Abstract 4140003: Impact of Long-Term Blood Pressure Variability on Adverse Cardiovascular Outcomes in High- and Low-Risk Populations

Background: Hypertension is a major preventable risk factor for cardiovascular (CV) disease. Emerging evidence suggests that in addition to blood pressure (BP) levels, controlling the consistency of BP is a key determinant of clinical outcomes. We aimed to assess the effects of consistency of BP control on adverse CV and renal outcomes using two metrics: long-term variability of systolic BP (LT-BPV) and the degree at which BP control is achieved, known as cumulative systolic BP load (CBPL). Methods: We collected clinic systolic BP (SBP) measurements from UK Biobank primary care records in those diagnosed with hypertension (N=39,816), chronic kidney disease (CKD, N=8,062), and neither of these (N=17,702), including a per-participant mean of 21 SBP values over 6 years. Instances of the primary outcome, 4-point major adverse cardiovascular events (MACE; stroke, acute myocardial infarction, heart failure hospitalization, and cardiovascular-related death), were collected from hospital records. Per-standard deviation (SD) and per-quartile hazard ratios (HR) were used to estimate LT-BPV, CBPL, and mean SBP separately using adjusted Cox regression. Sensitivity analyses were used to determine the independence of BP consistency effects from mean BP levels. Results: In those with hypertension, each per-SD increase in LT-BPV was associated with increased risk for MACE (HR=1.12, p=4.4E-11), specifically CV-related death, stroke, and HF hospitalization. LT-BPV effect estimates were strongest in those with controlled SBP and the strongest predictors for all events in those without high-risk comorbidities (coronary artery disease, CKD, peripheral vascular disease, and diabetes). CBPL was similarly associated with MACE (HR=1.13, p=8.1E-13), specifically stroke and acute myocardial infarction, but not with fatal events or HF. In those with CKD, both LT-BPV and CBPL were associated with progression to renal failure (HR=1.31-1.33, p<6.8E-11), but only LT-BPV was associated with MACE and mortality. CBPL showed a J-shaped mortality distribution in all populations, whereas LT-BPV consistently showed a stepwise increase in fatalities across quartiles. Conclusions: LT-BPV was a robust predictor of adverse CV outcomes across all populations, especially in typically “lower-risk” populations where the need for preventative measures would not usually account for LT-BPV. Our study indicates the importance of controlling the consistency of BP to optimize protection against CV events.