Prognostic value of average home blood pressure and variability: 19-year follow-up of the Didima study.

Abstract

The current general population study explored the prognostic value of home blood pressure (HBP) average and variability (BPV) versus office measurements (OBP). In 1997, 665 adults of Didima, Argolida, Greece were evaluated with OBP (two visits, six readings) and HBP (3 days, 12 readings) measurements. Total mortality and cardiovascular morbidity and mortality (CVD) were assessed after 19.0 ± 1.4 years. BPV was quantified by using SD and coefficient of variation. During follow-up, 216 deaths (124 cardiovascular) and 146 cardiovascular events (fatal and nonfatal) were documented. Hazard ratios for total mortality were 1.39/1.20 (P < 0.01/ < 0.01) per 10/5 mmHg increase in systolic/diastolic HBP and 1.36/1.02 (P < 0.01/NS) for OBP (similar hazard ratios for CVD). After adjustment for baseline risk factors, OBP and HBP lost their prognostic ability, except from diastolic OBP that predicted CVD (hazard ratio 1.10, P = 0.03). Sustained, masked and white-coat hypertension were associated with increased risk of death and CVD (P < 0.01 versus normotensive participants). After adjustment, the risk of death remained significant for all, but CVD only for white-coat hypertension. Systolic home BPV (not diastolic) predicted total mortality (adjusted hazard ratios 1.18/1.17 for 1-SD increase in SD/coefficient of variation; P < 0.05) and CVD. Unadjusted systolic/diastolic office BPV predicted CVD outcome (for total mortality only systolic), yet not after adjustment. Systolic home BPV predicted total mortality and CVD independent of office BPV. In this general population study with 19-year follow-up, average HBP predicted total mortality and CVD as reliably as OBP. Only normotension (low OBP and HBP) was associated with low risk. Systolic home BPV exhibited superior prognostic ability than OBP, deserving further research.