Atrial fibrillation and the risk of incident dementia: a meta-analysis

Heart Rhythm. 2012 Nov;9(11):1761-8. doi: 10.1016/j.hrthm.2012.07.026. Epub 2012 Aug 2.


Background: The risk of cerebrovascular embolic events with atrial fibrillation (AF) is potentially linked to an increased risk of cognitive decline and dementia. However, epidemiologic studies exploring the association between AF and incident dementia have reported conflicting results.

Objective: The purpose of this study was to perform a meta-analysis of observational studies specifically designed to evaluate the prospective relationship between AF and incident dementia.

Methods: We searched PubMed, CENTRAL, BioMedCentral, Embase, Cardiosource,, and ISI Web of Science (January 1980 to May 2012). No language restriction was applied. Two independent reviewers screened titles and abstracts to identify population-based studies that prospectively evaluated the association between AF and the incidence of dementia in patients not suffering an acute stroke and with normal cognitive function at baseline, providing the hazard ratio (HR) obtained in multiple Cox regression analyses, and adjusted for all confounding variables. Two independent reviewers assessed risk of bias according to the Cochrane Collaboration, and extracted patient and study characteristics and the adjusted HR of incident dementia with its 95% confidence interval (CI) of patients with AF vs those without AF.

Results: Eight studies with 77,668 patients were included in the analysis. All studies had a prospective observational design and included elderly patients (mean age range 61-84 years) with normal cognitive function at baseline, of whom 11,700 (15%) had AF. After a mean follow-up of 7.7 ± 9.1 years (range 1.8-30 years), 4773 of 73,321 (6.5%) patients developed dementia. Two studies did not report the rates of dementia at follow-up but reported the adjusted HR and were included in the pooled analysis. At pooled analysis adjusted for baseline confounders and covariates, AF was independently associated with increased risk of incident dementia (HR = 1.42 [95% CI 1.17-1.72], P <.001).

Conclusion: AF is independently associated with increased risk of dementia. Patients with AF should be periodically screened for dementia, which should be included among the outcomes assessed in AF treatment trials.

Publication types

  • Meta-Analysis

MeSH terms

  • Aged
  • Aged, 80 and over
  • Atrial Fibrillation / complications*
  • Dementia / epidemiology*
  • Dementia / etiology*
  • Female
  • Humans
  • Incidence
  • Male
  • Middle Aged
  • Risk