Arterial Stiffness Mechanisms and Orthostatic Hypotension in SPRINT

Figures

• 

  Central Illustration. Balancing Cardiovascular Benefits and Serious Adverse Event Risk With Intensive Blood Pressure Control In this secondary analysis of 642 SPRINT participants, baseline pulse wave velocity was partitioned into structural and load-dependent components by adjusting to a 120/80 mm Hg reference blood pressure. While intensive blood pressure goals reduce cardiovascular events, they are often limited by concerns for serious adverse events. Our findings demonstrate that higher load-dependent stiffness—reflecting blood pressure-driven arterial loading—is independently associated with a significantly increased risk of adjudicated orthostatic hypotension and total serious adverse events. Importantly, these associations remained consistent regardless of whether patients were randomized to intensive (<120 mm Hg) or standard (<140 mm Hg) blood pressure targets. Identifying elevated load-dependent pulse wave velocity at baseline may serve as a valuable clinical tool to recognize vulnerable patients who require enhanced surveillance and personalized management strategies during intensive antihypertensive therapy to minimize treatment-related complications (adapted from Pewowaruk et al25). BP =…

• 

  Figure 1. Pulse Wave Velocity Components and Risk of Adverse Clinical Events Restricted cubic splines (with 95% CIs) showing the relationship between the exposures of baseline total pulse wave velocity (A and B), structural pulse wave velocity (C and D), and mean arterial pressure (E and F) with the clinical outcomes for all participants regardless of the treatment group. A, C, and E, The frequency of orthostatic hypotension events; B, D, and F, the frequency of SAEs per participant. The negative binomial regression models were covariate-adjusted for age, sex, race, smoking status, mean arterial pressure, number of blood pressure medications, serum creatinine, and prior cardiovascular disease and were also offset for the duration of follow-up for each participant. Continuous variables were modeled as restricted cubic splines to allow for nonlinear effects, and a log-rank test comparing models with and without each exposure was used to calculate P values. MAP = mean arterial pressure; OH = orthostatic hypotension; PWV = pulse wave velocity; SAEs = serious adverse events.

• 

  Figure 2. Arterial Stiffness and Adverse Events Stratified by Treatment Group Restricted cubic splines (with 95% CIs) showing the relationship by treatment group between the exposures of baseline total pulse wave velocity (A and B), structural pulse wave velocity (C and D), and load-dependent (load-dep.) stiffness (E and F) with the clinical outcomes. G and H, The interaction between baseline mean arterial blood pressure and treatment group on the frequency of orthostatic hypotension and SAEs. Negative binomial regression models included an exposure*treatment group interaction and were covariate-adjusted for age, sex, race, smoking status, mean arterial pressure, number of blood pressure medications, serum creatinine, and prior cardiovascular disease; models were also offset for follow-up duration. Continuous variables were modeled as restricted cubic splines to allow for nonlinear effects, with P values calculated using a log-rank test comparing models with and without the interaction term. Abbreviations as in Figure 1.

• 

  Figure 3. Arterial Stiffness and Adverse Events Stratified by Frailty Status Restricted cubic splines (with 95% CIs) showing the relationship by degree of frailty between the exposures of baseline total pulse wave velocity (A and B), structural pulse wave velocity (C and D), and load-dependent (load-dep.) stiffness (E and F) with the clinical outcomes. G and H, The interaction between baseline mean arterial blood pressure and frailty status on the frequency of orthostatic hypotension and serious adverse events. Splines were calculated at the median frailty index for each categorical frailty level (fit, prefrail, frail). Negative binomial regression models included an exposure∗frailty index interaction and were covariate-adjusted for age, sex, race, smoking status, mean arterial pressure, number of BP medications, serum creatinine, and prior cardiovascular disease; models were also offset for follow-up duration. Continuous variables were modeled as restricted cubic splines to allow for nonlinear effects, though the interaction term was limited to linear terms to minimize overfitting; P values were calculated using a log-rank test comparing models with and without the interaction.…