Effects of blood-pressure-lowering treatment on outcome incidence in hypertension. 11. Effects of total cardiovascular risk and achieved blood pressure: overview and meta-analyses of randomized trials

Background: In recent meta-analyses of blood pressure (BP)-lowering randomized controlled trials (RCTs), we have shown that in hypertensive patients with diabetes, but not in those without, relative risk reduction of cardiovascular outcomes for a standardized BP reduction is significantly smaller at progressively lower SBP values achieved by treatment. Objectives: Whether this feature is typical of diabetes or is common to all hypertensive patients at high-very high cardiovascular risk is unknown. To clarify these points, we report a new set of meta-analyses, in which BP-lowering RCTs have been stratified in a double way, according to two levels of cardiovascular risk (below and above 5% cardiovascular death in 10 years) and three SBP levels attained by treatment (≥140, 130-139, and <130 mmHg). Methods: The database consisted of 72 BP-lowering RCTs including 260 210 patients, stratified in two ways (cardiovascular risk and achieved SBP) as indicated above. Risk ratios and 95% confidence intervals of six fatal and nonfatal cardiovascular outcomes and all-cause death were calculated (random effects model) for all patients and, separately, for those with and those without diabetes mellitus. Differences between treatment effects at different achieved SBP levels were evaluated by test of homogeneity or trend analysis. Results: When all patients at higher cardiovascular risk were analyzed (46 RCTs, 182 248 patients), no significant difference could be found in the relative risk reduction of any outcome in response to a standard BP reduction at any level of achieved SBP. On the other hand, in patients at a high level of cardiovascular risk, the presence of diabetes (29 RCTs, 52 350 patients) was associated with a significantly smaller outcome benefit of a standardized BP lowering to SBP less than 130 mmHg, and the opposite was found in absence of diabetes (22 RCTs, 102 792 patients). Similar findings were obtained in lower cardiovascular risk patients, but the smaller number of trials and, particularly, events weakens the evidence they provide, particularly on lower risk patients with diabetes. Conclusion: A high level of cardiovascular risk is not in itself a restraint to target at SBP values less than 130 mmHg, if treatment is well tolerated. Though a high cardiovascular risk associated with diabetes is not an indication for aiming at SBP less than 130 mmHg, current evidence is that SBP values slightly below 130 mmHg are not associated with harm.