Specific Association of Worry With Amyloid-β But Not Tau in Cognitively Unimpaired Older Adults

Soyoung Lee,Benjamin S Zide,Stephan T Palm,William J Drew,Reisa A Sperling,Heidi I L Jacobs,Shan H Siddiqi,Nancy J Donovan
DOI: https://doi.org/10.1016/j.jagp.2024.04.016
Abstract:Objective: Anxiety disorders and subsyndromal anxiety symptoms are highly prevalent in late life. Recent studies support that anxiety may be a neuropsychiatric symptom during preclinical Alzheimer's disease (AD) and that higher anxiety is associated with more rapid cognitive decline and progression to cognitive impairment. However, the associations of specific anxiety symptoms with AD pathologies and with co-occurring subjective and objective cognitive changes have not yet been established. Methods: Baseline data from the A4 and Longitudinal Evaluation of Amyloid Risk and Neurodegeneration studies were analyzed. Older adult participants (n = 4,486) underwent assessments of anxiety (State-Trait Anxiety Inventory-6 item version [STAI]), and cerebral amyloid-beta (Aβ; 18F-florbetapir) PET and a subset underwent tau (18F-flortaucipir) PET. Linear regressions estimated associations of Aβ in a cortical composite and tau in the amygdala, entorhinal, and inferior temporal regions with STAI-Total and individual STAI item scores. Models adjusted for age, sex, education, marital status, depression, Apolipoprotein ε4 genotype, and subjective and objective cognition (Cognitive Function Index-participant; Preclinical Alzheimer Cognitive Composite). Results: Greater Aβ deposition was significantly associated with higher STAI-Worry, adjusting for all covariates, but not with other STAI items or STAI-Total scores. In mediation analyses, the association of Aβ with STAI-Worry was partially mediated by subjective cognition with a stronger direct effect. No associations were found for regional tau deposition with STAI-Total or STAI-Worry score. Conclusion: Greater worry was associated with Aβ but not tau deposition, independent of subjective and objective cognition in cognitively unimpaired (CU) older adults. These findings implicate worry as an early, specific behavioral marker and a possible therapeutic target in preclinical AD.
What problem does this paper attempt to address?