Atrial fibrillation and the risk of incident dementia: A meta-analysis Pasquale Santangeli, MD, Luigi Di Biase, MD, PhD, FHRS, Rong Bai, MD, FHRS, Sanghamitra Mohanty, MD, Agnes Pump, MD, Mauricio Cereceda Brantes, MD, Rodney Horton, MD, J. David Burkhardt, MD, FHRS, Dhanunjaya Lakkireddy, MD, FHRS, Yeruva Madhu Reddy, MD, Michela Casella, MD, Antonio Dello Russo, MD, Claudio Tondo, MD, Andrea Natale, MD, FHRS Heart Rhythm Volume 9, Issue 11, Pages 1761-1768.e2 (November 2012) DOI: 10.1016/j.hrthm.2012.07.026 Copyright © 2012 Heart Rhythm Society Terms and Conditions
Figure 1 Selection process of studies included in the meta-analysis. Heart Rhythm 2012 9, 1761-1768.e2DOI: (10.1016/j.hrthm.2012.07.026) Copyright © 2012 Heart Rhythm Society Terms and Conditions
Figure 2 Forest plot showing the individual and pooled adjusted hazard ratios (HR) of incident dementia comparing patients with atrial fibrillation (AF) vs those without AF. Square boxes denote HR. The dimension of each square box denotes the weight from random effect analysis. Horizontal lines represent 95% confidence interval (CI). The study by Rastas et al14 did not report the adjusted HR and was included for sensitivity analysis (see Methods). Heart Rhythm 2012 9, 1761-1768.e2DOI: (10.1016/j.hrthm.2012.07.026) Copyright © 2012 Heart Rhythm Society Terms and Conditions
Figure 3 Effect of single study withdrawal on the pooled hazard ratio (HR) of incident dementia comparing patients with atrial fibrillation (AF) vs those without AF. Presence of AF consistently increased the risk of incident dementia also after excluding each study in turn. Heart Rhythm 2012 9, 1761-1768.e2DOI: (10.1016/j.hrthm.2012.07.026) Copyright © 2012 Heart Rhythm Society Terms and Conditions