Evaluation of the association between the first observation and the longitudinal change in C-reactive protein, and all-cause mortality

Objective: To evaluate the association between vascular inflammation as measured by subacute C-reactive protein (CRP; 1-10 mg/l) and all-cause mortality and the association between change in CRP status (normal (3 mg/l and elevated >3 mg/l) and all-cause mortality. Methods: Probabilistic record linkage was used to match hospital episode data, laboratory reports and mortality statistics in a large urban population. Survival was evaluated using Cox proportional hazards regression models. Results: 22 962 patients had their first CRP measurement in the subacute range (1-10 mg/l). Analysis grouped by each additional unit increase in CRP across the subacute range was associated with a 7.3% (95% CI 5.4% to 9.2%) increase in the hazard ratio (HR) of death over 4 years, after controlling for confounding factors (p<0.001). Repeated CRP observations around 1 year apart were recorded in 5811 subjects. After controlling for confounding factors, in patients whose CRP changed from normal (<= 3 mg/l) to elevated (>3 mg/l), the HR increased 6.7-fold (p< 0.001) relative to cases whose CRP remained normal. By comparison, among those subjects whose CRP was reduced from elevated to normal, the hazard ratio halved to 3.5 (p= 0.018). In an underpowered analysis of time to cardiovascular events, an identical pattern of risk emerged. Conclusions: CRP level predicted all-cause mortality, and additional inclusion of prior change in CRP level and current CRP level more so. Increasing vascular inflammation, as measured by CRP, increases the likelihood of death.