The brain in atrial cardiomyopathy: MRI evidence of subclinical injury in the UK Biobank
N L Van Vreeswijk, O T B Vad, Y Blaauw, A S Yassin, J K Kanters, C Graff, G J Biessels, M RienstraAbstract
Background
Atrial cardiomyopathy (ACMP) encompasses structural, functional, and electrical atrial abnormalities and is associated with adverse outcomes, including atrial fibrillation (AF), heart failure (HF), and stroke. Yet, evidence linking ACMP to subclinical brain injury is scarce, especially among individuals without a diagnosis of AF, HF and stroke, and large-scale population data integrating cardiac and brain magnetic resonance imaging (MRI) remain limited.
Purpose
To investigate whether markers of ACMP are associated with MRI-based indicators of brain injury in a population without a prior diagnosis of AF, HF, or stroke.
Methods
We included individuals from the UK Biobank of European ancestry and available cardiac electrocardiogram, cardiac MRI and brain MRI data. Individuals with a diagnosis of AF, HF, or stroke prior to imaging, and individuals not in sinus rhythm at imaging, were excluded. We examined four markers of ACMP: enlarged left atrial volume index (LAVImax >60 mL/m² or LAVImin >30 mL/m²), reduced left atrial ejection fraction (LAEF <45%), prolonged P-wave duration (>120 ms), and abnormal P wave terminal force (PTFV1 ≥5000 µV·ms). Associations with MRI-derived total brain volume, grey matter volume, white matter volume, and white matter hyperintensities (WMH) were assessed multivariable linear regression according to number of ACMP markers, and for each marker separately. Brain volumes were normalized for head size. WMH burden was analysed as the log-transformed ratio between WMH volume and total brain volume. Primary analyses were adjusted for age, sex, body mass index (BMI), and imaging centre. Sensitivity analyses were conducted further adjusting for, left ventricular ejection fraction (LVEF) and left ventricular dimensions.
Results
We included 26,467 individuals (52.6% women, median age 64.4 years). A total of 4,145 individuals (15.7%) exhibited at least one marker of ACMP and 619 (2.3%) had two or more markers. Compared with individuals with no markers of ACMP, having ≥2 ACMP markers was associated with smaller total brain volume (β:–7.7 mL, 95% confidence intervals [CI]: -12.7 to -2.7, p=0.003) and smaller grey matter volume (β:–7.0 mL, 95% CI: -10.0 to -4.0, p<0.001, figure 1). Among individual markers, reduced LAEF showed the strongest associations (lower total brain volume (–10.2 mL) and grey matter volume (–8.1 mL), and higher white matter hyperintensity burden (β=0.15 on the log scale; ≈16% increase; all p<0.001, figure 2), with grey matter volume and white matter hyperintensity burden remaining significant in sensitivity analysis. P-wave prolongation showed a modest association with lower grey matter volume that did not survive sensitivity analysis.
Conclusions
Markers of ACMP were associated with reduced grey matter volume and higher white matter hyperintensity burden in a population without a prior diagnosis of AF, HF or stroke.