Background Migraine is a familial, episodic disorder characterized by complex sensory processing dysfunction, with headache serving as its hallmark feature. While numerous risk factors have been proposed, the causal nature of these associations often remains ambiguous. Mendelian randomization (MR) represents a robust epidemiological framework that leverages genetic variants to infer causal relationships, thereby overcoming limitations of observational studies. This study systematically reviews and meta-analyzes MR evidence to elucidate bidirectional causal relationships between migraine and systemic diseases, identify novel risk determinants, and highlight critical gaps for future mechanistic investigations. Methods A comprehensive literature search was conducted across seven databases (PubMed, Embase, Cochrane Library, Web of Science, China National Knowledge Infrastructure, WanFang Data Knowledge Service Platform, and VIP China Science and Technology Journal Database) using predefined search strategies and exclusion criteria. The search time limit was from the construction of the database to July 3, 2024. Study eligibility was independently assessed by two reviewers, with data extraction processes adhering to STROBE-MR guidelines. Included studies were evaluated for quality using validated criteria, and relevant data (study design, participant demographics, genetic instruments, analytical methods, and outcomes) were systematically extracted. Data synthesis involved meta-analytical pooling of effect estimates using Review Manager 5.4, with forest plots generated to visualize results. Causal relationships were interpreted according to the WHO ICD-11 disease classification system, with subgroup analyses performed for migraine with aura (MWA) and migraine without aura (MOA). Results A total of 60 studies involving 331 MR analyses were included, revealing bidirectional causal relationships between migraine and multiple phenotypes: migraine was identified as a causal factor for 6 diseases (Alzheimer’s disease, cervical artery dissection, venous thromboembolism, coronary artery disease, angina, large artery stroke), 3 behavioral habits (delayed age at first sexual intercourse, maternal smoking, reduced physical activity), 1 dietary intakes (alcohol consumption), and 3 physiological indicators (elevated interleukin-2, increased Body Mass Index, higher serum vitamin D levels) ( p 0.05). Conversely, 6 diseases (venous thromboembolism, breast cancer, insomnia, difficulty awakening, major depressive disorder, depression), 5 behavioral factors (television watching, smoking initiation, delayed AFS, more schooling, reduced physical activity), 4 dietary determinants (coffee, alcohol, cheese, salad intake), 13 physiological parameters (hemostatic, cardiovascular, metabolic, and genetic markers), and 1 gut microbiota taxon ( LachnospiraceaeUCG001 ) were causal determinants of migraine risk ( p 0.05). Subtype-specific analyses showed MOA was causally associated with 4 diseases (AD, CeAD, CAD, LAS) and delayed AFS as an exposure, and influenced by breast cancer, celiac disease, TV watching, delayed AFS, increased schooling, and physiological parameters (DBP, PP, serum calcium, IGF-1) as an outcome; MWA demonstrated causal relationships with CeAD and LAS as an exposure, and associations with VTE, SLE, MDD, delayed AFS, coffee intake, and hemostatic markers as an outcome ( p 0.05 for all). Conclusion This systematic review provides robust genetic evidence supporting bidirectional causal relationships between migraine and multiple phenotypes, including systemic diseases, behavioral habits, dietary factors, and physiological parameters. Subtype-specific analyses highlight distinct causal pathways for MOA and MWA, underscoring the clinical heterogeneity of migraine. These findings advance our understanding of migraine pathogenesis and inform precision medicine approaches, while also identifying novel therapeutic targets for this disabling condition. More data will be needed in the future to obtain a more specific assessment. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/view/CRD42025636141 , Identifier CRD42025636141.
Li et al. (Tue,) studied this question.