Physical performance changes as clues to late-life blood pressure changes with advanced age: the osteoporotic fractures in men study

Objectives

This study examined whether changes in late-life physical performance are associated with contemporaneous changes in blood pressure (BP) in older men.

Design

prospective cohort study over 7 years.

Setting and Participants

Physical performance (gait speed, grip strength, chair stand performance) and clinic-measured BP at baseline and at least one follow-up (year 7 or 9) were assessed in 3,135 men aged ≥65 y enrolled in the Osteoporotic Fractures in Men Study (MrOS).

Methods

Generalized estimating equation analysis of multivariable models with standardized point estimates (β [95% CI]) described longitudinal associations between physical performance and BP changes in participants overall, and stratified by baseline cardiovascular disease (CVD), antihypertensive medication use (none, ≥1), and enrollment age (<75 years; ≥75 years).

Results

Overall, positive associations (z-score units) were found between each increment increase in gait speed and systolic (SBP) (0.74 [0.22, 1.26]) and grip strength (0.35 [0.04, 0.65]) or gait speed (0.55 [0.24, 0.85]) with diastolic (DBP). Better grip strength and chair stand performance over time were associated with 1.83 [0.74, 2.91] and 3.47 [0.20, 6.74] mmHg higher SBP, respectively in men with CVD at baseline (both interaction P < .05). Gait speed increases were associated with higher SBP in men without CVD (0.76 [0.21, 1.32]), antihypertensive medication non-users (0.96 [0.30, 1.62]), aged <75 years (0.73 [0.05, 1.41]) and ≥75 years (0.76 [0.06, 1.47]). Similar positive, but modest associations for DBP were observed with grip strength in men with CVD, antihypertensive medication non-users, and aged <75 years, and with gait speed in men without CVD, aged <75 years, and irrespective of antihypertensive medication use.

Conclusion

In older men, better physical performance is longitudinally associated with higher BP. Mechanisms and implications of these seemingly paradoxical findings, which appears to be modified by CVD status, antihypertensive medication use, and age, requires further investigation.