Temporal associations of neuropsychiatric symptoms, demographics and amyloid with subsequent tau burden in older adults.

Background: Psychiatric symptoms are increasingly recognized as early manifestations of Alzheimer's disease (AD). These symptoms may reflect or contribute to underlying neurobiological changes, including tau burden, which represents more advanced AD pathology. Understanding factors associated with tau burden may help identify individuals at elevated risk and improve early detection strategies.

Objectives: To investigate the temporal relationships between neuropsychiatric symptoms, demographic factors, and tau burden, by examining amyloid (Aβ)-dependent, -independent, and interactive associations.

Design: Retrospective cohort study.

Setting: Alzheimer's Disease Neuroimaging Initiative.

Participants: We included 681 participants without dementia (mean age = 71.2 years, 51.8 % female).

Measurements: The participants underwent tau PET scanning with prior amyloid PET and Neuropsychiatric Inventory (NPI) interview assessments clustered around three time periods relative to tau PET: closest (0 - 2 years), mid (3 - 5 years), and furthest (6 - 8 years). Linear regression analyses, adjusting for age and APOE ε4 alleles, examined associations of NPI scores, sex, education, and their Aβ-status interactions with tau burden.

Results: Higher total NPI scores up to 2 years prior to tau PET were associated with greater tau burden independently of Aβ status, whereas anxiety symptoms demonstrated an Aβ-dependent relationship with tau. (NPI: β=0.117, 95 % CI: 0.049 to 0.185, p=0.001, Anxiety: β=0.249, 95 % CI: 0.073 to 0.424, p=0.006). NPI measured up to 5 years prior to tau PET interacted with Aβ on tau burden (0-2 years: β=0.272, 95 % CI: 0.136 to 0.407, p<0.001, 3-5 years: β=0.336, 95 % CI: 0.127 to 0.544, p=0.002). Sex and education showed minimal associations with tau at uncorrected statistical levels.

Conclusions: Neuropsychiatric symptoms were associated with tau burden up to two years before tau sampling, independently of Aβ, and interacted with Aβ status up to five years prior, suggesting that neuropsychiatric symptoms are related to tau in the short term and may represent manifestations of advancing AD pathology. Demographic factors showed minimal associations. These findings highlight the importance of evaluating neuropsychiatric and anxiety symptoms as potential indicators of increased tau pathology.