Age- and Sex-Related Changes in Central and Peripheral Pulse Wave Velocity

Abstract

PURPOSE: To compare the age-associated changes in central (elastic) and peripheral (muscular) arterial stiffness in healthy samples of women and men. METHODS: Subjects were nonsmoking, normally active groups of women (n=58, 21–79 yrs) and men (n=68, 20–79 yrs). After resting supine for ∼10 minutes, estimations of carotid-to-femoral (PWVcentral) and femoral-to-ankle (PWVle) pulse wave velocities and blood pressure (BP) were made simultaneously using a computerized device incorporating oscillometric cuff measurements and applanation tonometry (Colin VP2000, Colin Medical). Additional predictors (blood lipid, % body fat and VO2peak) were acquired on a separate day. Simple regressions for age, VO2peak, and each pulse wave velocity were performed. Stepwise regressions including age, VO2peak, blood pressure, % body fat and blood lipid data were also performed. ANCOVA models with VO2peak as the covariate were run to control for fitness. Significance was accepted at p <0.05 for simple regressions and ANCOVA and p <0.15 for stepwise regressions. RESULTS: In men PWVcentral and PWVleg increased linearly with age (all p <0.001). PWV, (p <0.001) and PWV (p = 0.06) decreased linearly with increasing VO2peak Stepwise regression indicated age, LDL cholesterol, VO2peak, and brachial systolic BP (r2=0.57) were significant predictors of PWVcentral. For PWVleg ankle and brachial systolic BP and VLDL cholesterol were the strongest predictors (r2=0.52). When controlling for VO2peak by ANCOVA, age was not a significant predictor of either PWVcentral (p=0.27) or PWVleg (p=0.75). In women PWVcentral, and PWVleg increased with age (all p <0.001) and were best described by a piecewise linear regression model with the slope of the relationship increasing after age 50 yrs. PWV, and PWVcentral and PWVleg both decreased linearly with increasing VO2peak (all p <0.001). Stepwise regression indicated that age, brachial systolic BP and % body fat were the strongest predictors (r2=0.61) of PWV. For PWV age, brachial diastolic BP, % body fat and VLDLcentral leg & cholesterol were the strongest predictors (r2=0.57). When controlling for VO2peak by ANCOVA, age remained a significant predictor of PWVleg (p=0.001) but not PWVcentral (p=0.11). CONCLUSIONS: Central and peripheral pulse wave velocities increase with advancing age in healthy, normally active women and men. However, there appears to be sexual dimorphism in this age-associated increase, with women exhibiting an augmentation in its slope after menopause. Further, there is a fitness-independent component to this increase in arterial stiffness in women. Supported by: NIH R01-AG-18246, and PSU Health and Human Dev. Seed Grant