Associations Between Left Ventricular Dysfunction and Brain Structure and Function: Findings From the SABRE (Southall and Brent Revisited) Study.

Background-Subclinical left ventricular (LV) dysfunction has been inconsistently associated with early cognitive impairment, and mechanistic pathways have been poorly considered. We investigated the cross-sectional relationship between LV dysfunction and structural/functional measures of the brain and explored the role of potential mechanisms. Method and Results-A total of 1338 individuals (69+6 years) from the Southall and Brent Revisited study underwent echocardiography for systolic (tissue Doppler imaging peak systolic wave) and diastolic (left atrial diameter) assessment. Cognitive function was assessed and total and hippocampal brain volumes were measured by magnetic resonance imaging. Global LV function was assessed by circulating N-terminal pro-brain natriuretic peptide. The role of potential mechanistic pathways of arterial stiffness, atherosclerosis, microvascular disease, and inflammation were explored. After adjusting for age, sex, and ethnicity, lower systolic function was associated with lower total brain (beta +/- standard error, 14.9 +/- 3.2 cm(3); P<0.0001) and hippocampal volumes (0.05 +/- 0.02 cm(3), P=0.01). Reduced diastolic function was associated with poorer working memory (-0.21+0.07, P=0.004) and fluency scores (-0.18+0.08, P=0.02). Reduced global LV function was associated with smaller hippocampal volume (-0.10 +/- 0.03 cm(3), P=0.004) and adverse visual memory (-0.076 +/- 0.03, P=0.02) and processing speed (0.063+0.02, P=0.006) scores. Separate adjustment for concomitant cardiovascular risk factors attenuated associations with hippocampal volume and fluency only. Further adjustment for the alternative pathways of microvascular disease or arterial stiffness attenuated the relationship between global LV function and visual memory. Conclusions-In a community-based sample of older people, measures of LV function were associated with structural/functional measures of the brain. These associations were not wholly explained by concomitant risk factors or potential mechanistic pathways.