A total of 104 studies were included: 66 RCTs and 38 cohort studies. All non-statin, non-placebo cohorts were ignored. The numbers of participants were provided for each statin within each outcome but, because the tables were not clear and some studies were included in more than one outcome, the total number of included participants could not be calculated.
The numbers of participants for each outcome were as follows: CRP 5,237, fibrinogen 5,202, homocysteine 2,906, LDL cholesterol oxidation 243, tPA 459, PAI 738 and platelet aggregation 300.
CRP.
Thirteen studies compared statins with placebo: individuals treated with statins had greater reductions in CRP (median reduction: 13 to 50%) than those on placebo. The mean or median absolute reductions were generally less than 1 mg/L. Twenty-five studies (22 RCTs and 3 cohort studies) reported data on CRP with various statins. In half of the cohort studies, the change from baseline was statistically significant. [A:With the exception of one outlier study that reported an increase in CRP level], for all studies, the median change was -0.5 mg/L (range: 0 to -3.10). The results were inconclusive in the 5 studies that directly compared different statins, although 2 studies found that atorvastatin, pravastatin and simvastatin all had similar effects on CRP levels. When indirect comparisons were made, no differences among statins were found.
No association between changes in CRP levels and changes in lipid levels was found in 8 of the 10 studies addressing this issue. One study found that lovastatin reduced the risk of coronary events in patients with elevated LDL cholesterol or CRP levels.
Fibrinogen.
Twenty-nine RCTs and 26 cohort studies reported this outcome. The use of statins had no visible effect on fibrinogen levels. In the 13 placebo-controlled trials, the pooled change across studies was 0.2 micromol/L (95% CI: -0.3, 0.88). When all studies were considered, no class effect of statins on fibrinogen was seen in any study, and the range of effect (change from baseline) was -3.5 to 2.4 micromol/L. A meta-analysis was performed on data from 58 cohorts, giving a summary estimate of -0.2 micromol/L (95% CI: -0.4, 0.1). Substantial heterogeneity, which could not be accounted for, was found. No difference in effect on fibrinogen was found between the five statins in the only study that compared all five concurrently. The summary estimates for the 13 studies that compared fluvastatin, pravastatin, or simvastatin with placebo were similar for each statin; none was statistically significant. Seven studies that investigated all five statins found no association between changes in lipid levels and changes in fibrinogen levels.
Homocysteine.
Six RCTs and one cohort study reported this outcome. These showed that statins do not substantially lower homocysteine levels; only one study showed a statistically significant decrease in homocysteine levels with statins compared with placebo (median decrease 0.2 micromol/L more than with placebo). No study directly compared statins; no association between homocysteine and LDL cholesterol was seen. One study reported a statistically significant reduction in cardiovascular events when lovastatin was given to people with raised LDL and homocysteine levels.
LDL cholesterol oxidation.
There was no consistent evidence from 11 studies (8 RCTs and 3 cohort studies) that statins are of benefit with respect to the oxidative capacity of LDL cholesterol. Only one trial compared pravastatin and simvastatin; no significant difference between these was found for any of the four indicators of LDL cholesterol oxidation. No study evaluated the association between LDL oxidation and lipid levels.
tPA.
Fourteen studies (8 RCTs and 6 cohort studies) reported the effects of statins on tPA antigen levels. Evidence suggests a beneficial effect on tPA antigen levels only with pravastatin (2 RCTs), with significant net decreases of 31 and 38 pmol/L compared with placebo. The other results were inconclusive, although there was a trend towards a reduction in tPA with statins (further details given in the paper).
PAI.
Thirteen RCTs and 9 cohort studies reported this outcome. There was no conclusive evidence on the beneficial effect of statins on PAI antigen or activity levels. The results of the controlled studies were inconsistent, and the uncontrolled cohort studies also showed inconclusive results. There was no evidence of a correlation between PAI antigen or activity levels and changes in lipid levels.
Platelet aggregation.
Two RCTs and 12 cohort studies reported this outcome. One RCT showed that the statin-treated group had a statistically significantly greater reduction in platelet aggregation than the placebo group. The evidence from the other studies was inconsistent, although it showed a trend towards a reduction in platelet aggregability. Studies could not be combined due to the different methods used to evaluate platelet aggregation.