Disruptions in circadian rhythm alter amyloid‐dependent changes in tau and cognition

BACKGROUND: The circadian system drives biological and physiological processes aligning with a near 24-h light-dark cycle, referred to as circadian rhythms (CR). Disruptions in CR are an emerging feature of aging and Alzheimer disease (AD). Alterations in CR - including increased fragmentation of rest-activity rhythm (RAR) and decreased amplitude of RAR - are hypothesized to be contributors to AD pathogenesis and potentially modifiable risk factors for AD. Here we examined whether RAR features might be associated with amyloid-dependent cognitive and tau changes in a cohort of cognitively normal (CN) and cognitively impaired (CI) older adults. METHODS: CR analysis was performed on actigraphy data using nonparametric analyses and a data adaptive approach (empirical mode decomposition) to measure RAR fragmentation (intradaily variability [IV]), relative amplitude (RA), and mean amplitude (MA) in 171 individuals (n =135 CN and n=36 CI; Table 1). Individuals completed cross-sectional and longitudinal (in a subset of n=76) cognitive assessments (PACC and MMSE), tau-positron emission tomography (PET), and amyloid-PET. RESULTS: Cross-sectionally, increasing age was associated with lower RA (r=-0.23, p =0.0029; Figure 1A) and MA (r=-0.36, p =1.2e-06; Figure 1B), and higher IV (r=0.29, p =0.0001; Figure 1C), and male sex was associated with higher IV (p =0.034; Figure 1D). Using linear regressions, adjusting for age and sex, higher RA attenuated the effect of amyloid on and PACC scores (RA*amyloid: B[SE]=24.55[12.1], p =0.044; Figure 2A), MMSE scores (RA*amyloid: B[SE]=76.12[24.4], p =2.15e-03; Figure 2B) and inferior temporal tau-PET levels (RA*amyloid: B[SE]=-13.36[2.7], p =2.43e-06; Figure 2C). Using linear mixed effect models, adjusting for baseline age and sex, higher RA modified the effect of baseline amyloid on longitudinal PACC decline (time*RA*amyloid: B[SE]=3.16[1.1], p =0.007; Figure 2D), MMSE decline (time*RA*amyloid: B[SE]= 9.56[3.2], p =0.036; Figure 2E), and tau accumulation (time*RA*amyloid: B[SE]=-0.56[0.2], p =0.016; Figure 2D). CONCLUSION: In a sample composed of CN and CI individuals, disruptions in CR measures were associated with older age and male sex. Higher relative amplitude of a RAR was associated with diminished cross-sectional and longitudinal tau accumulation and cognitive decline in people with elevated amyloid. These findings suggest robust CR may provide resilience to amyloid-related cognitive decline and tau accumulation and supports circadian dysfunction as a potentially modifiable risk factor for AD progression.