Exploring the association between cumulative hs-CRP and all-cause mortality, with consideration of cardiometabolic mediators in middle-aged and elderly adults: Insights from observational study

Abstract

Background

Under a state of sustained relatively low inflammation, it is uncertain if cumulative high-sensitivity C-reactive protein (cumhs-CRP) can predict all-cause mortality, and its impact on it through cardiometabolic factors is also unknown.

Methods

This is a longitudinal national cohort study using CHARLS data. Specifically, we only included participants aged 45 and older with hs-CRP levels below 10mg/L in both 2012 and 2015. Cumhs-CRP derived through time-weighted averaging was employed as the key exposure. Death events were recorded as the primary outcome. All-cause mortality is defined as death occurring at any point from 2015 to 2018. Variables selection and interpretation were conducted by Boruta and Shapley Additive Explanations (SHAP). Further mediation analysis explored the possibility of cardiometabolic factors serving as mediators. Additional Mendelian randomization was conducted to demonstrate the robustness of the association.

Results

5052 Chinese adults over 45 years of age with hs-CRP levels below 10mg/L over time were included. Boruta identified 16 possible variables, with SHAP confirming the importance of cumhs-CRP. Multivariate analysis showed a significant association between cumhs-CRP and mortality (OR 1.05, 95 % CI 1.02 1.09, p = 0.003), and a linear relationship was observed (p-nonlinear = 0.184). Importantly, cumhs-CRP indirectly may affect the risk of death through systolic and diastolic blood pressure, fasting glucose, and uric acid, with proportions of 17.4 %, 8.2 %, 5.2 %, and 17 %, respectively.

Conclusions

Cumhs-CRP levels are associated with increased mortality risk, even in individuals with lower baseline levels. Regular monitoring of hs-CRP levels may aid in identifying individuals with elevated mortality risk.