Optimal target of LDL cholesterol level for statin treatment: challenges to monotonic relationship with cardiovascular events

Background Aggressive lipid lowering by high-dose statin treatment has been established for the secondary prevention of coronary artery disease (CAD). Regarding the low-density lipoprotein cholesterol (LDL-C) level, however, the "The lower is the better" concept has been controversial to date. We hypothesized that there is an optimal LDL-C level, i.e., a "threshold" value, below which the incidence of cardiovascular events is no longer reduced. We undertook a subanalysis of the REAL-CAD study to explore whether such an optimal target LDL-C level exists by a novel analysis procedure to verify the existence of a monotonic relationship. Methods For a total of 11,105 patients with CAD enrolled in the REAL-CAD study, the LDL-C level at 6 months after randomization and 5-year cardiovascular outcomes were assessed. We set the "threshold" value of the LDL-C level under which the hazards were assumed to be constant, by including an artificial covariate max (0, LDL-C - threshold) in the Cox model. The analysis was repeated with different LDL-C thresholds (every 10 mg/dl from 40 to 100 mg/dl) and the model fit was assessed by log-likelihood. Results For primary outcomes such as the composite of cardiovascular death, non-fatal myocardial infarction, non-fatal ischemic stroke, and unstable angina requiring emergency hospitalization, the model fit assessed by log-likelihood was best when a threshold LDL-C value of 70 mg/dl was assumed. And in the model with a threshold LDL-C [greater than or equal to] 70 mg/dl, the hazard ratio was 1.07 (95% confidence interval 1.01-1.13) as the LDL-C increased by 10 mg/dl. Therefore, the risk of cardiovascular events decreased monotonically until the LDL-C level was lowered to 70 mg/dl, but when the level was further reduced, the risk was independent of LDL-C. Conclusions Our analysis model suggests that a "threshold" value of LDL-C might exist for the secondary prevention of cardiovascular events in Japanese patients with CAD, and this threshold might be 70 mg/dl for primary composite outcomes. Trial registration Keywords: LDL cholesterol, Target value, Threshold value, Statin, Coronary artery disease, Proportional hazard, Bottoming-out model