Fifteen RCTs (37,348 participants) were included. One study included 18,790 participants, others ranged from 75 to 4,733. Mean follow-up ranged from 1.6 to 12.2 years. At follow-up the weighted mean difference in blood pressure between the more intensive and less intensive targets was 7.5mmHg for systolic and 4.5 mmHg for diastolic blood pressure.
Sequence generation was adequate in 12 studies and unclear in two. Allocation concealment was adequate in 11 studies and unclear in three. Investigators were unblinded to treatment allocation in all studies. Participants were blinded in only three studies and outcome assessors were blinded in nine studies; blinding was unclear in five studies. Eight studies addressed incomplete outcome data and this was unclear in six studies. One study was missed from the quality assessment table. Tests did not show evidence of publication bias.
Compared to less intensive blood pressure lowering targets, more intensive targets reduced the risk of major cardiovascular events (RR 0.89, 95% CI 0.79 to 0.99; Ι²=28.2%; 10 trials), stroke (RR 0.76, 95% CI 0.63 to 0.92; Ι²=23.2%; 10 trials), end stage kidney failure (RR 0.89, 95% CI 0.82 to 0.97; Ι²=0%; eight trials) and myocardial infarction (RR 0.87, 95% CI 0.75 to 1.00; Ι²=0%; nine trials); myocardial infarction narrowly failed to reach statistical significance.
There was no effect on all-cause mortality (number of trials not reported), cardiovascular mortality (Ι²=23.4%; nine trials) and heart failure (Ι²=26.7%; nine trials).
In studies on people with diabetes, more intensive targets reduced the risk of albuminuria progression (RR 0.90, 95% CI 0.84 to 0.96; Ι²=0.0%; three trials). There was some reduction in progression of retinopathy (RR 0.81, 95% CI 0.66 to 1.00; Ι²=65.5%; four trials) although this just failed to reach statistical significance; when one study with imbalances in baseline characteristics between groups was removed heterogeneity was reduced and result was statistically significant in favour of more intensive targets (Ι²=18.1%).
Adverse events were reported inconsistently but where reported there were no statistically significant differences between the treatments for severe adverse events (five trials) and drug discontinuation (four trials). Information on other adverse effects were reported.
Subgroup analyses showed no evidence of different treatment effects for any particular subgroup. Meta-regression investigating possible effects of baseline characteristics showed no evidence of heterogeneity.