Abstract
AIMS/HYPOTHESIS: We investigated whether atorvastatin 10 mg daily lowered C-reactive protein (CRP) and whether the effects of atorvastatin on cardiovascular disease (CVD) varied by achieved levels of CRP and LDL-cholesterol. METHODS: CRP levels were measured at baseline and 1 year after randomisation to atorvastatin in 2,322 patients with type 2 diabetes (40-75 years, 69% males) in a secondary analysis of the Collaborative Atorvastatin Diabetes Study, a randomised placebo-controlled trial. We used Cox regression models to test the effects on subsequent CVD events (n = 147) of CRP and LDL-cholesterol lowering at 1 year. RESULTS: After 1 year, the atorvastatin arm showed a net CRP lowering of 32% (95% CI -40%, -22%) compared with placebo. The CRP response was highly variable, with 45% of those on atorvastatin having no decrease in CRP (median [interquartile range, IQR] per cent change -9.8% [-57%, 115%]). The LDL-cholesterol response was less variable, with a median (IQR) within-person per cent change of -41% (-51%, -31%). Baseline CRP did not predict CVD over 3.8 years of follow-up (HRper SD log 0.89 [95% CI 0.75, 1.06]), whereas baseline LDL-cholesterol predicted CVD (HRper SD 1.21 [95% CI 1.02, 1.44]), as did on-treatment LDL-cholesterol. There was no significant difference in the reduction in CVD by atorvastatin, with above median (HR 0.57) or below median (HR 0.52) change in CRP or change in LDL-cholesterol (HR 0.61 vs 0.50). CONCLUSIONS/INTERPRETATION: CRP was not a strong predictor of CVD. Statin efficacy did not vary with achieved CRP despite considerable variability in CRP response. The use of CRP as an indicator of efficacy of statin therapy on CVD risk in patients with type 2 diabetes is not supported by these data. Trial registration NCT00327418.