Twenty-five RCTs were included (n=45,758): nine placebo controlled (n=17,357) and 16 with active controls (n=28,401). Twenty-three were prospective and two were post-hoc analyses. All were double-blinded and used intention to treat analysis. Twenty-four RCTs were described as double-blinded.
All nephropathy: there was no statistically significant difference between the groups for any outcome when RAS blockade was compared with active control therapy (nine RCTs). However, there was a non-significant trend for RAS to increase the risk of stroke compared to active controls (p=0.05, five RCTs). RAS blockade significantly reduced the risk of overall CV events (RR 0.84, 95% CI: 0.78, 0.91, p=0.0001, five RCTs), heart failure (RR 0.74, 95% CI: 0.58, 0.95, p=0.02, two RCTs) and MI (RR 0.78, 95% CI: 0.65, 0.97, p=0.03, five RCTs) compared to placebo. But, there was statistically significant heterogeneity in the findings for overall CV events (p=0.02).
Proteinuria: RAS significantly reduced the risk of overall CV events (p=0.04, 10 RCTs) and heart failure (p=0.009, three RCTs) compared to active controls and significantly reduced the risk of heart failure compared to placebo (p=0.005, two RCTs).
Diabetic nephropathy: RAS significantly reduced the risk of heart failure compared to active controls (p=0.003, three RCTs) and significantly reduced the risk of overall CV events (p=0.009, six RCTs) and heart failure (p=0.003, three RCTs) compared to placebo. There was some indication of publication bias for these findings.
Non-diabetic nephropathy: RAS significantly reduced the risk of overall CV events and heart failure compared to active controls (p=0.001, eight RCTs).
Hypertensive nephropathy: there was a non-significant trend for RAS to increase the risk of stroke compared to active controls (p=0.05, four RCTs).
Full RRS and 95% CIs were reported for all the above comparisons. The results of other reported comparisons were not statistically significant.
Subgroup and sensitivity analyses reported no significant findings.