THE IMPORTANCE OF BLOOD PRESSURE VARIABILITY FOR SUBCLINICAL ORGAN DAMAGE AND CARDIOVASCULAR EVENTS. A LIFE SUBSTUDY: 8C.07

Background: In daily practice, antihypertensive treatment is based on the mean value of several blood pressure (BP) measurements due to high day-to-day variation of BP. However, it is unknown whether high BP variability may be harmful in treated hypertensive patients. Methods: Therefore, we tested in the LIFE study whether BP variability assessed as standard deviation (SD) and range for BP(6–24months) measured at 6, 12, 18 and 24 months of treatment 1) was associated with subclinical organ damage (SOD) defined by left ventricular hypertrophy on ECG and urine albumin/creatinine ratio, and 2) predicted the composite cardiovascular endpoint (CEP) of cardiovascular death, myocardial infarction and stroke occurring after 24 months. We excluded patients with CEP before two years of treatment and patients with < two BP(6–24 months) leaving 8505 patients with 630 CEPs. Results: Patients randomized to losartan- vs. atenolol-based treatment had slightly lower systolic BP(6–24 months) (149 vs. 150mmHg), lower SD (10.2 ± 6.0 vs. 10.9 ± 6.3mmHg), lower range (22.1 ± 13.2 vs. 23.7 ± 14.0mmHg) and higher diastolic BP(6–24 months) (85 vs. 84mmHg, all P < 0.001), but not different diastolic BP(6–24months) SD (5.5 ± 3.2 vs. 5.5 ± 3.1mmHg) nor range (11.8 ± 7.1 vs. 11.9 ± 6.8mmHg). Diastolic as well as systolic BP(6–24months) SD (βdiastolic = 0.07 and βsystolic = 0.33) and range (βdiastolic = 0.06 and βsystolic = 0.31) increased with increasing mean BP(6–24months) (all P < 0.001). After adjusting for mean BP(6–24months), neither high BP(6–24months) SD nor wide range were related to SOD. Independently of treatment allocation, mean BP(6–24months), Framingham risk score and SOD at baseline, CEP was predicted by diastolic BP(6–24months) SD (hazard ratio [HR] = 1.04[1.02–1.07], P < 0.01), range (HR = 1.02[1.01–1.03], P < 0.01), systolic BP(6–24months) SD (HR = 1.01[0.99–1.03], P = 0.07) and range (HR = 1.007[1.001–1.01] per mmHg, P = 0.02). Adjusted for the same factors, stroke was predicted by diastolic BP(6–24months) SD (HR = 1.06[1.03–1.10], P < 0.01), range (HR = 1.03[1.01–1.04], P < 0.01), systolic BP(6–24months) SD (HR = 1.02[1.002–1.04], P = 0.03) and range (HR = 1.009[1.001–1.02] per mmHg, P = 0.04), but and myocardial infarction was not. Conclusions: In treated LIFE patients, BP(6–24months) variability predicted CEP and stroke but not myocardial infarction independent of mean BP(6–24months), treatment allocation, Framingham risk score and SOD at baseline.