Abstract
Atrial fibrillation has been associated with cognitive impairment. Whether subclinical abnormalities in atrial function and substrate predisposing to atrial fibrillation impact cognitive function has received limited attention. We tested associations of echocardiographic markers of atrial structure and function with cognitive function and decrease in 510 participants with obesity and metabolic syndrome (mean age SD of 64.4 [5.2] years in men and 66.5 [3.9] years in women). Left atrial (LA) markers (volume index, emptying fraction, strain, function index, and stiffness index) were estimated based on transthoracic echocardiography. General cognitive functioning (Mini-mental state examination), verbal ability (verbal fluency test), memory and attention (Digit Span Tests), and processing speed and executive function (Trail-Making Tests A and B) were assessed at baseline and at 2-year follow-up. Multiple linear regression was used to test associations of atrial markers (modeled in SD units) with baseline and 2-year changes in cognitive scores adjusted for demographic and health covariates. LA structure and function were not associated with cognitive function at baseline. Larger LA volume index (standardized β [95% confidence interval] -0.13 [-0.22 to -0.03]), lower peak longitudinal strain (-0.11 [-0.20 to -0.01]), and higher stiffness index (-0.18 [-0.28 to -0.08]) were associated with 2-year worsening in Trail-Making Test A. Strain measurements were also associated with a 2-year change in the Controlled Oral Word Association Test. In conclusion, overall, adverse markers of LA structure and function were associated with 2-year detrimental executive function-related cognitive changes in a sample of participants at high risk for cardiovascular disease, highlighting LA substrate as a potential risk factor for cognitive decrease and dementia.