Abstract
BACKGROUND: Blood pressure (BP), autonomic function and atherosclerosis are associated with adverse bone-related outcomes. These entities may affect blood pressure variability (BPV). The few studies that have directly investigated the association of BPV with hip fracture or bone mineral density (BMD) were limited to Asian populations and reported few BPV measures.
METHODS: Individuals with BP measurements at ≥4 of the 6 Cardiovascular Health Study (CHS), a large, prospective, observational cohort, visits between 1989 and 90 and 1994-95 and who attended the 1994-95 visit were included. Systolic (SBP), diastolic (DBP) and pulse pressure (PP) variability were estimated as: between-visit mean, between-visit slope (linear trajectory over time) and between-visit standard deviation (SD) of the observed residuals (departure from linear trajectory). Incident hip fractures between 1994 and 95 and 2015 were recorded. The association of BPV with hip fracture was estimated with Cox models (95% confidence intervals, CI). Primary analyses were conducted in individuals not using BP-lowering medications at baseline. Secondary analyses were conducted in those with stable use of BP-lowering medications during look back. Total hip BMD was measured at the 1994-95 study visit in a subset of individuals and analysed for cross-sectional associations with BPV.
RESULTS: Among 1820 individuals not using BP-lowering medications at baseline (60% women; age 76 ± 5 years), 292 incident hip fractures (222 women, 70 men) occurred. After multivariable adjustment, a one SD increase (3.12 mmHg) in DBP residual was associated with an 8.9% increased relative risk (95%CI, 0.4% to 18.2%) in men only. No other BPV measures were associated with hip fractures in men or women. No BPV measure was associated with BMD. Among 1958 individuals with stable BP medication use, 297 incident hip fractures (229 women, 68 men) occurred. In this group, no BPV measure was associated with incident hip fracture or BMD.
CONCLUSION: We found no consistent association of BPV with BMD or hip fracture, which did not validate results from prior studies. Low BPV in this cohort may have limited our ability to identify associations with adverse skeletal outcomes but also may reflect limited clinical importance of BPV for bone.