Do estimated 24-h pulse pressure components affect outcome? The Ohasama study

Twenty-four-hour ambulatory pulse pressure (PP) is a powerful predictor of outcome. We attempted to apply the recently described PP components, an elastic (elPP), and systolic stiffening (stPP) components from 24-h ambulatory blood pressure (BP) monitoring (AMBP), and examine their influence on outcome in the Ohasama study population.

Included were participants of the Ohasama study without history of cardiovascular disease (CVD), who were followed-up for total and CVD mortality, and for stroke morbidity. The PP components were derived from 24-h SBP and DBP using a model based on the nonlinear pressure--volume relationship in arteries expressing pressure stiffness relationship. Outcome predictive power was estimated by Cox regression models; hazard ratio with 95% confidence interval (CI), applied to elPP, and stPP, adjusted for age, sex, BMI, smoking, alcohol drinking, diabetes mellitus, total cholesterol, antihypertensive treatment, and mean arterial pressure (MAP), whenever appropriate.

Of 1745 participants (age 61.4 ± 11.6, 65% women), 580 died, 212 of CVD, and 290 experienced a stroke during 17 follow-up years. PP was strongly correlated with elPP (r = 0.89) and less so with stPP (r = 0.58), and the correlation between the two components was weak (r = 0.15). After the adjustment, hazard ratio of PP per 1 SD increment for total mortality, CVD mortality, and stroke morbidity were 1.095 (95% CI 0.973-1.232), 1.207 (1.000-1.456), and 0.983 (0.829-1.166), respectively. Corresponding hazard ratios and 95% CIs were nonsignificant for elPP, and stPP. However, among participants with median pulse rate 68.5 bpm or less (median, n = 872), total (327 deaths) and CVD (131 deaths) mortality were predicted by elPP (per 1 SD increment), hazard ratio 1.231 (95% CI, 1.082-1.401), and 1.294 (95% CI, 1.069-1.566), respectively. In the subgroup of treated participants with hypertension and pulse rate 68.5 or less bpm (n = 309), total (177 deaths) and CVD (77 deaths) mortality were predicted by elPP, hazard ratio of 1.357 (95% CI, 1.131-1.628), and 1.417 (95% CI, 1.092-1.839), respectively. Stroke morbidity was not predicted by either PP or the PP components.

In a rural Japanese population, elPP but not stPP was predictive of total and CVD mortality even when adjusted for MAP and conventional risk factors in the subpopulation with slower pulse rate. This was mostly among the treated hypertensive patients.