Skip to main content
Download PDF

Genetic evidence for causal association between migraine and dementia: a mendelian randomization study

BMC Medical Genomics volume 17, Article number: 180 (2024) Cite this article

Abstract

Background

There is an association between migraine and dementia, however, their causal relationship remains unclear. This study employed bidirectional two-sample Mendelian randomization (MR) to investigate the potential causal relationship between migraine and dementia and its subtypes: Alzheimer’s disease (AD), vascular dementia (VaD), frontotemporal dementia (FTD), and dementia with Lewy bodies (DLB).

Methods

Summary-level statistics data were obtained from publicly available genome-wide association studies (GWAS) for both migraine and five types of dementia. Single nucleotide polymorphisms (SNPs) associated with migraine and each dementia subtype were selected. MR analysis was conducted using inverse variance weighting (IVW) and weighted median (WM) methods. Sensitivity analyses included Cochran’s Q test, MR pleiotropy residual sum and outlier (MR-PRESSO) analysis, the intercept of MR-Egger, and leave-one-out analysis.

Results

Migraine showed a significant causal relationship with AD and VaD, whereas no causal relationship was observed with all-cause dementia, FTD, or DLB. Migraine may be a potential risk factor for AD (odds ratio [OR]: 1.09; 95% confidence interval [CI]: 0.02–0.14; P = 0.007), while VaD may be a potential risk factor for migraine (OR: 1.04; 95% CI: 0.02–0.06; P = 7.760E-5). Sensitivity analyses demonstrated the robustness of our findings.

Conclusion

Our study suggest that migraine may have potential causal relationships with AD and VaD. Migraine may be a risk factor for AD, and VaD may be a risk factor for migraine. Our study contributes to unraveling the comprehensive genetic associations between migraine and various types of dementia, and our findings will enhance the academic understanding of the comorbidity between migraine and dementia.

Peer Review reports

Introduction

Migraine is a highly prevalent neurological disorder characterized by severe headache, throbbing, and associated symptoms such as photophobia, phonophobia, nausea, and vomiting [1]. It affected approximately 14.4% of the world population and was considered the second largest contributor to the disability, owing to pain and recurrent attacks [2]. People who suffer from migraine is twice as likely as the general population to develop dementia, and migraine is linked to an increased risk of subsequent dementia [3, 4]. Although there may be an association between migraine and dementia, migraine pathophysiology remains poorly understood and requires further exploration. Dementia primarily includes Alzheimer’s disease (AD), vascular dementia (VaD), frontotemporal dementia (FTD), and Lewy body dementia (DLB). Several researches have reported an increased risk of all-cause dementia, AD, or VaD in patients with migraine compared to individuals without migraine [5,6,7]. However, little research has been conducted on the relationship between migraine and FTD or DLB.

Furthermore, previous research has demonstrated that individuals diagnosed with migraine exhibit a heightened susceptibility to dementia, with this correlation being predominantly observed among female patients [8]. Attention, memory, and language skills are impaired in patients with migraine because they develop abnormalities in the white matter at a faster rate than those without migraine. Damage to the white matter of the brain, where neuronal fibers gather, can lead to the above symptoms of dementia [9, 10]. In addition, depression is a common comorbidity between migraine and dementia, which suggests a central mechanism between migraine and dementia [11, 12]. Notably, these results indicate that migraine may be a risk factor for dementia, and identifying the risk factors for dementia may help in the early identification of individuals at risk and the development of prevention strategies. However, the results of these studies are not completely consistent, and research exploring the causal association between migraine and dementia is lacking.

Traditional observational studies’ conclusions are usually impacted by confounding factors and cannot establish causation. Additionally, randomized clinical trials may be limited by ethical considerations. Therefore, effective strategies must be developed to determine the causal relationships between exposure factors and response outcomes. Mendelian randomization (MR) was proposed in the 1990s, and with the emergence of large-scale genome-wide association studies (GWAS), MR has increasingly improved the research framework and methods [13, 14]. MR uses single nucleotide polymorphism (SNP) associations from GWAS to study and uncover the causal relationships between complex traits [15]. In our study, we used bidirectional two-sample MR analysis to investigate the genetic relationships between migraine risk and dementia, as well as performed a causal inference analysis.

