Table 1.

A summary of the main population-based longitudinal studies assessing the association between atrial fibrillation and risk of cognitive decline or dementia

Study, country	Study population	Assessment of atrial fibrillation	Assessment of cognitive outcomes	Main findings
Nishtala et al. [14], USA	Framingham Heart Study original and offspring cohort; 3- or 6-year follow-up; mean age 83 for the original and 68 for the offspring cohort; N = 2628	Self-reports, ECG, and medical registers	Cognitive function: a neuropsychological battery on major cognitive domains	• Prevalent AF is associated with faster decline in executive function (β coefficient = −0.31; 95% CI −0.37, −0.25)
Chen et al. [15••], USA	Atherosclerosis Risk in Communities Study; 20-year follow-up; mean age 56.9 at baseline; N = 12,515	ECG and ICD-9 codes (427.31 and 427.32)	Cognitive function: 3 neuropsychological tests Dementia: diagnostic algorithm and ICD-9 codes	• AF is associated with faster global cognitive decline (difference in Z score = 0.12, 95% CI 0.03–0.23) • AF is associated with incident dementia (HR = 1.31, 95% CI = 1.11–1.55)
Singh-Manoux et al. [16••], UK	Whitehall II Study; 26.6-year follow-up; age range 45–69 at baseline; N = 10,217	12-lead ECG and ICD-9/10 codes (437.3 and I48)	Cognitive function: a cognitive test battery (memory, reasoning, and verbal fluency) Dementia: ICD-10 codes	• Longer exposure to AF is associated with faster cognitive decline (p for trend = 0.01) • AF is associated with incident dementia (HR = 1.87, 95% CI 1.37–2.55) • Stroke does not explain these associations
de Bruijn et al. [17••], The Netherlands	Rotterdam Study; 21-year follow-up; age 55+ at baseline; N = 6514	ECG, physician diagnosis, and medical registers	Dementia: DSM-III-R criteria	• Incident AF is related to dementia only in people < 67 years (HR = 1.81, 95% CI 1.11–2.94) • Duration of AF is associated with dementia risk only in people < 67 years (p for trend = 0.003)
Rusanen et al. [18], Finland	CAIDE study; mean 7.8-year follow-up; age range 65–79; N = 1510	Medical registers	Dementia: DSM-IV criteria	• Prevalent AF is associated with dementia (HR = 2.61, 95% CI 1.05–6.47) and Alzheimer’s disease (HR = 2.54, 95% CI 1.04–6.16); the association is evident only in APOE non-carriers
Thacker et al. [19], USA	Community-dwelling people; mean 7-year follow-up; age 73 at baseline; N = 5150	ECG, ICD-9 codes	Cognitive function: modified MMSE (3MSE) and Digit Symbol Substitution Test	• Incident AF is associated with accelerated 5-year cognitive decline for age 70, 75, 80, and 85 years
Haring et al. [20], USA	RCTs of postmenopausal women; median 8.6-year follow-up; age 60+ at baseline; N = 7479	Self-reports or physical measure	MCI and probable dementia: DSM-IV criteria	• Prevalent AF is not related to probable dementia (HR = 1.12, 95% CI 0.59–2.14) • Prevalent AF is not associated with mild cognitive impairment (HR = 1.46, 95% CI 0.90–2.37)
Marzona et al. [21], 40 countries	Two RCTs of patients with CVD or diabetes; median follow-up 56 months; mean age 66.5 at baseline; N = 31,506	12-lead ECG	Cognitive function: MMSE Dementia: new dementia diagnosis, reported severe cognitive impairment, or MMSE ≤ 23	• Prevalent and incident AF is associated with ≥ 3 points decline in MMSE during the follow-up (HR = 1.14, 95% CI 1.03–1.26) • Prevalent and incident AF is associated with dementia (HR = 1.30, 95% CI 1.14–1.49)
Dublin et al. [22], USA	Community-dwelling people; mean 6.8-year follow-up; mean age 74.3 at baseline; N = 3045	At least two documented ICD-9 codes within 12 months	Dementia: DSM-IV criteria AD: NINCDS-ADRDA criteria	• Prevalent AF is associated with dementia (HR = 1.38, 95% CI 1.10–1.73) • Prevalent AF is associated with AD (HR = 1.50, 95% CI 1.16–1.94)
Marengoni et al. [23], Sweden	Kungsholmen Project; 6-year follow-up; age 75+ at baseline; N = 685	Physician diagnosis, medical records, drug use, and ICD-9 codes	Dementia: DSM-III-R criteria	• No association between AF and dementia (HR = 0.9, 95% CI 0.5–1.7) or AD (HR = 0.8, 95% CI 0.4–1.5)
Bunch et al. [24], USA	Health care patients; mean 5-year follow-up; mean age 60.6 at baseline; N = 37,025	ECG and ICD-9 codes	Dementia: ICD-9 codes	• Prevalent AF is associated with vascular dementia (HR = 1.73, p = 0.001), senile dementia (HR = 1.39, p = 0.005), and non-specific dementia (HR = 1.44, p < 0.001); highest risk was in younger group (< 70 years)
Peters et al. [25], UK	RCT of hypertensive patients; mean 2-year follow-up; age 80+ at baseline; N = 3336	ECG	Cognitive decline: decrease to MMSE < 24 or by > 3 point annually Dementia: DSM-IV criteria, a CT scan, and modified ischemic score	• No association between prevalent AF and dementia (HR = 1.03, 95% CI 0.62–1.72) • No association between prevalent AF and cognitive decline (HR = 1.08, 95% CI 0.80–1.46) • No association between prevalent AF and annual change of MMSE (β coefficient = −0.26; 95% CI −0.66, 0.13)
Rastas et al. [26], Finland	Community-dwelling people; 9-year follow-up; age 85+ at baseline; N = 553	12-lead ECG or 1-h Holter ECG; health records	Dementia: DSM-III-R criteria	• No association between prevalent AF and dementia
Tilvis et al. [27], Finland	Community-dwelling people; 10-year follow-up; age 75, 80, and 85 at baseline; N = 650	Clinical examinations	Cognitive decline: increase in Clinical Dementia Rating class or at least 4 point decrease in MMSE	• AF is associated with 5-year cognitive decline (RR = 2.88, 95% CI 1.26–6.06)

AF atrial fibrillation, ECG electrocardiogram, HR hazard ratio, CI confidence interval, RR relative risk, OR odds ratio, ICD-9/10 International Statistical Classification of Diseases and Related Health Problems, 9th Revision/10th revision, DSM-III-R Diagnostic and Statistical Manual of Mental Disorders, 3rd Edition, Revised, DSM-IV Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, NINCDS-ADRDA National Institute of Neurological and Communicative Disorders and Stroke-Alzheimer's Disease and Related Disorders Association, MCI mild cognitive impairment, CAIDE Cardiovascular Risk Factors, Aging and Dementia. MMSE Mini-Mental State Examination, RCT randomized control trial, CVD cardiovascular disease, CT computed tomography