Sleep as an Active Modifier of Blood-Based Biomarkers in Alzheimer’s Disease

Article information

Sleep Med Res. 2026;17(1):1-2

Publication date (electronic) : 2026 March 31

doi : https://doi.org/10.17241/smr.2026.03384

Yoo Hyun Um, MD, PhD

Department of Psychiatry, St. Vincent’s Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea

Corresponding Author Yoo Hyun Um, MD, PhD Department of Psychiatry, St. Vincent’s Hospital, College of Medicine, The Catholic University of Korea, 222 Banpo-daero, Seocho-gu, Seoul 06591, Korea Fax +82-31-248-6758 E-mail cherubic712@naver.com

Blood-based biomarkers (BBMs) for Alzheimer’s disease (AD), including phosphorylated tau at threonine 217, 181 (pTau217, pTau181), the amyloid-β 42 to amyloid-β 40 ratio (Aβ42/40), glial fibrillary acidic protein (GFAP), and neurofilament light chain (NfL), are emerging as a cost-effective platform for early screening and detection. Increase in relevant research is grounded in the assumption that plasma biomarkers primarily reflect stable, pathologydriven brain processes. However, emerging evidence suggests that diverse demographic factors and clinical comorbidities can influence BBM levels [1]. One of them is sleep, not merely a behavioral state but a powerful biological modulator of protein clearance, neuroinflammation, and brain-blood transport [2,3]. For this reason, I propose that sleep should be conceptualized as an active modifier of BBMs, as interpretations that do not account for sleep may lead to misclassification, particularly in preclinical disease where biomarker changes are subtle.

Although BBMs are often viewed as highly stable indicators, accumulating findings suggest they are sensitive to sleep-related processes. A recent study demonstrated that disrupted rest-activity rhythm in midlife women has been linked to poorer processing speed and higher plasma phosphorylated tau at threonine 231 (p-Tau231) levels. Moreover, poorer sleep quality was associated with higher plasma p-Tau231 levels and interacted with AD genetic risk to influence p-Tau231 concentrations [4]. Sleep disorders such as obstructive sleep apnea (OSA) also affected the BBM levels, with blood total Aβ, Aβ42, Aβ40, and total tau all elevated in OSA patients [5]. Furthermore, p-Tau217 was reported to have a diurnal variation, with the lowest level observed in the morning and the highest values observed in the afternoon/early evening [6]. Interestingly, these dynamic changes in BBMs were influenced by other known risk factors for AD, such as sex and apolipoprotein E genotype [7].

In Aβ+ cognitively unimpaired individuals, sleep plays a critical role in AD at the stage when biomarkers are most informative for early detection. In this preclinical phase—where overt cognitive symptoms are absent—biomarkers play an integral role in early detection and risk stratification, yet their absolute levels and longitudinal changes are usually subtle and vulnerable to fluctuation. Accumulating evidence indicates that older adults at genetic or biomarker-defined risk of AD frequently exhibit early sleep disturbances, including reduced slow-wave activity, impaired sleep spindle density, circadian misalignment, and increased nocturnal fragmentation—alterations that have been linked to Aβ deposition and tau propagation [8,9].

A central paradox emerges at this stage: individuals in the earliest phases of AD pathology, in whom biomarkers are most clinically valuable, may also experience the greatest sleep-related variability in biomarker levels. Experimental and human studies demonstrate that sleep-wake dynamics directly influence Aβ and tau concentrations through activity-dependent release and clearance mechanisms [10]. As a result, a BBM drawn after a night of poor sleep may be interpreted as abnormal or “borderline positive,” whereas the same individual could fall below diagnostic thresholds following consolidated sleep. This variability has important implications not only for clinical diagnosis but also for longitudinal tracking, risk stratification, and treatment-response monitoring in prevention-orientated settings.

These considerations have several implications. First, BBM collection should be standardized: morning fasting samples after adequate sleep may reduce physiological noise. Many large biomarker trials do not document sleep the night before sampling, despite the likelihood of substantial variation. Second, incorporating simple tools such as the Pittsburgh Sleep Quality Index, Epworth Sleepiness Scale, or actigraphy-derived sleep efficiency could contextualize biomarker levels at minimal cost. Third, borderline biomarker results should be interpreted cautiously, especially in individuals with recent sleep deprivation, insomnia, or circadian disruption. Finally, interventions that enhance slow-wave sleep may stabilize biomarker variability and sharpen the detection of true pathological change, an area ripe for translational research.

As BBMs move toward widespread clinical adoption, the field must recognize that these markers are shaped by fundamental aspects of sleep biology. Glymphatic clearance, circadian timing, neuroinflammation, and barrier integrity all modulate the production and peripheral appearance of tau, amyloid, GFAP, and NfL. Treating BBMs as static indicators risks overinterpretation or misinterpretation, particularly in earlier disease stages when sleep impairment is subtle yet meaningful. Integrating sleep metrics into biomarker protocols is a simple and scalable strategy that can enhance diagnostic precision and accelerate translational progress. Rather than viewing sleep as a confounder, the field should begin to treat it as an essential component of biomarker biology and a window into early neurodegenerative change.