Materials and methods

Study design

This study was designed as a bidirectional two-sample MR research study, an overview of which is presented in Fig. 1. First, we used linkage disequilibrium score regression (LDSC) to evaluate the genetic links between migraine and the five types of dementia. Second, we conducted an MR analysis and a series of sensitivity analyses to assess the causal relationship between migraine and dementia. Finally, we employed reverse MR to determine whether dementia was linked with migraine.

Fig. 1
figure 1

Overview of this bidirectional MR study design

Full size image

To ensure that the causal relationships derived from the MR analysis are reliable, instrumental variables (IVs) should satisfy three core assumptions: (1) IVs are significantly associated with the exposure feature, (2) IVs are unrelated to other confounding factors, and (3) IVs do not affect the outcome through any other possible pathways (Fig. 2) [14, 16].

Fig. 2
figure 2

IVs should satisfy three core assumptions

Full size image

Data sources

We acquired GWAS summary-level data for migraine from the 2022 International Headache Genetics Consortium (IHGC), which aggregates datasets from the UK Biobank, GeneRISK, HUNT, and the GWAS of the 2016 IHGC, for a total of 24 cohorts, 48,975 patients with migraine and 540,381 migraine-free individuals of European ancestry [17]. This 2022 IHGC migraine GWAS contained association results for 123 independent risk loci. To protect the privacy of participants in the 23andMe cohort, the migraine GWAS summary statistics applied in our analysis excluded individuals from this cohort (Table 1).

The summary-level GWAS dataset for AD was derived from a recent GWAS that included 788,989 individuals (111,326 patients and 677,663 controls) from Europe [18]. We acquired GWAS summary statistics for VaD from the FinnGen Consortium (Round 9, https://www.finngen.fi/en), comprising a total of 363,113 individuals of European descent (2,335 patients and 360,778 controls). The GWAS summary statistics for FTD were obtained from a study conducted by Ferrari et al. This study included a total of 12,932 individuals, consisting of 3,526 patients and 9,402 control individuals [19]. The DLB GWAS dataset was extracted from a study of 6,618 individuals (2,591 patients and 4,027 controls) [20]. Table 1 presents the essential characteristics of the dataset included in this study.

Table 1 Details of GWAS datasets included in MR analysis
Full size table

Ethical approval

All summary-level datasets in our study were obtained from de-identified public data/studies. Ethical approval and informed consent were previously obtained from the ethics committee. Thus, the requirement for ethical approval was waived for this study.

Instrumental variables selection

The key to the selection of IVs is to ensure the accuracy and robustness of the causal inferences. The following steps were performed: First, single nucleotide polymorphisms (SNPs) that were strongly linked with exposures were screened at genome-wide significance (P < 5 × 10− 8). For FTD and VaD, only a few or no SNPs reached the significance threshold (P < 5 × 10− 8), so we chose a more relaxed threshold (P < 1 × 10− 5) to select instrumental variables. This threshold was used in previous MR studies when a limited number of SNPs met the conventional threshold [21,22,23]. Second, we merged the IVs with the aforementioned significance and the outcome data to avoid losing too many IVs; therefore, we did not search for SNPs. The identified SNPs underwent clumping for LD using a stringent threshold of clumping r² < 0.001 within a 10,000 kb window. LD was estimated using European samples from the 1000 Genome Project as a reference [24]. The IVs are listed in Supplementary Table S1. Thirdly, in order to evaluate the presence of weak instrument bias, we utilized the F statistic, which was calculated according to previous study described methods, to assess the association of the retained SNPs [25]. SNPs with an F statistic < 10 were regarded as weak instruments and were subsequently excluded from the analysis [26]. The F-statistic is calculated using the following formula: F = R² × (N − 2)/(1 − R²); R² = 2 × EAF × (1 − EAF) × β², in which, R² refers to the cumulative explained variance of the selected IVs on migraine, EAF refers to the effect allele frequency, β refers to the estimated effect of SNPs, and N refers to the sample size of the GWAS [27]. Because of the lack of an FTD EAF, we did not calculate the R² and F statistics for the FTD. Finally, after harmonizing the exposure and outcome datasets and excluding palindromic SNPs, the remaining SNPs were utilized for the MR analysis.

