Alexander R van Rosendael1, Aukelien C Dimitriu-Leen1, Philippe J van Rosendael1, Melissa Leung1, Jeff M Smit1, Antti Saraste1, Juhani Knuuti1, Rob J van der Geest1, Britt W H van der Arend1, Erik W van Zwet1, Arthur J Scholte1, Victoria Delgado1, Jeroen J Bax2. 1. From the Department of Cardiology (A.R.v.R., A.C.D.-L., P.J.v.R., M.L., J.M.S., B.W.H.v.d.A., A.J.S., V.D., J.J.B.), Division of Image Processing (R.J.v.d.G.), Department of Medical Statistics and Bio-informatics (E.W.v.Z.), Leiden University Medical Center, The Netherlands; Netherlands Heart Institute, Utrecht, The Netherlands (A.R.v.R.); Department of Cardiology, Liverpool Hospital, University of New South Wales, Sydney, Australia (M.L.); and The Turku PET Centre, Finland (A.S., J.K.). 2. From the Department of Cardiology (A.R.v.R., A.C.D.-L., P.J.v.R., M.L., J.M.S., B.W.H.v.d.A., A.J.S., V.D., J.J.B.), Division of Image Processing (R.J.v.d.G.), Department of Medical Statistics and Bio-informatics (E.W.v.Z.), Leiden University Medical Center, The Netherlands; Netherlands Heart Institute, Utrecht, The Netherlands (A.R.v.R.); Department of Cardiology, Liverpool Hospital, University of New South Wales, Sydney, Australia (M.L.); and The Turku PET Centre, Finland (A.S., J.K.). j.j.bax@lumc.nl a.r.rosendael@lumc.nl.
Abstract
BACKGROUND: Epicardial adipose tissue located close to the atrial wall can change the electric conduction of the left atrium, potentially leading to atrial fibrillation (AF). The aim of this study was to assess whether an increased atrial adipose tissue mass posterior to the left atrium is related to AF independent of demographical and cardiovascular risk factors. METHODS AND RESULTS: Two hundred patients with AF and 200 patients without AF who underwent computed tomographic angiography were included. The posterior left atrial adipose tissue mass was quantified on computed tomographic angiography images as tissue with Hounsfield Units between -195 and -45. The adipose tissue mass was significantly larger in patients with AF compared with patients with sinus rhythm: 10.6±5.5 versus 4.7±3.5 g, P<0.001. In a multiple variable model (including age, body mass index, sex, coronary artery calcium score, diabetes mellitus, hypertension, hypercholesterolemia, family history of coronary artery disease, and known coronary artery disease), each gram increase of posterior left atrial adipose tissue was associated with 1.32 odds ratio of having AF (95% confidence interval, 1.22-1.43; P<0.001). Furthermore, the addition of the adipose tissue mass to the multiple variable analysis significantly increased the discriminatory ability to predict AF: increase in the area under the receiver operating characteristic, 0.88 (95% confidence interval, 0.84-0.91) versus 0.81 (0.76-0.85), P<0.001. CONCLUSIONS: Posterior left atrial adipose tissue mass is significantly larger in patients with AF versus without AF. An increase in adipose tissue was independently associated with AF and provided incremental value over well-known predictors of AF. These findings add to the hypothesis that the posterior left atrial adipose tissue mass contributes to structural and electric remodeling leading to AF.
BACKGROUND: Epicardial adipose tissue located close to the atrial wall can change the electric conduction of the left atrium, potentially leading to atrial fibrillation (AF). The aim of this study was to assess whether an increased atrial adipose tissue mass posterior to the left atrium is related to AF independent of demographical and cardiovascular risk factors. METHODS AND RESULTS: Two hundred patients with AF and 200 patients without AF who underwent computed tomographic angiography were included. The posterior left atrial adipose tissue mass was quantified on computed tomographic angiography images as tissue with Hounsfield Units between -195 and -45. The adipose tissue mass was significantly larger in patients with AF compared with patients with sinus rhythm: 10.6±5.5 versus 4.7±3.5 g, P<0.001. In a multiple variable model (including age, body mass index, sex, coronary artery calcium score, diabetes mellitus, hypertension, hypercholesterolemia, family history of coronary artery disease, and known coronary artery disease), each gram increase of posterior left atrial adipose tissue was associated with 1.32 odds ratio of having AF (95% confidence interval, 1.22-1.43; P<0.001). Furthermore, the addition of the adipose tissue mass to the multiple variable analysis significantly increased the discriminatory ability to predict AF: increase in the area under the receiver operating characteristic, 0.88 (95% confidence interval, 0.84-0.91) versus 0.81 (0.76-0.85), P<0.001. CONCLUSIONS: Posterior left atrial adipose tissue mass is significantly larger in patients with AF versus without AF. An increase in adipose tissue was independently associated with AF and provided incremental value over well-known predictors of AF. These findings add to the hypothesis that the posterior left atrial adipose tissue mass contributes to structural and electric remodeling leading to AF.