Forty-nine RCTs were included: 12 compared ARBs with placebo (n=3,551); 9 compared ARBs with calcium-channel blockers (n=1,892); 23 compared ARBs with ACE inhibitors (n=1,301); 16 compared combination therapy (ARB plus ACE inhibitors) with ARBs (n=571); and 23 compared combination therapy with ACE inhibitors (n=711). Some studies contained more than one treatment comparison. The quality of the studies varied: 7 studies met all three quality criteria (allocation concealment, blinding and ITT analysis), 15 met two, 12 met one and 15 met none. For crossover studies, all but 3 studies reported a prolonged washout period or no statistically significant carryover effects.
ARB monotherapy.
For outcomes up to 4 months, ARBs achieved significantly lower proteinuria levels compared with placebo (means ratio 0.57, 95% confidence interval, CI: 0.47, 0.68; based on 8 comparisons) and calcium-channel blockers (means ratio 0.69, 95% CI: 0.62, 0.77; based on 7 comparisons), but there was no evidence of a difference in comparison with ACE inhibitors. For outcomes between 5 and 12 months, ARBs achieved significantly lower proteinuria levels compared with placebo (means ratio 0.66, 95% CI: 0.63, 0.69; based on 6 comparisons) and calcium-channel blockers (means ratio 0.62, 95% CI: 0.55, 0.70; based on 5 comparisons), but there was no evidence of a difference in comparison with ACE inhibitors.
Combined ARB and ACE inhibitor therapy.
For outcomes up to 4 months, combination therapy achieved significantly lower proteinuria levels compared with ARB alone (means ratio 0.76, 95% CI: 0.68, 0.85; based on 14 comparisons) and ACE inhibitors alone (means ratio 0.78, 95% CI: 0.72, 0.84; based on 21 comparisons). For outcomes between 5 and 12 months, combination therapy achieved significantly lower proteinuria levels compared with ARB alone (means ratio 0.75, 95% CI: 0.61, 0.92; based on 7 comparisons) and reduced levels compared with ACE inhibitors alone, although this result was not statistically significant.
Only the comparison of ARB with placebo for short-term outcomes showed substantial heterogeneity (I2=86%). Subgroup analyses of clinical characteristics did not show any possible reasons for this heterogeneity. However, larger treatment effects were seen in studies that did report allocation concealment compared with those that did not (p=0.016), and in crossover compared with parallel-group studies (p=0.02); these results were consistent for the monotherapy and combination therapy analyses. Larger treatment effects were also seen in the blinded studies compared with non-blinded studies (p=0.04) for the combination therapy analyses. There was no evidence of publication bias.