Response to: Cardiovascular events with blood pressure lowering in patients with diabetes and systolic blood pressure below 140 mm Hg

Salam et al1 questioned the correctness of our conclusions2 because of inappropriate inclusion of some trials and missing relevant data from other trials. Here, we respond to them point by point. These authors suggest that ALTITUDE3 and VANEPHRON trial4 should be excluded in our recent meta-analysis.2 However, considering the following points, our study2 incorporated the data from ALTITUDE and VANEPHRON trial. First, although statistical weight given to dual renin-angiotensin-aldosterone (RAAS) therapy trials ALTITUDE and VANEPHRON was really large, the incorporation of ALTITUDE and VANEPHRON trial did not produce a significant increase of heterogeneity and the exclusion of the two trials did not significantly change the pooled estimate in analysis of main outcome, eliminating the potential bias on results caused by dual RAAS therapies. Second, it is unclear that to which extent the blood pressure (BP) reduction is appreciable degree and there seems to be no certain definition of appreciable degree of BP reduction. What is clear is that dual RAAS therapies do have a certain ability to reduce BP, as ALTITUDE and VANEPHRON trial both show that dual RAAS therapies can further reduce BP by about 1.5/0.6 mm Hg on the basis of standard therapy with RAAS blockade. Third, dual RAAS therapies were shown to increase the risks of hyperkalemia and acute kidney injury (AKI). However, our primary and secondary outcomes did not include hyperkalemia and AKI. Although ALTITUDE was stopped due to an increased risk of stroke, stroke is an outcome with clear pathophysiological linkage to BP. Finally, consistent with our study, previous two meta-analyses5, 6 also included ALTITUDE and VANEPHRON trial to evaluate the effects of BP lowering in patients with diabetes. Since neither the recent systemic review7 nor Salam et al1 provided the references for the 12 trials for 17974 (not 21044) patients with diabetes whose baseline SBP was below 140 mm Hg, we do not know exactly which trials they think we should include. But according to total trials included in this systematic review,7 we inferred that the 12 trials probably included CHARM-Alternative,8 DREAM,9 MERIT-HF,10 and NAVIGATOR trial.11 The four trials either enrolled diabetics with heart failure or performed on prediabetics. It should be noted that evidence from diabetics with heart failure and prediabetics can not inform our BP management in patients with diabetes, because diabetics with heart failure benefit from antihypertensive drugs in different mechanisms and diabetics and prediabetics have different pathophysiological processes. Our meta-analysis2 has a detailed explanation about the reason why evidence from diabetics with heart failure and from prediabetics can not inform our clinical practice of patient with diabetes in the paragraph 2 and 3 of “Discussion” section. Although in the recent systemic review,7 the relative risk (RR) for the composite outcome of coronary heart disease or stroke in diabetics was compatible with for all patients with baseline SBP <140 mm Hg, the RR of 0.90 (95% confidence intervals 0.81-1.01) for diabetics had no statistical significance, which is consistent with major cardiovascular outcome 1.01 (95% 0.93-1.10) of our meta-analysis.2 Additionally, the recent systemic review7 only provided the potential benefit of composite outcome but not that of individual outcomes, and this composite outcome was defined by authors themselves. In this case, it is difficult to determine whether the benefits are real rather than artificial. No conflict of interest.