0758 REDUCED SPINDLE FREQUENCY AND DENSITY IN STAGE 2 NREM SLEEP IS ASSOCIATED WITH INCREASED CSF P-TAU IN COGNITIVELY NORMAL ELDERLY

• Published in: Sleep (New York, N.Y.) Vol. 40; no. suppl_1; p. A281
• Main Authors: Sharma, RA, Kam, K, Parekh, A, Uribe-Cano, S, Tweardy, S, Bubu, OM, Ayappa, I, Rapoport, DM, Varga, AW, Osorio, RS
• Format: Journal Article
• Language: English
• Published: US Oxford University Press 28.04.2017
• Subjects: NREM sleep; Older people; Pathology; Sleep
• ISSN: 0161-8105; 1550-9109; 1550-9109
• DOI: 10.1093/sleepj/zsx050.757

Abstract Introduction: Sleep may play a role in AD pathogenesis, but the timing, role, and extent to which sleep disturbances in late-life are associated with increasing burden of AD neuropathology remains unclear. Sleep spindles have been implicated in sleep quality. Wakefulness is mediated by an arousal system beginning in the brainstem and continuing on to the diencephalon and innervating the thalamus, the region where sleep spindle oscillations are generated. In AD pathology, hyperphosphorylated tau (P-Tau) protein accumulates in the brainstem, from where it spreads to the entorhinal cortices, hippocampi and other brain regions. These tau aggregates may interfere with the sleep-wake cycle resulting in down-regulation of sleep spindles and associated sleep disruption. Increased CSF P-tau and T-tau levels are likely related to the formation of neurofibrillary tangles in the brainstem and limbic system (Braak stages I-IV). Methods: 49 cognitively normal (CDR=0) elderly (66.95 ± 7.76 years) subjects completed a structural MRI, lumbar puncture (LP) and nocturnal polysomnography (NPSG) within 4.65 ± 6.81 months of the LP. From the NPSG, spindle frequency and density were analyzed for stages NREM1, NREM2 and NREM3, using an automated optimization algorithm which decomposes the EEG as a sum of transient and oscillatory components. This was used to detect the spindles and a Fourier analysis was performed to evaluate the spindle frequency in Hz. Results: Spindle frequency and density in NREM2 sleep were inversely associated with CSF P-tau (r= -0.355, p<0.05; r=-0.476, p<0.05) and CSF T-tau (r=-0.405, p<.05; r=-0.542, p<.05) using partial correlation controlling for age and ApoE4 allele. There were no associations between spindle frequency or density and CSF P-tau or CSF T-tau in stages NREM1, NREM3. Conclusion: The association of spindle frequency and density in NREM2 to CSF P-tau and CSF T-tau in cognitively normal elderly suggest either that tau pathology may produce an early downstream effect on sleep spindles, or that changes in sleep spindles can identify a process relating to tau pathology. Whether the association of tau to spindles is a non-specific effect of tau on increasing sleep fragmentation in general remains an area of active investigation. Support (If Any): R01H2118624.