Abstract
OBJECTIVES: To examine whether the benefit of statins varied according to cardiovascular (CV) and non-CV mortality of the treated population. DESIGN: Meta-analysis and meta-regression of 16 randomized placebo-controlled trials. SETTING: Community and hospital. PARTICIPANTS: Statin- (n = 59,671) and placebo-treated (n = 59,707) individuals with and without CV disease (mean age 55 to 75). MEASUREMENTS: Meta-regression was used to model relative risks (RRs) of major CV events (myocardial infarction and stroke) and total mortality for statins versus placebo as a function of CV and non-CV mortality risks of the study population. RESULTS: Every 1% increase in 5-year non-CV mortality risk of the study population was associated with a 3.7% (95% confidence interval (CI) = 1.2 to 6.3%) greater RR of major CV events and a 4.4% (95% CI = 2.1 to 6.9%) greater RR of total mortality. (Higher RRs indicate smaller benefits.) CV mortality was not associated with statin effects (P > .05). In stratified analysis according to CV (≥5.3% vs <5.3%) and non-CV mortality (≥3.8% vs <3.8%) of the study population, statins had little mortality benefit in populations with high non-CV mortality, regardless of CV mortality (random-effects pooled RR = 0.81, 95% CI = 0.72 to 0.91, for low CV and low non-CV mortality; random-effects pooled RR = 0.90, 95% CI = 0.76 to 1.06 for low CV and high non-CV mortality; random-effects pooled RR = 0.79, 95% CI = 0.72 to 0.87 for high CV and low non-CV mortality; random-effects pooled RR = 0.94, 95% CI = 0.87 to 1.02 for high CV and high non-CV mortality). The CV event reduction was also attenuated in populations with high non-CV mortality (random-effects pooled RR = 0.67, 95% CI = 0.60 to 0.75, for low CV and low non-CV mortality; random-effects pooled RR = 0.73, 95% CI = 0.66 to 0.81 for low CV and high non-CV mortality; random-effects pooled RR = 0.77, 95% CI = 0.69 to 0.87 for high CV and low non-CV mortality; random-effects pooled RR = 0.83, 95% CI = 0.74 to 0.92 for high CV and high non-CV mortality). CONCLUSION: Benefits of statins may depend on the non-CV mortality risk of the treated population. This should be confirmed using individual-level data.