Neuroinflammation, cerebrovascular dysfunction and diurnal cortisol biomarkers in a memory clinic cohort: Findings from the Co-STAR study

Abstract

Cortisol dysregulation, neuroinflammation, and cerebrovascular dysfunction are biological processes that have been separately shown to be affected in Alzheimer's disease (AD). Here, we aimed to identify biomarker signatures reflecting these pathways in 108 memory clinic patients with subjective cognitive decline (SCD, N = 40), mild cognitive impairment (MCI, N = 39), and AD (N = 29). Participants were from the well-characterized Cortisol and Stress in Alzheimer's Disease (Co-STAR) cohort, recruited at Karolinska University Hospital. Salivary diurnal cortisol measures and 41 CSF proteins were analyzed. Principal component analysis was applied to identify combined biosignatures related to AD pathology, synaptic loss, and neuropsychological assessments, in linear regressions adjusted for confounders, such as age, sex, education and diagnosis. We found increased CSF levels of C-reactive protein (CRP), interferon γ-inducible protein (IP-10), thymus and activation-regulated chemokine (TARC), intercellular adhesion molecule-1 (ICAM-1), and vascular cell adhesion molecule-1 (VCAM-1) in MCI patients. Further, markers of cortisol dysregulation (flattened salivary cortisol awakening response and flattened cortisol slope) correlated with increased levels of placental growth factor (PlGF), IP-10, and chitinase 3-like 1 (YKL-40) in the total cohort. A biosignature composed of cortisol awakening response, cortisol slope, and CSF IL-6 was downregulated in AD patients. Moreover, biomarker signatures reflecting overlapping pathophysiological processes of neuroinflammation and vascular injury were associated with AD pathology, synaptic loss, and worsened processing speed. Our findings suggest an early dysregulation of immune and cerebrovascular processes during the MCI stage and provide insights into the interrelationship of chronic stress and neuroinflammation in AD.

Fig. 1. Correlation heatmap of the analysed CSF and diurnal cortisol biomarkers.

CSF Aβ42, t-tau, p-tau, and all biomarkers from Table 2 are included in the analysis. Correlation coefficients (Spearman’s test) are indicated in red/blue depending on the direction of the association (red: positive association, blue: negative association). ns p ≥ 0.05, *p < 0.05, **p < 0.01, ***p < 0.001.

Fig. 2. Box-plots depicting component scores among SCD, MCI, and AD patients.

A–F Principal components 1-6 (PC1-PC6). Component scores are in y-axis. The median is shown as a horizontal line, and the whiskers represent the 10th and 90th percentiles. P values were calculated by analysis of covariance (ANCOVA), adjusting for age and sex. Only p-values for significant differences are presented (P < 0.05 was considered statistically significant). AD Alzheimer’s disease, MCI mild cognitive impairment, PC1 principal component 1, SCD subjective cognitive decline.

Fig. 3. Associations of components with CSF biomarkers of AD pathology and synaptic damage.

Scatter plots representing associations of principal component 1 (PC1) with A t-tau, B p-tau, C SNAP-25, D NG, E SYT-1 and associations of F principal component 2 (PC2) with Aß42 and G principal component 3 (PC3) with Aß42. Scatter plots representing associations of PC1 with t-tau, p-tau, SNAP-25, NG, and SYT-1 and associations of PC2 and PC3 with Aβ42. The standardized regression coefficients (β) and p values shown were computed using linear regression models adjusting for age, sex, and diagnosis.