Independent and joint effect of central and brachial SBP on incident stroke in hypertensive adults

Objective: It is unclear whether central SBP (cSBP) is an independent predictor of stroke above and beyond brachial SBP (bSBP). This study aimed to investigate the difference between cSBP and bSBP in predicting first stroke and the joint effect of cSBP and bSBP on the risk of first stroke in hypertensive adults. Methods: A total of 8122 hypertensive adults without stroke history were included in this study. cSBP was measured noninvasively using A-Pulse CASPro device. The outcome was first stroke. During a median follow-up of 4.4 years, 579 first strokes were identified. Results: A total of 8122 hypertensive adults without stroke history were included in this study. cSBP was measured noninvasively using A-Pulse CASPro device. The outcome was first stroke. During a median follow-up of 4.4 years, 579 first strokes were identified. The risk of first stroke increased by the same 16% [hazard ratio: 1.16, 95% confidence interval (95% CI): 1.07–1.26] for per SD increment in both cSBP and bSBP. The differences in areas under the curves, continuous net reclassification indices, and integrated discrimination indices of bSBP and cSBP models for predicting first stroke were 0.003 (95% CI: −0.003, 0.008), 0.007 (95% CI: −0.058, 0.071), and −0.0002 (95% CI: −0.0028, 0.0013), respectively. When cSBP and bSBP were evaluated jointly, participants in the highest tertiles of both cSBP and bSBP had the highest risk of first stroke compared with their counterparts (hazard ratio: 1.59, 95% CI: 1.29–1.96; P -interaction = 0.034). Similar results were found for ischemic stroke and hemorrhagic stroke. Conclusion: Although cSBP was not found to be superior to bSBP in predicting first stroke, cSBP and bSBP were jointly associated with the risk of first stroke among hypertensive adults.