Statistical analysis

To assess the causal impact of migraine on dementia, we conducted the random-effects inverse variance-weighted (IVW) method in a two-sample MR analysis, which is extensively used in MR investigations and offers reliable causal effect estimation. We employed the weighted median method to further validate the results, and this method often exhibits larger positive causal effects, especially as the number of invalid instrumental variables increased. Combining these two analysis methods can enhance the reliability of the conclusions [28]. Moreover, we performed extensive sensitivity analyses to analyze potential model assumptions violations in the MR study and improve the robustness of the MR results. Cochran’s Q test was used to analyze the heterogeneity of individual causal effects. Furthermore, MR pleiotropy residual sum and outlier (MR-PRESSO) analysis was conducted to detect outlier instrumental variables [29], and horizontal pleiotropy was examined using the MR-PRESSO global test. In order to minimize the influence of horizontal pleiotropy, outlier IVs found by the MR-PRESSO outlier test were deleted step by step. The MR-PRESSO distortion test was used to analyze whether there were significant differences in the results before and after excluding outliers, and the MR-Egger intercept was used to determine pleiotropy [30]. Finally, we performed a leave-one-out analysis to explore the impact of excluding each selected individual SNP on the overall results. Moreover, we conducted a reverse MR analysis to test the potential for reverse causation (i.e., we assessed the effects of dementia on migraine). Statistical significance was set at P < 0.05 each analysis, and statistical power was calculated using an online tool at https://shiny.cnsgenomics.com/mRnd [31]. Owing to the lack of an FTD EAF, we did not calculate the statistical power for FTD when it was used as an exposure risk factor. The statistical analysis carried out in R 4.3.1 using the ‘TwoSampleMR’ package (version 0.5.7) [32].

Results

When migraine was considered as an exposure factor under the more stringent threshold (P < 5 × 10− 8 and clumped at LD r² = 0.001), 35 SNPs were selected for any dementia, AD, VaD, and DLB, and 33 SNPs for FTD, after harmonization. When the five types of dementia were considered as exposure factors, under the condition of P < 5 × 10− 8 and clumped at LD r² = 0.001, 14 SNPs were selected for any dementia, 51 SNPs for AD, and five SNPs for DLB. Additionally, under the condition of P < 1 × 10− 5, we chose 23 SNPs for VaD and 10 SNPs for FTD. The calculated F-statistics of these SNPs were 136.50-51700.56, suggesting that strong instrumental variables were used (Table 2). In the forward MR analysis of migraine and AD, the statistical power was > 80%. When dementia and VaD were considered as risk factors, a statistical power of > 80% was observed in the reverse MR analysis (Table 2). Detailed information on the SNPs associated with migraine and the five types of dementia is provided in Supplementary Table S1.

The main results of the bidirectional MR analysis are presented in Table 3. The scatter and leave-one-out plots for each analysis are shown in Figs. 3, 4, 5 and 6.

Fig. 3
figure 3

Scatter plots of the causal impact of migraine on dementia. The scatter plots show the causal effects of migraine on any dementia (A), Alzheimer’s disease (B), vascular dementia (C), frontotemporal dementia (D), dementia with lewy bodies (E), when migraine is the risk factor. Three lines reveal the estimated effect sizes by MR methods (IVW, MR-Egger and weighted median)

Full size image
Fig. 4
figure 4

Scatter plots of the causal impact of dementia on migraine. The scatter plots show the causal effects of any dementia (A), Alzheimer’s disease (B), vascular dementia (C), frontotemporal dementia (D), dementia with lewy bodies (E) on migraine, when five types of dementia are the risk factors. Three lines reveal the estimated effect sizes by MR methods (IVW, MR-Egger and weighted median)

Full size image
Fig. 5
figure 5

Leave-one-out plots of the causal impact of migraine on dementia. Leave-one-out plots show the association between any dementia (A), Alzheimer’s disease (B), vascular dementia (C), frontotemporal dementia (D), dementia with lewy bodies (E) and migraine when migraine is the risk factor, using IVs that excluded 1 SNP at a time from the overall instrumental variable or using IVW as an MR method to exclude all SNPs. Bars show effect size and 95% confidence interval

Full size image
Fig. 6
figure 6

Leave-one-out plots of the causal impact of dementia on migraine. Leave-one-out plots show the association between any dementia (A), Alzheimer’s disease (B), vascular dementia (C), frontotemporal dementia (D), dementia with lewy bodies (E) and migraine when 5 types of dementia are the risk factors, using IVs that excluded 1 SNP at a time from the overall instrumental variable or using IVW as an MR method to exclude all SNPS. Bars show effect size and 95% confidence interval

