M. Guo, L. Cong, Z. Liu
Background: Lacunar stroke accounts for approximately 25% of ischaemic strokes and is linked to metabolic and vascular risk factors. Abdominal adiposity, measured as waist-to-hip ratio adjusted for BMI (WHRadjBMI), is a heritable predictor of cerebrovascular disease independent of overall adiposity, but its biological mediators remain unclear. We performed a two-sample Mendelian randomization (MR) mediation analysis of 16 candidate biomarkers. Methods: Summary-level GWAS data were obtained from the IEU Open GWAS and EBI GWAS Catalog for WHRadjBMI (exposure), 16 mediators, and lacunar stroke (outcome). Instruments met genome-wide significance (P < 5 x 10), LD clumping (r2 < 0.001, 10,000 kb; 1000 Genomes European panel), and F-statistic 10. Inverse variance weighted regression was the primary analysis, with MR-Egger, weighted median, and MR-PRESSO sensitivity analyses. Mediation was quantified using the product-of-coefficients method with delta-method standard errors. Results: Systolic blood pressure showed the largest mediation proportion (29.6%; 95% CI, 12.7%-46.5%), followed by diastolic blood pressure (29.2%; 12.1%-46.3%) and glycated hemoglobin (23.0%; 11.2% to 57.2%). Other mediators included triglycerides (10.8%), fasting insulin (9.9%), blood glucose (7.0%), HDL cholesterol (3.2%), and LDL cholesterol (1.1%). Sensitivity analyses were generally consistent, though MR-Egger intercepts suggested directional pleiotropy for glycoprotein acetyls and CRP (Path A only). Because mediators are intercorrelated, proportions cannot be summed; multivariable MR is needed for joint effects. The HbA1c estimate was imprecise and requires cautious interpretation. Conclusions: Blood pressure, and possibly glycemic control, are the leading mediators linking abdominal adiposity to lacunar stroke. These findings support prioritizing blood pressure management-and potentially glycemic control-to reduce small vessel cerebrovascular disease in individuals with elevated WHRadjBMI.