Abstract
BACKGROUND: The associations of modified triglyceride-glucose (TyG) indices with risks of dementia subtypes and brain structural changes remain unclear. This study prospectively examines whether modified TyG indices, including TyG with body mass index (TyG-BMI) and TyG with waist circumference (TyG-WC), are associated with the risks of Alzheimer's disease (AD) and vascular dementia (VaD) and with structural brain alterations. MATERIALS AND METHODS: This study analyzed 356,454 dementia-free participants from the UK Biobank. Multivariable-adjusted Cox proportional hazards models were used to estimate hazard ratios (HRs) for incident AD and VaD. Restricted cubic spline (RCS) analyses assessed nonlinear relationships. Linear regression models evaluated associations between modified TyG indices and brain structures, including hippocampal volume and white matter hyperintensity (WMH) volume. Subgroup analyses and sensitivity analyses were performed to test robustness. RESULTS: During follow-up, 2,594 AD and 1,386 VaD cases were identified. In fully adjusted Cox models, both TyG-BMI and TyG-WC showed clear dose-response patterns with dementia risk. Using Q5 as the reference, participants in the lowest sextile (Q1) had a 47% higher risk of AD for TyG-BMI (HR = 1.47, FDR-adjusted p < 0.001) and a 23% higher risk for TyG-WC (HR = 1.23, FDR-adjusted p = 0.019), while those in the highest sextile (Q6) also tended to have increased AD risk. By contrast, VaD risk increased with higher modified TyG levels, and participants in the highest sextile had 32 and 45% higher VaD risk for TyG-BMI and TyG-WC, respectively (TyG-BMI: HR = 1.32, FDR-adjusted p = 0.029; TyG-WC: HR = 1.45, FDR-adjusted p = 0.011). Multivariable-adjusted restricted cubic spline analyses confirmed significant nonlinear relationships, showing a broad U-shaped association of modified TyG indices with AD and a J-shaped association with VaD. Higher modified TyG indices were additionally linked to larger hippocampal volume but greater WMH burden. The associations remained robust in multiple sensitivity analyses. CONCLUSION: Modified TyG indices show nonlinear, differential associations with AD and VaD risks, and are linked to structural brain alterations. These findings highlight the importance of metabolic health in dementia prevention and brain aging.