Full size image

Forward MR analysis

For any dementia, using the IVW method, we did not identify any potential causal relationship between migraine and any dementia (odds ratio [OR]: 1.02; 95% confidence interval [CI]: -0.08–0.11; P = 0.711), which is consistent with the results obtained from the weighted median method (OR: 1.00; 95% CI: -0.09–0.09; P = 0.994). In the sensitivity analysis, the MR analysis revealed some heterogeneity (Cochran’s Q, P < 0.001). The Global test using the MR PRESSO method detected a certain level of horizontal pleiotropy (P < 0.001), and the subsequent outlier test identified two SNPs. However, the results of the distortion test showed no difference before and after correction (P = 0.39), and the intercept from the MR-Egger test was not significant (P = 0.910, Table 3). Furthermore, the leave-one-out plot indicated that no SNP had a significant impact on the outcome (Fig. 5).

For AD, consistent results from both the IVW (OR: 1.09; 95% CI: 0.02–0.14; P = 0.007) and weighted median methods (OR: 1.13; 95% CI: 0.05–0.18; P = 3.000E-04) revealed a positive genetic correlation with migraine. The sensitivity analysis revealed heterogeneity (Cochran’s Q, P = 0.001) and horizontal pleiotropy (MR PRESSO Global test, P = 0.002), and MR PRESSO outlier testing identified one SNP. However, the results of the distortion test (P = 0.614) and the intercept of the MR-Egger’s test (P = 0.178) were not statistically significant (Table 3). Furthermore, the leave-one-out plot indicated the robustness of the findings (Fig. 5).

For VaD, this MR study did not find any potential risk relationship between VaD and migraine using the IVW (OR: 0.88; 95% CI: -0.32–0.05; P = 0.163) and weighted median methods (OR: 0.84; 95% CI: -0.42–0.07; P = 0.158). Moreover, the sensitivity analysis did not reveal any heterogeneity (Cochran’s Q P = 0.100), horizontal pleiotropy (MR PRESSO Global test P = 0.098), or pleiotropy (MR Egger intercept test P = 0.098; Table 3). Furthermore, the leave-one-out plot visually demonstrates the stability of the results (Fig. 5).

For FTD, the IVW (OR: 0.88; 95% CI: -0.36–0.10; P = 0.277) and weighted median methods (OR: 0.74; 95% CI: -0.65–0.05; P = 0.089) did not reveal any potential causal effect between FTD and migraine. Additionally, no evidence of heterogeneity (Cochran’s Q, P = 0.628), horizontal pleiotropy (MR PRESSO Global test, P = 0.617), or pleiotropy (MR Egger intercept test, P = 0.276) was observed (Table 3). Furthermore, the stability of the results was demonstrated by the leave-one-out plot (Fig. 5).

For DLB, using the IVW (OR: 0.98; 95% CI: -0.22–0.18; P = 0.850) and weighted median methods (OR: 0.93; 95% CI: -0.36–0.22; P = 0.621), we did not identify any potential causal effect. The sensitivity analysis showed a slightly significant result in the pleiotropy test (MR-Egger intercept test, P = 0.047); however, the results of the heterogeneity test (Cochran’s Q, P = 0.604) and MR PRESSO Global test for horizontal pleiotropy (P = 0.615) were not statistically significant (Table 3). Furthermore, the leave-one-out plot supported the credibility of the results (Fig. 5).

Reverse MR analysis

For the reverse MR analysis, we used the IVW method to investigate whether the five types of dementia are potential causal risk factors for migraine. Notably, we did not find any potential causal effects of dementia, AD, FTD, or DLB on migraine. In the reverse MR analysis for VaD, the IVW method did not show statistical significance (OR: 1.02; 95% CI: 0.00–0.04; P = 0.060). However, the sensitivity analysis revealed heterogeneity (Cochran’s Q P = 0.034) and horizontal pleiotropy (MR PRESSO Global test P = 0.047), and the leave-one-out plot indicated a potential bias from SNPs influencing the results. After excluding the five SNPs with significant effects from the leave-one-out analyses, we conducted a reanalysis with the filtered SNPs and identified a significant causal effect of VaD on migraine (OR: 1.04; 95% CI: 0.02–0.06; P = 7.760E-05). The results did not exhibit significant heterogeneity (Cochran’s Q, P = 0.638), pleiotropy (MR Egger intercept test, P = 0.122), or horizontal pleiotropy (MR PRESSO Global test, P = 0.576; Table 3), revealing the robustness and credibility of the MR analysis results (Fig. 6).

