Non-linear associations between blood pressure and clinical outcomes in atrial fibrillation: A systematic review and dose-response meta-analysis

BACKGROUND AND AIMS: Hypertension carries the highest attributable risk for atrial fibrillation (AF) and strongly influences clinical outcomes such as stroke and death; however, the optimal blood pressure targets in individuals with AF remain poorly defined. We aimed to characterise the dose-response relationship between systolic blood pressure (SBP) and diastolic blood pressure (DBP) and ischaemic stroke, all-cause mortality, and intracranial haemorrhage. METHODS: We systematically searched MEDLINE, Embase, and Scopus for studies reporting outcomes in AF stratified by SBP and DBP. A two-stage dose-response meta-analysis used restricted cubic splines to model non-linearity, with within-study correlation handled by the Greenland-Longnecker method. RESULTS: Twenty-three studies (17 cohorts; n = 476,750) were included. SBP demonstrated a U-shaped association with all-cause mortality, with the lowest risk at 134 mmHg (RR 0.92, 0.86-0.98 versus 120 mmHg). DBP exhibited a J-shaped association, with the lowest risk at 83 mmHg (RR 0.99, 0.97-1.02). In contrast to mortality, the risks of ischaemic stroke, intracranial haemorrhage, and stroke or systemic embolism rose progressively with SBP, with a steepening gradient. At 140 versus 120 mmHg, the RR was 1.51 for intracranial haemorrhage against 1.08 for ischaemic stroke, reaching 1.99 versus 1.14 at 150 mmHg. CONCLUSION: Systolic blood pressure exhibits a U-shaped relationship with all-cause mortality in patients with AF, balancing against a monotonic increase in ischemic stroke and intracranial haemorrhage at higher pressures. A pragmatic target SBP of 120-130 mmHg with DBP kept above approximately 75-80 mmHg, appears to optimize the balance between stroke prevention, bleeding risk, and survival.