0.0068%, which was 0.0117% in males and 0.0019% in females.

Significant variability in the prevalence of migraine and TTH was observed, with high levels of heterogeneity (I² > 99%, p < 0.001), which persisted in subgroup analyses. Egger’s test revealed significant publication bias in studies concerning the prevalence of migraine and TTH (p = 0.001). A trim and fill method was employed to address this bias. However, the results remained unchanged following its application.

This systematic review and meta-analysis demonstrated that the prevalence of migraine and TTH in China was 6.1% and 13.4%, respectively. These results aligned with global trends in primary headache prevalence, with TTH (26.1%) being more prevalent than migraine (14.4%) [1]. The overdiagnosis of TTH in cases where migraine criteria are not met may contribute to an inflated prevalence of TTH and a corresponding underestimation of migraine prevalence.

The prevalence of migraine in China was found to be higher than in Africa (5.61%) yet lower than in the USA (11.7%), Japan (8.4%), Brazil (15.2%), all European countries (9.6–24.6%), and globally (14.4%) [1, 3, 15, 30‒32]. Additionally, TTH was less prevalent in China compared to the global average [33‒35]. These discrepancies may be attributed to various factors, including variations in migraine prevalence among different countries and human races, as well as differences in survey methodologies [36]. This meta-analysis incorporated epidemiological studies utilizing the ICHD diagnostic criteria published from 1988 to 2023, while Woldeamanuel et al. [30] included prevalence studies predating 1988 that employed alternative diagnostic criteria [33]. Furthermore, the quality assessment indicated that the majority of the epidemiological studies analyzed in this meta-analysis exhibited moderate quality, suggesting a deficiency in precision in the epidemiological studies of migraine and other primary headaches in China. After our careful screening, only 18 articles met the criteria. Due to this small sample size, our meta-analysis data may differ from other studies.

Our findings revealed that individuals aged 40–49 years exhibited the highest prevalence of migraine, while those below 30 years of age showed the highest prevalence of TTH. Limited data stratified by age in the various headache diagnoses analyzed in this review and meta-analysis may have contributed to the high prevalence rates observed. The findings regarding migraine and TTH prevalence partly aligned with previous researches indicating a peak prevalence of migraine in the general population globally at 35–39 years of age [1] and suggesting TTH to be more prevalent in individuals of both sexes aged 20–64 [1, 33]. This inconsistency may be attributed to the limited number of studies included on migraine and TTH and the lack of age-stratified data. Additionally, the exclusion of studies focusing solely on individuals over the age of 64 by Stovner et al. may have contributed to this discrepancy. Further detailed studies on the prevalence of migraine and other types of primary headache in China according to both age and sex are needed.

The findings of this review and meta-analysis indicated that the weighted-pooled prevalence of both migraine and TTH were higher in females than in males. These results were in line with previous research suggesting that females exhibited a higher susceptibility to both migraine and TTH compared to males [37‒41]. The observed sex-related disparities in migraine and TTH prevalence may be attributed to hormonal fluctuations in females during the menstrual cycle and across the lifespan [42, 43]. It is advisable to allocate supplementary resources and attention toward females to improve the prevention and management of their headaches in the future.

The prevalence of migraine and TTH in the Chinese mainland was found to be higher than in Hong Kong and Taiwan. Our meta-analysis also found that the prevalence of migraine is higher in the north than in the south. These variations may be attributed to various methodological factors, such as the sampling method, the number of conditions examined, the nature of the screening question, and the application of ICHD criteria. Additionally, differences in cultural activities and socioeconomic status among the southern Chinese, northern China, Hong Kong, and Taiwan populations may also contribute to these disparities. However, the urban/rural divide and relative wealth or poverty showed weak associations with years lived with disability rate of headache in some studies [33, 36, 44, 45]. Between the years 1988–1999 and 2000–2023, there was an observed rise in the prevalence of migraine and TTH, a trend that aligned with previous researches. Also, study showed migraine incidence had an overall increasing trend in China during 1990–2019 [46]. Previous studies have reported a similar uptick in migraine prevalence within the same demographic across various time periods, as well as a general increase in global migraine prevalence from 11% to 14% [8, 33, 45].

The results pertaining to the prevalence of cluster headache in China were largely in agreement with studies predominantly conducted in European and North American nations [47, 48]. Limited data on the prevalence rate of cluster headache were identified in this review, underscoring the necessity for additional research on less prevalent but equally debilitating primary headaches in China. During the COVID-19 pandemic (2019–2021), although there was no nationwide epidemiological survey, many studies have shown that COVID-19 infection can increase the frequency of migraine attacks and exacerbate anxiety and depression [49‒51].

Several limitations should be acknowledged. The meta-analysis was constrained by a limited number of studies, which may have impeded a thorough interpretation of the results. Nevertheless, our findings indicated a higher prevalence of TTH compared to migraine in China. Additionally, notable heterogeneity was observed in the prevalence of migraine and other primary headaches, despite attempts to mitigate bias. Subgroup analyses were restricted by data availability, and the influence of age, sex, and geographical region on the prevalence of migraine and other primary headaches had not been sufficiently examined.

Although the prevalence of migraine and other primary headache disorders in China was lower compared to other countries, our study indicated an increasing trend in the prevalence of migraine and TTH in China over time, with females being more susceptible to primary headaches than males. This underscores the significance of not disregarding these conditions, especially in females. It is advisable for future research to prioritize comprehensive investigations to evaluate the prevalence of these disorders in China, considering various factors such as demographics, socioeconomic status, and geographic location.

This article is meta-analysis and does not require Ethics Committee approval or a consent statement. The decision was made by the Ethics Committee of West China Hospital, Sichuan University.

The authors have no conflicts of interest to declare.

This project was supported by the Postdoctor Research Fund of West China Hospital, Sichuan University (Grant No. 2024HXBH069) and the Natural Science Foundation of China (Grant No. 82301404).

Y.Z. and C.L.: study conception and design. L.H.: supervision and administration. Y.Z., C.L., and N.C.: writing – manuscript preparation, intellectual input, and data analysis. Y.Z., C.L., N.C., J.F., M.M., P.H., and L.H.: data interpretation. Y.Z., C.L., N.C., J.F., M.M., and P.H.: experiment or data collection.

Yang Zhang and Changling Li contributed equally to this work.

Data are available on the requirement and will undergo approval.