Table 2 The characteristics of each group in MR analysis
Full size table
Table 3 The main results of the bidirectional MR analysis
Full size table

Discussion

To the best of our knowledge, this is the first bidirectional two-sample MR analysis conducted to assess the causal relationship between migraine and dementia, including its subtypes. In our MR study, we primarily investigated the potential causal relationships between migraine and the five types of dementia. Our study contributes to unraveling the genetic associations between migraine and various types of dementia, and our findings will facilitate the advancement of clinical medicine and enhance the academic understanding of the comorbidity between migraine and dementia.

We found a positive association between genetically predicted migraine and AD, and this causal relationship remained significant when using the weighted median method. Although some bias risks were identified in heterogeneity and horizontal pleiotropy detection, pleiotropy was not significant, and the MR-PRESSO distortion test showed no statistical difference in the results before and after outlier correction. Notably, the leave-one-out plot analysis also revealed the robustness of the results. In the context of the reverse MR analysis, after removing five SNPs with substantial bias effects based on the leave-one-out plot, VaD ultimately exhibited a significant causal effect on migraine.

Several scholars have investigated the association between migraine and dementia. For instance, in a meta-analysis, researchers have found that migraine may be associated with an increased risk of all-cause dementia, AD, and VaD [33]. However, our results do not genetically support migraine as a risk factor for all-cause dementia or VaD. Additionally, a previous study conducted a MR analysis to appraise the potential influences of 18 neurodegenerative and neuropsychiatric disorders on AD. They found that four disorders, including migraine, bipolar disorder, schizophrenia, and Parkinson’s disease, are causally associated with an increased risk for AD [34]. These findings are consistent with our study results. Although our study suggests that VaD may be an important risk factor for migraine, the association between migraine and VaD remains controversial. Some studies have indicated a causal relationship between migraine and dementia, particularly VaD [35]. However, another study suggested a significant association between migraine and all-cause dementia and AD, but not VaD [36]. The heterogeneity among the different research conclusions may be attributed to differences in study designs, such as the specific diagnostic subtypes of patients with migraine, geographical variations, follow-up duration, and different outcome measures (all-cause dementia, AD, or VaD) included in previous cohort studies. In addition, cohort studies often face challenges in terms of obtaining adequate sample sizes and sufficient statistical power. In this study, we used publicly available large-scale databases to provide robust statistical power. Additional research is warranted to elucidate the association between migraine and dementia, validating the findings of our study.

Our study suggests that migraine may have potential causal relationships with AD and VaD. Specifically, migraine may be a risk factor for AD, and VaD may be a risk factor for migraine. Given these associations, several potential future directions for clinical practice emerge. First, routine cognitive impairment screenings for migraine patients could facilitate early detection of cognitive decline, enabling timely interventions to potentially slow dementia progression. Second, considering the complex relationship between migraine and various cerebrovascular conditions, early management of vascular risk factors is crucial. This could involve regular monitoring and control of blood pressure, lipid levels, and other vascular health indicators in migraine patients, thereby reducing the incidence of vascular events and associated dementia risk. Third, developing individualized treatment plans by tailoring strategies based on the specific medical history and risk factors of migraine patients is essential. For instance, migraine patients with a family history of hypertension or cardiovascular diseases might benefit from more aggressive preventive measures. Lastly, encouraging close collaboration between neurologists, cardiovascular and cerebrovascular specialists can provide a more holistic assessment and management of these patients’ health risks, improving overall outcomes.

This study has several limitations. First, the bidirectional two-sample MR analysis offers advantages over traditional observational studies, as it can mitigate confounding factors and distinguish between the directions of causal relationships. Second, comprehensive sensitivity analyses were conducted to ensure the robustness and credibility of the results. Additionally, we categorized dementia into five types, to provide richer information and more reliable results.

