Abstract
BACKGROUND: Inflammation is one of the major pathways in the progression of hypertension (HTN), and the related inflammatory markers have demonstrated certain predictive values. The current study aimed to integrate these markers to construct an inflammatory prognostic scoring (IPS) system and to assess the prognostic values of IPS in patients with HTN. METHODS: A total of 9846 adult participants with HTN from NHANES 1999-2010 were enrolled and followed up. Demographic characteristics and the related laboratory results for the 12 inflammatory markers were collected. LASSO-COX regression, Kaplan-Meier, restricted cubic spline COX regression (RCS), receiver operator characteristic curve (ROC), and random survival forests (RSF) analysis were applied to explore the values of individual and IPS parameters. RESULTS: At the census date of follow-up, 2387 (24.2%) were identified as all-cause deaths and 484 (4.9%) as cardiovascular deaths. All inflammatory markers showed certain prognostic values. Then, based on the LAASO analysis, LDH, ALP, LYM, NLR, MLR, SIRI, and RDW were included in the construction of the IPS system. The higher IPS had significantly worse long-term prognosis in Kaplan-Meier analysis (p log-rank <0.001). Also, IPS remained an independent prognosticator compared to the lowest quartile (All p for trend <0.001), and the ROC showed satisfactory values in the long-term prognosis of both cardiovascular and all-cause mortality. RCS further showed a linear association of IPS with cardiovascular mortality and all-cause mortality (p for non-linearity >0.05). Two different algorithms of RSF, variable importance and minimal depth, to evaluate the prognostic importance showed that IPS was the best in survival prediction. CONCLUSION: Our results highlight that a higher IPS (system integrating the inflammatory markers) was associated with the increased risk of cardiovascular and all-cause mortality in patients with HTN, suggesting that IPS is a useful method for risk stratification in HTN.