Determinants and Prognostic Value of Blood Pressure Trajectories During Graded Exercise

Objective: The systolic blood pressure (SBP) response to exercise reflects cardiovascular health, with both exaggerated and blunted responses being linked to adverse outcomes. Prior studies relied typically on peak SBP or two-point slopes for evaluation, but these metrics capture only a fraction of the SBP response. In contrast, we may gain deeper physiological and prognostic insights when analysing the full SBP trajectory (from rest to maximal exertion). We aimed to identify distinct SBP responses from full times-series recorded during a maximal, graded exercise test and assess their clinical determinants and prognostic relevance. Design and method: We retrospectively analysed SBP recordings from 6107 patients (mean age, 55.4 years; 45% women) who underwent maximal cycle ergometry. Group-based trajectory modelling (GBTM) extracted sex-specific responses from absolute SBP levels and the relative change (deltaSBP) traces. Associations with clinical factors and incident major adverse cardiovascular events (MACE) were assessed using ordinal logistic regression and Cox survival analyses, respectively. Results: Per sex, GBTM identified four distinct SBP and four deltaSBP trajectories (see Figure). Agreement between SBP and deltaSBP trajectory assignments was weak, suggesting the two metrics captured different components of the exercise SBP response. Higher SBP and lower deltaSBP trajectories were associated with adverse clinical profiles, including higher age and lower exercise capacity. Survival analyses revealed a graded increase in MACE incidence from low to high SBP response and from high to low deltaSBP rise (P<0.0001 for both). In men, high SBP and low deltaSBP independently predicted MACE in multivariable-adjusted analyses (adjusted hazard ratio (HR) relative to the cohort's average risk: 1.16 and 1.19, respectively). In women, only a mid-high SBP response was associated with increased MACE risk after full adjustment (adjusted HR: 1.22; P=0.007). In both sexes, patients presenting the combination of high average SBP response (high SBP) and low SBP rise (low deltaSBP) conferred the highest risk. Conclusions: Time-series analysis of SBP measurements during graded exercise revealed distinct response patterns with complementary value for MACE prediction (see Figure). Integrating SBP time-series analyses into exercise testing may refine cardiovascular risk stratification, particularly when integrating absolute and relative SBP dynamics.