Nevertheless, our study design and results had some limitations. First, our data sources were limited to European populations, which limits the generalizability of our findings to other populations. Second, owing to the lack of explicit differentiation between chronic and episodic migraine in the available data, we were unable to perform additional MR analyses. Moreover, because the GWAS data were aggregated, subgroup analyses were not feasible. Third, some heterogeneity and pleiotropy were observed during the analysis, which may introduce biases that need to be interpreted in conjunction with the sensitivity analysis results. Fourth, although we made efforts to select the largest available datasets to date, some combinations of exposures and outcomes may have low statistical power owing to the small sample size, potentially leading to false-negative results. Thus, future large-scale GWAS are needed to explore the association between migraine and dementia. Fifth, regarding FTD and VaD as exposure variables, the significance threshold for SNP selection was relatively low. Although the F-statistic suggests a low likelihood of weak instrument bias, false-positive variations and the resulting biases cannot be completely avoided, and caution is advised when interpreting the significant causal effect of VaD on migraine. For VaD and FTD, further MR analyses should be conducted when a sufficient number of SNPs are obtained by screening using conventional thresholds. Sixth, there may be sample overlap between dementia, VaD, and migraine in some data sources, which could lead to a winner’s curse and weak instrument bias. Further MR analyses should be conducted with larger sample sizes and different data sources from other GWAS studies, as they become available.

Conclusion

Our finding suggests that migraine may have a potential causal relationship with AD and VaD. On one hand, migraine may be a risk factor for AD, while on the other hand, VaD may be a risk factor for migraine. Our research contributes to a better understanding of the causal relationship among migraine, dementia, and its subtypes from a genetic perspective, filling a gap in the academic field and enhancing the academic understanding of the comorbidity between migraine and dementia. Further exploration of the shared biological mechanisms underlying migraine and dementia is required. Updated MR studies will be necessary to validate our findings when more extensive GWAS summary data are available.

Data availability

The data used in this study are publicly available. The complete migraine dataset is available for download from the 2022 International Headache Genetics Consortium (IHGC) website: http://www.headachegenetics.org. The dementia GWAS summary results are publicly available for download from the FinnGen Consortium (https://www.finngen.fi/en) and three articles (https://doi.org/10.1038/s41588-022-01024-z; https://doi.org/10.1016/s1474-4422(14)70065-1; https://doi.org/10.1038/s41588-021-00785-3).

Abbreviations

MR:

Mendelian randomization

AD:

Alzheimer’s disease

VaD:

Vascular dementia

FTD:

Frontotemporal dementia

DLB:

Dementia with Lewy bodies

GWAS:

Genome-wide association studies

SNPs:

Single nucleotide polymorphisms

IVW:

Inverse variance weighting

WM:

Inverse variance weighting

MR-PRESSO:

Mendelian randomization pleiotropy residual sum and outlier

LDSC:

Linkage disequilibrium score regression

IVs:

Instrumental variables

IHGC:

International Headache Genetics Consortium

OR:

Odds ratio

95% CI:

95% Confidence interval

References

  1. Headache Classification Committee of the International Headache Society (IHS) The International Classification of Headache Disorders, 3rd edition. Cephalalgia. 2018;38(1):1-211.

  2. Global regional, national burden of neurological disorders. 1990–2016: a systematic analysis for the global burden of Disease Study 2016. Lancet Neurol. 2019;18(5):459–80.

    Article  Google Scholar 

  3. Hurh K, Jeong SH, Kim SH, Jang SY, Park EC, Jang SI. Increased risk of all-cause, Alzheimer’s, and vascular dementia in adults with migraine in Korea: a population-based cohort study. J Headache Pain. 2022;23(1):108.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Morton RE, St John PD, Tyas SL. Migraine and the risk of all-cause dementia, Alzheimer’s disease, and vascular dementia: a prospective cohort study in community-dwelling older adults. Int J Geriatr Psychiatry. 2019;34(11):1667–76.

    Article  PubMed  Google Scholar 

  5. Hagen K, Stordal E, Linde M, Steiner TJ, Zwart JA, Stovner LJ. Headache as a risk factor for dementia: a prospective population-based study. Cephalalgia. 2014;34(5):327–35.

    Article  PubMed  Google Scholar 

  6. Islamoska S, Hansen ÅM, Wang HX, Garde AH, Andersen PK, Garde E, et al. Mid- to late-life migraine diagnoses and risk of dementia: a national register-based follow-up study. J Headache Pain. 2020;21(1):98.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. George KM, Folsom AR, Sharrett AR, Mosley TH, Gottesman RF, Hamedani AG, et al. Migraine Headache and Risk of Dementia in the atherosclerosis risk in communities Neurocognitive Study. Headache. 2020;60(5):946–53.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Kostev K, Bohlken J, Jacob L. Association between Migraine Headaches and Dementia in more than 7,400 patients followed in General practices in the United Kingdom. J Alzheimers Dis. 2019;71(1):353–60.

