Abstract
BACKGROUND: Chronic low-grade inflammation may mediate the relationship between obesity and diabetes, yet clinical research in this area remains scarce. Thus, this study aimed to explore the mediating role of chronic low-grade inflammation in this relationship using the National Health and Nutrition Examination Survey (NHANES). METHODS: This study involved 2,482 participants enrolled in the NHANES between 2005 and 2016. Based on the complex sampling survey weights of NHANES, logistic regression models were fitted, adjusting for various covariates to investigate the relationship between BMI, INFLA score, and diabetes. Moreover, weighted quantile sum (WQS) regression models were fitted to analyze the proportional contribution of individual components within the INFLA score. Finally, mediation analysis was conducted to quantitatively assess the magnitude of the mediating effect of the INFLA score on the relationship between BMI and diabetes. RESULTS: After adjusting for all potential confounding factors, a significant positive correlation was noted between INFLA score and diabetes [OR (95% CI), 1.038(1.003-1.075), p = 0.035]. Additionally, a significant positive correlation was observed between the high INFLA group and diabetes compared to the low INFLA group [OR (95% CI), 1.599(1.031-2.481), p = 0.037]. WQS regression models revealed that the proportional contributions of C-reactive protein, white blood cell count, platelet count, and neutrophil-to-lymphocyte ratio (NLR) were 55.5%, 34.8%, 8.46%, and 1.19%, respectively. Finally, the results of the mediation analysis indicated that the indirect effect of the INFLA score accounted for 10.20%. CONCLUSIONS: Chronic low-grade inflammation was associated with diabetes and partially mediates the relationship between obesity and diabetes.