The Leukocyte Subtype Counts and Ratios Can Effectively Predict the Risk of Arterial Stiffness Assessed by Cardio-Ankle Vascular Index: A Retrospective Study

Background: Arterial stiffness was the pathological basis and risk factor of cardiovascular diseases, with chronic inflammation as the core characteristic. We aimed to analyze the association between the arterial stiffness measured by cardio-ankle vascular index (CAVI) and indicators reflecting the inflammation degree, such as count of leukocyte subtypes, platelet, and monocyte-to-lymphocyte ratio (MLR), etc. Methods: The data of inpatients from November 2018 to November 2019 and from December 2019 to September 2020 were continuously collected as the training set (1,089 cases) and the validation set (700 cases), respectively. A retrospective analysis of gender subgroups was performed in the training set. The association between inflammatory indicators and CAVI or arterial stiffness by simple linear regression, multiple linear regression, and logistic regression was analyzed. The effectiveness of the inflammation indicators and the CAVI decision models to identify arterial stiffness by receiver operating curve (ROC) in the training and validation set was evaluated. Results: The effect weights of MLR affecting the CAVI were 12.87% in men. MLR was the highest risk factor for arterial stiffness, with the odds ratio (95% confidence interval) of 8.95 (5.04-184.79) in men after adjusting the covariates. A cutpoint MLR of 0.19 had 70% accuracy for identifying arterial stiffness in all participants. The areas under the ROC curve of the CAVI decision models for arterial stiffness were >0.80 in the training set and validation set. Conclusions: The MLR might be a high-risk factor for arterial stiffness and could be considered as a potential indicator to predict arterial stiffness.

Automated summary

The study suggests that MLR may be a high-risk factor for arterial stiffness, especially significant in men, with a cutoff point of MLR=0.19 being able to accurately identify arterial stiffness in all participants. The CAVI decision models showed good performance in identifying arterial stiffness with areas under the ROC curve >0.80 in both the training and validation sets.