    Article  CAS  PubMed  Google Scholar 

  9. Sas K, Párdutz A, Toldi J, Vécsei L. Dementia, stroke and migraine–some common pathological mechanisms. J Neurol Sci. 2010;299(1–2):55–65.

    Article  PubMed  Google Scholar 

  10. Hamedani AG, Rose KM, Peterlin BL, Mosley TH, Coker LH, Jack CR, et al. Migraine and white matter hyperintensities: the ARIC MRI study. Neurology. 2013;81(15):1308–13.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Christensen MC, Schmidt SN, Grande I. Effectiveness of vortioxetine in patients with major depressive disorder and early-stage dementia: the MEMORY study. J Affect Disord. 2023;338:423–31.

    Article  CAS  PubMed  Google Scholar 

  12. George A, Minen MT. Episodic migraine and Psychiatric Comorbidity: a narrative review of the literature. Curr Pain Headache Rep. 2023;27(9):461–9.

    Article  PubMed  Google Scholar 

  13. Smith GD, Ebrahim S. Mendelian randomization’: can genetic epidemiology contribute to understanding environmental determinants of disease? Int J Epidemiol. 2003;32(1):1–22.

    Article  PubMed  Google Scholar 

  14. Davies NM, Holmes MV, Davey Smith G. Reading mendelian randomisation studies: a guide, glossary, and checklist for clinicians. BMJ. 2018;362:k601.

    Article  PubMed  PubMed Central  Google Scholar 

  15. Boehm FJ, Zhou X. Statistical methods for mendelian randomization in genome-wide association studies: a review. Comput Struct Biotechnol J. 2022;20:2338–51.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Skrivankova VW, Richmond RC, Woolf BAR, Davies NM, Swanson SA, VanderWeele TJ, et al. Strengthening the reporting of observational studies in epidemiology using mendelian randomisation (STROBE-MR): explanation and elaboration. BMJ. 2021;375:n2233.

    Article  PubMed  PubMed Central  Google Scholar 

  17. Hautakangas H, Winsvold BS, Ruotsalainen SE, Bjornsdottir G, Harder AVE, Kogelman LJA, et al. Genome-wide analysis of 102,084 migraine cases identifies 123 risk loci and subtype-specific risk alleles. Nat Genet. 2022;54(2):152–60.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Bellenguez C, Küçükali F, Jansen IE, Kleineidam L, Moreno-Grau S, Amin N, et al. New insights into the genetic etiology of Alzheimer’s disease and related dementias. Nat Genet. 2022;54(4):412–36.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Ferrari R, Hernandez DG, Nalls MA, Rohrer JD, Ramasamy A, Kwok JB, et al. Frontotemporal dementia and its subtypes: a genome-wide association study. Lancet Neurol. 2014;13(7):686–99.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Chia R, Sabir MS, Bandres-Ciga S, Saez-Atienzar S, Reynolds RH, Gustavsson E, et al. Genome sequencing analysis identifies new loci associated with Lewy body dementia and provides insights into its genetic architecture. Nat Genet. 2021;53(3):294–303.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Deng Y, Wong MCS. Association between rheumatoid arthritis and osteoporosis in Japanese populations: a mendelian randomization study. Arthritis Rheumatol. 2023;75(8):1334–43.

    Article  CAS  PubMed  Google Scholar 

  22. Li N, Wang Y, Wei P, Min Y, Yu M, Zhou G et al. Causal effects of specific gut microbiota on chronic kidney diseases and renal Function-A two-sample mendelian randomization study. Nutrients. 2023;15(2).

  23. Liu YQ, Liu Y, Chen ZY, Li H, Xiao T. Rheumatoid arthritis and osteoporosis: a bi-directional mendelian randomization study. Aging. 2021;13(10):14109–30.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Auton A, Brooks LD, Durbin RM, Garrison EP, Kang HM, Korbel JO, et al. A global reference for human genetic variation. Nature. 2015;526(7571):68–74.

    Article  PubMed  Google Scholar 

  25. Burgess S, Davies NM, Thompson SG. Bias due to participant overlap in two-sample mendelian randomization. Genet Epidemiol. 2016;40(7):597–608.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Palmer TM, Lawlor DA, Harbord RM, Sheehan NA, Tobias JH, Timpson NJ, et al. Using multiple genetic variants as instrumental variables for modifiable risk factors. Stat Methods Med Res. 2012;21(3):223–42.

