Association of metabolic insulin resistance with white matter hyperintensity burden and plasma proteomic signatures in the UK Biobank

Metabolic insulin resistance is closely related to cardiometabolic dysfunction and vascular injury. Its association with subclinical cerebral small vessel disease, and the circulating molecular signals that may accompany this association, remain incompletely characterised. We analysed 59,826 UK Biobank participants with brain magnetic resonance imaging to examine the association between the metabolic score for insulin resistance (METS-IR) and white matter hyperintensity (WMH) burden. Associations were evaluated using multivariable linear regression and restricted cubic spline analysis, with spline models restricted to participants with complete covariate data. Proteomic data were available for 6,790 participants. Protein-wide association analyses, functional annotation and LASSO prioritisation explored circulating protein links between METS-IR and WMH burden, followed by exploratory mediation models in 1,309 participants with complete model data. Higher METS-IR was associated with greater WMH burden after adjustment for demographic, socioeconomic, lifestyle and vascular risk factors. Complete-case spline analysis suggested a nonlinear association, with steeper fitted slopes at higher METS-IR values. Proteomic screening identified 1,266 proteins associated with METS-IR and 122 proteins associated with WMH after false discovery rate correction, yielding 119 overlapping proteins. These proteins were enriched for immune-inflammatory processes and cell adhesion. LASSO prioritisation selected eight candidate proteins, and exploratory mediation models yielded indirect-effect estimates for selected proteins, with GAL showing the largest estimated proportion mediated. METS-IR was associated with WMH burden after adjustment for measured covariates in this large UK Biobank analysis. Integrated plasma proteomic analyses identified immune-inflammatory and cell-adhesion signals that may help characterise molecular correlates of metabolic insulin resistance and cerebral small vessel disease burden.