and clinical outcomes but it has not been investigated as a measure of outcomes in atrial fibrillation (AF) patients. Methods We performed a post-hoc analysis of the AFFIRM trial including patients with available echo data. Patients were stratified based on gender-adjusted echo derived interventricular septal (IVS) thickness, relative wall thickness (RWT), gender-adjusted LV mass and type of LV remodelling (normal LV geometry, concentric hypertrophy, eccentric hypertrophy, and concentric

more »

... ling). Cox proportional hazards models were used for multivariate analyses of time to death and time to ischaemic stroke. Results Of 4060 patients recruited in AFFIRM, sufficient echo data were available in 2433 patients (60%). Multivariate analysis showed that moderate-severe LV (IVS diastolic dimension >1.2 cm for women, >1.3 cm for men) was associated with all cause mortality (HR 1.45, 95% CI 1.13 to 1.86, p¼0.003). Concentric LV hypertrophy was associated with the worst outcome (defined as RWT>0.42 and LV mass >224 g for men or LV mass >162 g for women) (p¼0.008 vs, normal geometryddefined as RWT #0.42 and LV mass #224 g for men or LV mass #162 g for women). In a multivariate model, including established clinical, demographic and echo risk factors, moderate-severe LV hypertrophy assessed by IVS thickness was the strongest echo predictor of stroke (HR 2.2, 95% CI 1.3 to 3.7, p¼0.002). Conclusion In the AFFIRM Trial, LV hypertrophy assessed by gender-adjusted IVS thickness is an important risk factor for ischaemic stroke in patients with AF. LV hypertrophy assessed by gender-adjusted IVS thickness is associated with increased all cause mortality in AF patients.