Dynamic changes in anaemia and phenotype-specific risk of adverse outcomes in patients with atrial fibrillation treated with NOACs

Background Anaemia is a common comorbidity in patients with atrial fibrillation (AF) receiving oral anticoagulants. Although its association with bleeding is well established, its prothrombotic role and the prognostic relevance of specific anaemia phenotypes and their longitudinal trajectories remain unclear. Purpose To investigate the prognostic impact of anaemia phenotypes and their temporal trajectories on thromboembolic, mortality, and bleeding outcomes in AF patients initiating non–vitamin K antagonist oral anticoagulants (NOACs). Methods This prospective study included consecutive AF patients starting NOACs therapy between 2013 and 2019. Anaemia was defined according to World Health Organization criteria and classified as microcytic, normocytic, or macrocytic. Two co-primary outcomes were assessed: (i) a hard composite event (including all-cause death and cardiovascular events) and (ii) clinically relevant bleeding (CRB). Secondary outcomes were the individual components of the hard composite. Associations were evaluated using multivariable Cox regression models. In addition, landmark analyses were performed to account for temporal changes in anaemia status, classifying patients as having incident, persistent, resolved, or stable anaemia. Results Among 940 patients (median age 80.2 years; 49% women), hard composite event occurred in 345 (36.7%) and CRB in 249 (26.5%) during 3-year follow-up. After multivariable adjustment, microcytic, normocytic, and macrocytic anaemia were independently associated with increased risk of the hard composite (HR 2.44, 95% CI 1.62–3.67; HR 1.60, 95% CI 1.26–2.03; HR 4.92, 95% CI 2.65–9.14, respectively). Normocytic anaemia was independently associated with risk for the CRB (HR 1.33, 95% CI 1.00–1.75). In trajectory analyses, both incident and persistent anaemia, but not recovered anaemia, were independently associated with a higher risk of the hard composite (HR 2.54, 95% CI 1.75–3.67 and HR 2.91, 95% CI 2.15–3.92, respectively) and the bleeding composite (HR 1.85, 95% CI 1.20–2.85 and HR 2.17, 95% CI 1.54–3.04), compared with patients with stable non-anaemic status. Conclusions In AF patients treated with NOACs, anaemia identifies a subgroup at markedly higher risk of adverse outcomes, with a gradient of risk across anaemia phenotypes. Dynamic changes in haemoglobin provide incremental prognostic information, as incident and persistent anaemia are associated with substantially worse prognosis, whereas resolution mitigates risk.