Biomarkers of endothelial dysfunction and cognition: A two-step IPD meta-analysis

Abstract

Introduction: This study assessed the association of plasma biomarkers of endothelial dysfunction with cognitive performance and decline.

Methods: Data from 9414 individuals from eight Dutch cohorts were included (Ø age-range: 57-93 years). Plasma biomarkers of endothelial dysfunction (soluble intercellular adhesion molecule-1 (sICAM-1), soluble vascular cell adhesion molecule-1 (sVCAM-1), soluble E-selectin) were combined into a standardized composite score. Cognitive outcomes included executive function, processing speed, immediate and delayed memory, attention, and language. Linear regressions and linear mixed models were run in the individual cohorts and standardized coefficients were subsequently pooled using random-effects meta-analyses.

Results: A higher endothelial dysfunction composite score was cross-sectionally associated with worse performance on executive function, processing speed, delayed memory, and attention, but not immediate memory or language (pooled β-range: -0.04, -0.02). We found no association with change in cognition over time.

Discussion: This comprehensive two-step, individual participant data (IPD) meta-analysis showed a small, consistent cross-sectional association between endothelial dysfunction and worse cognitive performance across multiple domains but no support for a longitudinal association.

Highlights: Prior evidence on endothelial dysfunction (ED) biomarkers and cognition is conflicting. This two-step, individual participant data (IPD) meta-analysis used data from eight Dutch cohorts. ED was consistently associated with concurrent cognition. ED was not associated with a change in cognition over time. The association of ED with current cognition may be generic.

Keywords: biomarkers; cognition; cohort; endothelial dysfunction; meta‐analysis.

FIGURE 1

Cross‐sectional associations between endothelial dysfunction and executive function (A), processing speed (B), immediate memory (C), delayed memory (D), attention (E), and language (F).Note: Results are expressed as standard deviation difference in the cognitive performance measures per standard deviation higher plasma biomarkers of endothelial dysfunction composite score. The size of the square symbol is proportional to the sample size of the individual studies, with larger‐sized studies showing larger squares. Analyses are adjusted for age, sex, education, diabetes status, body mass index, cardiovascular diseases, hypertension, antihypertensive medication, dyslipidemia, lipid‐lowering medication, smoking, and alcohol use. CI, confidence interval; RE model, random‐effects model.

FIGURE 2

Longitudinal associations between endothelial dysfunction and executive function (A), processing speed (B), immediate memory (C), delayed memory (D), attention (E), and language (F).Note: Results are expressed as the interaction coefficients of the standardized endothelial dysfunction composite score with time per year in relation to the standardized cognitive performance measures. The size of the rectangles (i.e., effect estimates of the individual studies) corresponds to the sample size of the study, with larger‐sized studies having larger rectangles. Analyses are adjusted for age, sex, education, diabetes status, body mass index, cardiovascular diseases, hypertension, antihypertensive medication, dyslipidemia, lipid‐lowering medication, smoking, and alcohol use. CI, confidence interval; RE model, random‐effects model.