    Article  PubMed  PubMed Central  Google Scholar 

  27. Pierce BL, Ahsan H, Vanderweele TJ. Power and instrument strength requirements for mendelian randomization studies using multiple genetic variants. Int J Epidemiol. 2011;40(3):740–52.

    Article  PubMed  Google Scholar 

  28. Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estimation in mendelian randomization with some Invalid instruments using a weighted median estimator. Genet Epidemiol. 2016;40(4):304–14.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from mendelian randomization between complex traits and diseases. Nat Genet. 2018;50(5):693–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Burgess S, Thompson SG. Interpreting findings from mendelian randomization using the MR-Egger method. Eur J Epidemiol. 2017;32(5):377–89.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Brion MJ, Shakhbazov K, Visscher PM. Calculating statistical power in mendelian randomization studies. Int J Epidemiol. 2013;42(5):1497–501.

    Article  PubMed  Google Scholar 

  32. Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D et al. The MR-Base platform supports systematic causal inference across the human phenome. Elife. 2018;7.

  33. Gu L, Wang Y, Shu H. Association between migraine and cognitive impairment. J Headache Pain. 2022;23(1):88.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Baranova A, Zhao Q, Cao H, Chandhoke V, Zhang F. Causal influences of neuropsychiatric disorders on Alzheimer’s disease. Transl Psychiatry. 2024;14(1):114.

    Article  PubMed  PubMed Central  Google Scholar 

  35. Jiang W, Liang GH, Li JA, Yu P, Dong M. Migraine and the risk of dementia: a meta-analysis and systematic review. Aging Clin Exp Res. 2022;34(6):1237–46.

    Article  PubMed  Google Scholar 

  36. Wang L, Wu JC, Wang FY, Chen X, Wang Y. Meta-analysis of association between migraine and risk of dementia. Acta Neurol Scand. 2022;145(1):87–93.

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

Thanks to all researchers and editors who contributed to this study.

Funding

This study was supported by China National Natural Science Foundation (82374575, 82074179), Beijing Natural Science Foundation (7232270), Capital Health Development Scientific Research Project Independent Innovation (Capital Development 2024-2-2235), Outstanding Young Talents Program of Capital Medial University (B2207), R&D Program of Beijing Municipal Education Commission (KM202110025005), and China Association for Science and Technology Young Talent Lifting Project (2019-2021ZGZJXH-QNRC001).

Author information

Author notes

  1. Qiuyi Chen and Chengcheng Zhang contributed equally to this work.

Authors and Affiliations

  1. Beijing Key Laboratory of Acupuncture Neuromodulation, Department of Acupuncture and Moxibustion, Beijing Hospital of Traditional Chinese Medicine, Capital Medical University, Beijing, 100010, China

    Qiuyi Chen, Chengcheng Zhang, Yuhan Liu, Libin Zheng, Bin Li & Lu Liu

  2. Beijing Institute of Brain Disorders, Laboratory of Brain Disorders, Collaborative Innovation Center for Brain Disorders, Ministry of Science and Technology, Capital Medical University, Beijing, 100069, China

    Shiyang Wu & Guiyou Liu

  3. State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications, Beijing, 100876, China

    Yiwei He

Authors
  1. Qiuyi Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Chengcheng Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Shiyang Wu
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Yiwei He
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Yuhan Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Libin Zheng
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Bin Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Guiyou Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Lu Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

LL, LGY, LB, CQY and ZCC contributed to the study conception and design. Material preparation, data collection and analysis were performed by ZCC, HYW and WSY. The first draft of the manuscript was written by CQY, ZCC, LYH and ZLB, and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Guiyou Liu or Lu Liu.

Ethics declarations

Ethics approval and consent to participate

Ethical approval and consent to participate is waived by the Research Ethics Committee of Beijing Hospital of Traditional Chinese Medicine, Capital Medical University is not required.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Chen, Q., Zhang, C., Wu, S. et al. Genetic evidence for causal association between migraine and dementia: a mendelian randomization study. BMC Med Genomics 17, 180 (2024). https://doi.org/10.1186/s12920-024-01956-x

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12920-024-01956-x

Keywords

BMC Medical Genomics

ISSN: 1755-8794

Contact us