Background: Patients suffering chronic heart failure (HF) are at increased risk for accelerated cognitive decline, including development of vascular and Alzheimer’s dementia. These complications are highly relevant as they significantly increase morbidity and health care costs. However, comprehensive cognitive testing, which is needed to identify patients at risk, remains expensive and thus unavailable to many patients. As atherosclerotic changes of cerebral vessels are thought to fuel cognitive impairment, we questioned the role of non-invasive measurement of pulse wave velocity (PWV), which represents a measurement of arterial stiffness and is an independent predictor of cardiovascular risk. We hypothesized that PVW relates to 1) severity of chronic HF, 2) brain morphologic measures and 3) cognitive decline in chronic HF.

Methods: We investigated participants of the prospective Cognition.Matters-HF cohort study, which included patients with chronic, stable systolic or diastolic heart failure, but clinically unapparent with regard to history or signs/symptoms of cerebrovascular accidents. Besides carotid-femoral PWV measurement, extensive neurological, cardiological and neuropsychological testing (five domains of cognition: intensity of attention, selectivity of attention, visual/verbal memory, working memory and visual/verbal fluency) and brain magnetic resonance imaging was done. Cerebral atrophy was visually rated from 1 to 8, medial temporal lobe atrophy by Schelten’s score ranging from 0 (normal) to 4 (severe atrophy).

Results: We measured PWV in 106 study participants. Their median age was 65 years (quartiles 56, 73) and ranged from 35 to 85 years. The majority of patients was male (87.7%). Median left ventricular ejection fraction (LVEF) was 44 % (37, 50), levels of NT-pro BNP were 471 pg/ml (197, 1114) and 6-minute walking distance was 400 m (360, 460).

1) PWV (9.3 m/s (7.9, 10.8)) was highly age-dependent (Spearman’s ρ=0.93; p<0.001), but also related to surrogates of HF severity, such as NT-proBNP (ρ=0.36; p<0.001) and 6-MWD (ρ=-0.36; p<0.001), but not with LVEF (ρ=0.011; p=0.911). Using a multivariable regression model with backward elimination including the above-mentioned parameters, both age (T=24.2; p<0.001) and NT-proBNP (T=-2.1; p=0.038) remained significant, independent correlates of PVW with very high explainational variance (R²=0.86).

2) Cerebral MRI revealed cerebral atrophy score of 3.0 (2.5, 4.0), hippocampal atrophy of 2.0 (1.5, 3.0) and white matter hyperintensity volume (WMH) of 3.5 cm³ (1.9, 5.8). PVW strongly related to cerebral atrophy (ρ=0.55; p=0.001), hippocampal atrophy (ρ=0.32; p=0.002) and WMH volume (ρ=0.53; p<0.001), but not in age-adjusted models.

3) Cognitive testing revealed intensity attention deficits (age-adjusted T-score <40) in 50 %, memory deficits in 19%, and executive dysfunction in 21 % of examined patients. PWV correlated with T-scores of executive functions (ρ=-0.27; p=0.005), but not with intensity of attention (ρ=-0.15; p=0.132) or memory (ρ=-0.13; p=0.195).

Discussion: We found that in chronic HF, PWV independently relates to levels of NT-proBNP and ageing. Further, high PWV, indicating increased vascular stiffness, associated with worse performance in executive functioning. Despite limitations due to small sample size and cross-sectional design, our findings implicate the need for further investigations on the role of vascular stiffness in cognitively impaired HF patients.