Association between Plasma LRG1 and Lower Cognitive Function in Asians with Type 2 Diabetes Mellitus

Leucine-rich α-2-glycoprotein 1 (LRG1) has been implicated in the pathogenesis of diabetic complications, but its association with cognitive function remains unclear.Our primary objective is to investigate the longitudinal association between LRG1 and cognitive function in patients with type 2 diabetes mellitus (T2DM). Secondarily, we determine the causal relationship using Mendelian Randomization (MR), and the role of arterial stiffness as a potential mediator.T2DM patients (n = 1039; age = 64.1 ± 6.4 years) were followed-up for 5.3 ± 1.2 years. Plasma LRG1 was measured at baseline using enzyme-linked immunosorbent assay. Baseline and follow-up cognitive function was assessed using Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). One-sample MR was performed with rs4806985 as plasma LRG1-associated single-nucleotide polymorphism (SNP). Mediation analysis was performed to examine if pulse wave velocity (PWV), an arterial stiffness index, mediated the association between plasma LRG1 and follow-up cognitive function.Elevated baseline natural log (Ln)-transformed LRG1 was inversely associated with baseline and follow-up RBANS total score with adjusted coefficients -1.38 (95%CI -2.55 to -0.21; p = 0.021) and -1.38 (95%CI -2.70 to -0.07; p = 0.039), respectively. Genetically-predicted higher levels of plasma LRG1 was associated with lower follow-up RBANS total score with coefficient -7.44 (95%CI -14.14 to -0.74; p = 0.030) per unit increase in LnLRG1. Higher PWV accounted for 27.7% of the association between LnLRG1 and follow-up RBANS total score.Baseline plasma LRG1 was associated with lower cognitive function at follow-up in patients with T2DM, mediated by PWV. MR analysis provided evidence of an association between genetically influenced plasma LRG1 and lower cognitive function at follow-up.