Association of Lipid- and Weight-Related Indices with Cognitive Impairment: Insights from ROC Analysis in Middle-Aged and Elderly Chinese Adults

OBJECTIVE: The aim of this study was to systematically assess the predictive ability of multiple weight-related indicators (including body mass index (BMI), waist circumference (WC), and triglyceride-glucose index (TyG)) on cognitive decline, and to explore the differences in the sensitivity of these indicators in different gender populations. By comparing the predictive effects of each index, this study hopes to identify more clinically valuable screening tools for early detection of cognitive impairment, thus providing a scientific basis for public health interventions. METHODS: Based on the China Health and Retirement Longitudinal Study (CHARLS) database, 8,255 middle-aged and elderly subjects aged 45 to 98 years old who met the criteria were screened for a total of 14 weight-related indices, including BMI, WC, weight-adjusted waist circumference index (WWI), waist-to-height ratio (WHtR), TyG index and its derivatives (TyG-BMI, TyG-WC, TyG-WHtR), visceral adiposity index (VAI), body size index (ABSI), body roundness index (BRI), lipid accumulation index (LAP), and lipid accumulation index (LAP). WC, TyG-WHtR), visceral adiposity index (VAI), body size index (ABSI), body roundness index (BRI), lipid accumulation index (LAP), circularity index (CI), and Chinese visceral adiposity index (CVAI). The predictive ability of each index for cognitive impairment was assessed using the subject's work characteristic curve (ROC) analysis, and the optimal cut-off value, sensitivity, specificity and area under the curve (AUC) were calculated. In addition, associations between indicators and severe cognitive impairment (SCI) were analyzed using binary logistic regression models and adjusted for confounders such as age, blood pressure, and chronic diseases to improve the robustness of the models. RESULTS: Among the 8255 middle-aged and elderly subjects included, the AUC values of most weight-related indicators ranged from 0.500 to 0.590, with limited overall predictive power. Sex-stratified analysis showed that BMI (AUC=0.591), TyG-BMI (AUC=0.585), and WC (AUC=0.569) had high predictive value in the male group, while WWI (AUC=0.591), BMI (AUC=0.588), and TyG-BMI (AUC=0.572) performed relatively better. In a binary regression analysis based on the best cutoff values and after adjusting for confounders such as age, blood pressure, and chronic diseases, among men, BMI (OR=1.31, 95% CI: 1.12-1.54), WC (OR=1.21, 95% CI: 1.03-1.41), TyG-BMI (OR=1.36, 95% CI: 1.16- 1.60), TyG-WC (OR=1.23, 95% CI: 1.06-1.44), TyG-WHtR (OR=1.19, 95% CI: 1.02-1.39), VAI (OR=1.25, 95% CI: 1.07-1.46) and LAP (OR=1.24, 95% CI: 1.06-1.44) were all significantly and positively correlated with SCI (p < 0.05). In the female group, BMI (OR = 1.49, 95% CI: 1.26-1.76), WC (OR = 1.39, 95% CI: 1.17-1.65), TyG-BMI (OR = 1.43, 95% CI: 1.21-1.69), TyG-WC (OR = 1.40, 95% CI: 1.19-1.66) and LAP (OR = 1.36, 95% CI: 1.16-1.61) likewise showed significant associations (P < 0.05). The results indicated that compared with the traditional single weight indicators, TyG-BMI, TyG-WC, LAP and CVAI could reflect the weight distribution, fat accumulation and metabolic characteristics more comprehensively, and have a higher application value in the identification of SCI risk. CONCLUSION: This study shows that weight-related indicators have some application value in the prediction of SCI. Based on ROC analysis alone, the indicators had limited discriminatory ability. However, after stratified analysis based on optimal cutoff values and controlling for confounders, BMI, WC, TyG-related index, LAP and CVAI showed independent predictive ability in different gender groups. These indices, which can comprehensively reflect weight distribution, morphologic characteristics and lipid metabolism, are more clinically relevant in the early identification of SCI in middle-aged and elderly people than the traditional single weight indices. The findings emphasize that comprehensive weight-related indices can be used to screen people at high risk of cognitive impairment and provide a scientific basis for individualized health management and intervention. These findings can help promote the application of weight management in cognitive health and inform future public health strategies.