From the University of Milano-Bicocca and IRCCS Istituto Auxologico Italiano, Italy (G.M.); Statistical Consultant, Ingelheim, Germany (H.S.); Klinik für innere Medizin III, Universitätsklinikum des Saarlandes, Homburg/Saar, Germany (M.B.); Hypertension Clinic, Department of Internal Medicine, Hospital Clinico Universitario de Valencia INCLIVA, University of Valencia and CIBERObn, ISCIII, Madrid, Spain (J.R.); Nephrologie und Hypertensiologie, Universitätsklinikum Erlangen, Erlangen, Germany (R.E.S.); Struttura Complessa di Medicina, Ospedale di Assisi, Assisi (PG), Italy (P.V.); CV Medicine, John Radcliffe Hospital, Oxford, United Kingdom (P.S.); and Population Health Research Institute, McMaster University, Hamilton, Canada (K.T., S.Y.).
Hypertension. 2017 Nov;70(5):938-948. doi: 10.1161/HYPERTENSIONAHA.117.09714. Epub 2017 Oct 3.
In 28 790 patients recruited for the ONTARGET (Ongoing Treatment Alone and in Combination With Ramipril Global End Point Trials) and TRANSCEND (Telmisartan Randomized Assessment Study in ACE Intolerant Subjects With Cardiovascular Disease) trials, we investigated the prognostic value for cardiovascular events (primary outcome) of (1)on-treatment visit-to-visit systolic blood pressure (SBP) variability versus mean SBP and (2) the 2 measures together. SBP variability was measured by the coefficient of variation (CV) of mean SBP to which it was unrelated. Confounders such as variable time and number of visits from which to calculate SBP-CV were avoided by using the same number of visits at identical times in all patients. The covariate-adjusted risk of the primary outcome (Cox models) increased as SBP-CV or mean on-treatment quintile SBP increased, but only for mean on-treatment SBP, the relationship achieved statistical significance: global test for trend, =0.12 versus <0.0001. SBP-CV showed a relationship with fatal events, but it was unrelated to the risk of myocardial infarction and stroke, which were predicted by on-treatment mean SBP. Prediction of the primary outcome improved by the combined use of both measures: global test for trend, <0.0001; hazard ratio for combined fifth versus first quintile, 1.42 (1.20-1.68) compared with 1.13 (1.01-1.27) for SBP-CV and 1.24 (1.11-1.40) for mean SBP. Thus, in the present study, on-treatment mean SBP provided an overall better prediction of cardiovascular risk than visit-to-visit SBP-CV. Prediction improved by their combined use, which may thus offer a more precise estimate of the protective effect of treatment.
URL: http//www.clinicaltrial.gov. Unique identifier: NCT00153101
在 ONTARGET（单独治疗和与雷米普利联合的全球终点试验）和 TRANSCEND（不耐受 ACE 的心血管疾病患者的替米沙坦随机评估研究）试验中，我们招募了 28790 名患者，研究了心血管事件（主要结局）的预后价值，包括（1）治疗期间的收缩压（SBP）变异性与平均 SBP，以及（2）这两个指标的共同作用。SBP 变异性通过平均 SBP 的变异系数（CV）来测量，与平均 SBP 无关。通过在所有患者中使用相同数量的访视时间和数量来避免与计算 SBP-CV 相关的变量时间和访视次数，从而避免了混杂因素。调整协变量后，主要结局（Cox 模型）的风险随着 SBP-CV 或治疗期间平均 SBP 五分位数的增加而增加，但仅对于治疗期间的平均 SBP，这种关系才具有统计学意义：整体趋势检验，=0.12 对<0.0001。SBP-CV 与致命事件有关，但与心肌梗死和中风无关，后者由治疗期间的平均 SBP 预测。使用两种测量方法的联合使用提高了主要结局的预测能力：整体趋势检验，<0.0001；与 SBP-CV 的第五五分位与第一五分位相比，危险比为 1.42（1.20-1.68），与平均 SBP 的 1.13（1.01-1.27）相比；与平均 SBP 的 1.24（1.11-1.40）相比。因此，在本研究中，治疗期间的平均 SBP 比治疗期间的收缩压变异性（CV）提供了更好的心血管风险预测。联合使用两种测量方法可提高预测能力，从而更准确地估计治疗的保护作用。
