Lower in‐vivo locus coeruleus integrity is associated with lower cortical thickness in older individuals with elevated Alzheimer’s pathology

Abstract

AbstractBackgroundCortical neurodegeneration is, along with Alzheimer’s Disease (AD) pathology, an important feature of AD, which correlates highly with tauopathy and cognitive decline. Post‐mortem data show that the locus coeruleus (LC) loses approximately 8.4% volume per cortical Braak stage. This work investigates whether LC integrity, as indexed by LC MRI signal intensity, is associated with cortical thickness, and whether this is exacerbated by the presence of AD pathology.MethodCross‐sectional data from 165 Harvard Aging Brain Study participants (age range=50‐94 years; female=∼60%; cognitively unimpaired: n=149, cognitively impaired: n=16; Table 1) who underwent 3T‐MRI, PiB‐PET (cut‐off=1.324 DVR) and T807‐PET scanning was selected. LC MRI signal intensity was assessed by normalizing the LC to the pontine tegmentum (reference region) for each individual and averaging clusters of 5 voxels with the highest intensities. Unilateral cortical thickness (CT) values were extracted utilizing the Desikan‐Killiany atlas (FreeSurfer 6) and averaged across hemispheres to obtain regional bilateral values. PET measures were corrected for partial‐volume effects. Associations between CT and i) LC intensity, or ii) LC intensity interacted with pathology (i.e., inferior temporal (IT) tau, entorhinal (EC) tau and global amyloid), were examined with bootstrapped linear regressions, while controlling for age, sex, and education, and applying FDR‐correction.ResultWe found a positive association between LC intensity and CT for medial and lateral temporal regions (p‐FDR<0.05; Figure 1). Models interacting LC intensity with IT or EC tau revealed that lower LC intensity was associated with lower CT especially in temporal regions. This effect was greater in individuals with elevated tau (p‐FDR<0.05; Figure 2‐3). Results still held after controlling for amyloid (p‐FDR<0.05). Similarly, interacting LC intensity with amyloid showed a relation between lower LC intensity and lower CT in temporal and some occipital regions, which was stronger in individuals with elevated amyloid (p‐FDR<0.05; Figure 4).ConclusionThis work demonstrates that lower LC integrity is associated with bilateral cortical thinning in regions aligning with the expected topographical distribution of tau in preclinical AD. Additionally, these spatial patterns are more widespread in individuals with elevated AD pathology. These results suggest that LC integrity is closely related to the downstream neurodegenerative sequela of AD pathology.