Internal carotid artery blood flow and pulsatility index in cognitively impaired people with chronic heart failure

Aims: Mild cognitive impairment and dementia are common and serious co-morbidities in people with chronic heart failure (HF) as they increase hospitalization rates, mortality and health care costs. Upon other factors, dysregulated cerebral perfusion might contribute to brain pathology. We aimed to evaluate the association of non-invasively measured blood flow (BF) and pulsatility index (PI) of the internal carotid artery (ICA) with (i) chronic HF parameters, (ii) brain morphologic measures and (iii) cognitive impairment.

Methods and results: This post-hoc analysis of the observational, prospective Cognition.Matters-HF study included 107 chronic HF patients without atrial fibrillation or carotid artery stenosis (aged 63 ± 10 years; 19% women). Using extracranial sonography, we measured ICA-BF and ICA-PI 1.5 cm distal of the carotid bifurcation. Brain magnetic resonance imaging was performed on a 3-Tesla scanner to quantify cerebral atrophy, hippocampal atrophy and white matter hyperintensities. Extensive neuropsychological testing tested the cognitive domains intensity of attention, visual/verbal memory and executive function (including its subdomains selectivity of attention, visual/verbal fluency and working memory) using a comprehensive test battery. (i) Neither ICA-BF (median 630 (quartiles 570, 700) mL/min) nor ICA-PI (1.05 (0.96. 1.23)) related to left ventricular ejection fraction, left atrial volume index or NT-proBNP. (ii) Higher ICA-PI (r = 0.25; P = 0.011), but not ICA-BF (r = 0.08; P = 0.409), associated with increased volume of white matter hyperintensities beyond ageing, while neither ICA-PI nor ICA-BF related to cerebral or hippocampal atrophy indices. (iii) ICA-BF, but not ICA-PI, positively correlated with age-adjusted T-scores of executive function (r = 0.38; P < 0.001) and its subdomains working memory (r = 0.32; P < 0.001) and visual/verbal fluency (r = 0.32; P < 0.001). In a multivariate linear model of executive function, only ICA-BF (T = 3.79; P < 0.001), but not HF or magnetic resonance imaging parameters, remained a significant correlate of executive function.

Conclusions: ICA-BF and ICA-PI, measured in broadly available extracranial sonography, independently related to measures of functional and structural brain changes in people with chronic HF, respectively. Due to limitations of this cross-sectional approach without a healthy control group, larger controlled longitudinal studies are needed to further elucidate the role of ICA-BF dysregulation and its implication for clinical care in this vulnerable cohort.