Abstract: Objective  To investigate the correlation between triglyceride-glucose index (TyG) and cognitive impairment in elderly hemodialysis patients.  Methods  A total of 204 elderly hemodialysis patients from January 2020 to June 2020 were enrolled in this study. According to score of the Montreal Cognitive Assessment (MoCA) Chinese version, patients were divided into a normal group (MoCA score≥26 points, n=67) and an obstacle group (MoCA score <26 points, n=137). Clinical data including TyG were collected from the hospital information management system. Multivariate logistic regression was used to explore the relationship between TyG and cognitive impairment in elderly hemodialysis patients. ROC curve was used to evaluate the predictive ability of TyG for cognitive impairment.  Results  There were no statistical differences in age, gender, hypertension, BMI and TC between the two groups (χ2/t=0.593, 0.496, 0.027, 0.581 and 0.245 respectively; P=0.554, 0.481, 0.869, 0.562 and 0.806 respectively). Diabetes, stroke, dialysis age, education level and TyG were significantly different between normal group and obstacle group (χ2/t=9.566, 23.511, 7.122, 8.058 and 6.003 respectively; P=0.002, <0.001, <0.001, <0.001 and <0.001 respectively). Multivariate logistic regression showed that diabetes (OR=1.713, 95% CI: 1.512～2.013, P<0.001), stroke (OR=2.023, 95% CI: 1.647～2.216, P<0.001), dialysis age (OR=1.527, 95% CI: 1.376～1.786, P<0.001), education level (OR=0.613, 95% CI: 0.472～0.874, P<0.001), and TyG (OR=1.476, 95% CI: 1.309～1.738, P<0.001) were the influencing factors for cognitive dysfunction. ROC curve analysis showed that the area under the curve of cognitive dysfunction in elderly hemodialysis patients predicted by diabetes was 0.774, the area under the curve of cognitive dysfunction in elderly dialysis patients predicted by TyG was 0.855 (95% CI: 0.752～0.957), and the maximum Yodon index was 0.515; when TyG was set at 9 points, the optimal cutoff value was obtained, with the sensitivity of 81.25, and the specificity of 79.12. According to the optimal cutoff value from ROC curve, patients were subdivided into TyG <9 group (n=55) and TyG≥9 group (n=149). The readmission rates were 40.00% (22/55) and 59.73% (89/149) in TyG <9 group and TyG≥9 group respectively (χ2=6.305, P=0.012), and the infection rates were 20.00% (11/55) and 37.11% (59/149) in TyG <9 group and TyG≥9 group respectively (χ2=6.845, P=0.009).  Conclusion  The increase of TyG is an independent risk factor for cognitive dysfunction, and has a better predictive value for cognitive dysfunction in elderly hemodialysis patients.

Key words:  Older age, Hemodialysis, Triglyceride-glucose index, Cognitive dysfunction, Correlation