418Pulse pressure and cardiovascular outcomes in high-risk individuals enrolled in the Systolic Blood Pressure Intervention Trial (SPRINT)

Abstract

Background: It is unclear whether the efficacy and safety of intensive blood pressure (BP) lowering varies by pulse pressure (PP), defined as the difference between systolic and diastolic BP. Patients with increased PP, which may track with age and arterial stiffness, may be at higher risk for adverse events with intensive blood pressure lowering. Purpose: To assess the relationship between baseline PP, treatment response to intensive BP lowering, and cardiovascular (CV) outcomes. Methods: SPRINT was a randomized, controlled trial in which 9,361 individuals ≥50 years of age at high CV risk but without diabetes who had a systolic BP 130‐180 mmHg were randomized to intensive (target systolic BP <120mmHg) or standard antihypertensive treatment (target systolic BP <140mmHg). The primary efficacy endpoint was the composite of myocardial infarction, other acute coronary syndromes, stroke, heart failure, or death from CV causes. The primary safety endpoint was the composite of serious adverse events (SAE). We examined the prognostic implications of baseline PP using 1) restricted cubic splines for the test of linear, non‐linear, and overall trends with clinical outcomes; 2) Cox proportional‐hazards regression models with PP quartiles, adjusted for demographic, clinical, and laboratory variables; and 3) restricted cubic splines for the effects of intensive BP lowering on clinical outcomes across the spectrum of PP values (test for interaction). SPRINT is registered at ClinicalTrials.gov. Results: Mean baseline PP was similar between the two study groups (intensive treatment 61 mmHg vs. standard treatment 62 mmHg; P=0.59). The four quartiles were defined as follows: quartile 1, ≤51 mmHg; quartile 2, 52‐60 mmHg; quartile 3, 61‐70 mmHg; quartile 4, ≥71 mmHg. Median follow‐up duration was 3.3 years (range 0‐4.8), with 562 composite primary efficacy events (6%) occurring over the course of the study. Except stroke, for which the association with PP was best defined as linear, PP displayed a non‐linear relationship with the risk of all tested clinical endpoints (test for non‐linearity, P<0.05; test for overall trend, P<0.05). The Figure shows the incidence rate of these endpoints as a function of PP. Although quartiles 2 and 3 generally appeared to harbor a lower risk than quartiles 1 and 4, no associations remained significant upon multivariable adjustment (P>0.05). However, the benefit of intensive BP management on mortality was greatest in patients with a PP ∼60 mmHg (P=0.03 for interaction). PP did not modify the risks and benefits of intensive BP lowering for other clinical endpoints (P>0.05 for interaction). Conclusions: In SPRINT, PP served as an easily calculated marker of CV risk and further identified patients who were at increased risk for SAE. However, the excess risk associated with PP was accounted for by major CV risk factors. Patients with a PP ∼60 mmHg appeared to have a greater mortality benefit from intensive BP lowering.