Fifty RCTs with 54 comparisons (N = 30,144) were included in the review. 26 RCTs assessed statins in pre-dialysis patients, 11 in dialysis patients, and 17 in transplanted patients. Study quality was suboptimal in many instances: allocation concealment was adequate in only 11 trials, triple-blinding in 10 trials, and an ITT analysis was performed in 10 trials. Dropout rates were low, being less than 10% in 43 trials. Sample size ranged from 6 to 16,824.
There was no statistically significant difference between statins and placebo in overall mortality (RR 0.92, 95% CI: 0.82, 1.03, no statistically significant heterogeneity), although an isolated significant effect was observed in the pre-dialysis population. Significant reductions were found in cardiovascular mortality (RR 0.81, 95% CI: 0.73, 0.90, 43 RCTs, N = 23,266) and non-fatal cardiovascular events (RR 0.78, 95% CI: 0.73, 0.84, 8 RCTs, N = 22,863) in the statin groups, with no statistically significant heterogeneity. These effects were consistent across the population groups.
Statistically significant reductions relative to placebo were found in total cholesterol (WMD -42.28 mg/dl, 95% CI: -47.25, -37.32, 42 comparisons, N = 6390), low density lipoprotein cholesterol (WMD -43.12 mg/dl, 95% CI: -47.85, -38.40, 39 comparisons, N = 6216) and triglycerides (WMD -23.71 mg/dl, 95% CI: -33.52, -13.90, 39 comparisons, N = 5569). 24 hour protein excretion was significantly reduced in pre-dialysis patients (WMD -0.73 g/24 hr, 95% CI: -0.95, -0.52, 6 RCTs, N = 311). There were no significant differences in high density lipoprotein cholesterol or in creatinine clearance. Significant statistical heterogeneity was present in all these analyses.
There were no significant differences in the incidence of adverse events or withdrawals due to adverse events, and no significant heterogeneity in these analyses. In the pos-transplant population there were no statistically significant differences in the risk of acute allograft rejection between the groups, but significant heterogeneity was detected.
The results of meta-regression analyses were also reported.