Insights into modifiable risk factors of migraine: a Mendelian randomization analysis

Increasing epidemiological evidence has reported that various factors are associated with migraine risk and subtypes. Nevertheless, definitive conclusions regarding whether the putative modifiable risk factors are causally related to the pathogenesis of migraine have not been drawn. Using single-nucleotide polymorphisms as instrumental variables, we conducted a two-sample Mendelian randomization (MR) analysis to investigate the causal effects of 38 modifiable factors, including dietary nutrients, lifestyle factors, cardiometabolic diseases, and associated traits, as well as reproductive characteristics and sex hormones, on the risk of migraine, migraine with aura (MA), and migraine without aura (MO). Subsequently, meta-analyses were performed to combine causal estimates from two independent genome-wide association studies. In the combined findings with multiple test correction, genetically predicted higher alcohol intake frequency (odds ratio [OR]: 1.25; 95% confidence interval [CI]: 1.12–1.40), lifetime smoking index (OR: 1.24; 95% CI: 1.08–1.42), insomnia (OR: 1.20; 95% CI: 1.17–1.24), long sleep duration (OR: 1.26; 95% CI: 1.07–1.50), and hypertension (OR: 1.76; 95% CI: 1.47–2.11) were causally linked to migraine incidence. Subgroup analyses revealed higher carbohydrate and sugar intake, alcohol consumption frequency, lifetime smoking index, insomnia, and hypertension causally increased susceptibility to MA, while later age at first birth (AFB) had a protective effect on MA risk. Meanwhile, the MR findings revealed a detrimental association between alcohol intake frequency, insomnia, hypertension, and early AFB and MO incidence. Overall, our study demonstrated various causal risk factors for migraine and its subtypes risk, providing insights into its pathogenesis and potential prevention strategies. Further research is needed to validate these findings and explore their clinical implications and underlying mechanisms.

越来越多的流行病学证据表明，多种因素与偏头痛及其亚型的发病风险相关。然而，针对潜在可改变风险因素是否与偏头痛发病机制存在因果关联，学界尚未得出明确结论。
本研究以单核苷酸多态性（single-nucleotide polymorphisms）作为工具变量，开展两样本孟德尔随机化（two-sample Mendelian randomization, MR）分析，探究38种可改变因素对偏头痛、伴先兆偏头痛（migraine with aura, MA）及无先兆偏头痛（migraine without aura, MO）发病风险的因果效应；上述可改变因素涵盖膳食营养素、生活方式因素、心脏代谢疾病及其相关性状，以及生殖特征与性激素。
后续，本研究通过荟萃分析整合两项独立全基因组关联研究（genome-wide association study, GWAS）的因果效应估计值。
经多重检验校正的合并分析结果显示，遗传预测的更高饮酒频率（比值比[OR]：1.25；95%置信区间[CI]：1.12~1.40）、终身吸烟指数（OR：1.24；95%CI：1.08~1.42）、失眠（OR：1.20；95%CI：1.17~1.24）、长睡眠时间（OR：1.26；95%CI：1.07~1.50）以及高血压（OR：1.76；95%CI：1.47~2.11）与偏头痛发病存在因果关联。
亚组分析结果显示，更高的碳水化合物与糖类摄入量、饮酒频率、终身吸烟指数、失眠及高血压会通过因果通路增加伴先兆偏头痛的发病易感性；而较晚的初产年龄（age at first birth, AFB）则对伴先兆偏头痛的发病风险具有保护作用。
与此同时，孟德尔随机化分析结果显示，饮酒频率升高、失眠、高血压以及较早的初产年龄与无先兆偏头痛的发病呈有害关联。
综上，本研究明确了偏头痛及其亚型发病的多种因果风险因素，为阐释其发病机制及探索潜在预防策略提供了新的视角。未来仍需开展进一步研究以验证本研究结果，并探讨其临床应用价值及潜在分子机制。