Summary: Utilizing sophisticated, non-linear mathematical modeling to evaluate a cohort of over 2,400 older adults, the investigators unmasked a distinct, curving biological relationship: regularly sleeping 8.5 hours or more per night is strongly tied to elevated blood levels of phosphorylated tau 181 (p-tau181), a hallmark diagnostic protein of Alzheimer’s pathology.
The research suggests that excessive sleep duration does not necessarily cause decay, but rather serves as a crucial behavioral footprint of early, silent neurodegeneration.
Key Facts
- The 8.5-Hour Inflection Point: By moving past standard flat averages, the non-linear RCS models revealed that plasma p-tau181 levels begin accumulating noticeably once self-reported regular sleep durations reach 8.5 to 9 hours per night. Beyond the 10-hour threshold, the risk curve spikes dramatically upward, signaling intense target biomarker concentration.
- The Specificity of p-tau181: While the team simultaneously tracked up to four blood-based proteins linked to cellular damage and neurodegeneration, the associations for the other three markers evaporated the moment the math adjusted for baseline kidney function. Only p-tau181 withstood the adjustment, proving its unique status as a specific, highly isolated mirror of Alzheimer’s-related pathways.
- Rigorous Covariate Shielding: The non-linear association held firm even after the mathematical models were heavily adjusted to rule out an array of confounding health factors, including biological age, sex variations, clinical sleep apnea, deep depression profiles, systemic kidney function metrics, and the presence of the high-risk APOE varepsilon4 genotype.
- A Behavioral Footprint, Not a Cause: Dr. Vanessa M. Young explicitly clarifies that because this study represents a snapshot in time rather than a multi-decade controlled trial, long sleep cannot be labeled as a direct cause of Alzheimer’s disease. Instead, excessive sleep is highly likely to be a downstream behavioral symptom, an early physiological reaction as the brain attempts to cope with initial, silent neurodegenerative changes.
- Building on Prior Cognitive Telemetry: This biomarker analysis extends a 2025 investigative milestone by the same team, which proved that sleeping 9 hours or more per night was significantly associated with a drop in processing speed and general cognitive performance, particularly among individuals managing co-occurring depression.
- The Clinical Conversation Starter: The public health takeaway translates into a simple, practical diagnostic boundary: if an older adult routinely requires 9 to 10 hours of sleep or more just to feel rested, it should serve as an immediate behavioral cue to initiate an open conversation with their physician regarding sleep quality, objective biomarker screening, and proactive brain health monitoring.
Source: UT Health San Antonio
Regularly sleeping long hours each night is associated with higher levels of an Alzheimer’s-related protein in the blood, even accounting for other health factors, suggests a new study from UT Health San Antonio, the academic health center of The University of Texas at San Antonio.
Modeling across a sample of 2,410 study participants found a link between sleep duration and phosphorylated tau 181, or p-tau181, a modified form of tau protein that is a hallmark of Alzheimer’s disease, and is now detectable in blood.
Regular sleep durations beginning at eight-and-a-half to nine hours per night were associated with higher p-tau181 levels, increasing most sharply beyond 10 hours, suggesting that long sleep might reflect early neurodegenerative processes.
“A lot of people worry about whether their sleep habits are affecting their brain health,” said Vanessa M. Young, PhD, MS, a postdoctoral research fellow at the Glenn Biggs Institute for Alzheimer’s and Neurodegenerative Diseases at UT Health San Antonio.
“Because this is a snapshot in time rather than a long-term study, we cannot say that long sleep causes Alzheimer’s, but the findings suggest it may be worth monitoring, and that more sleep is not always better for brain health.”
Young, also a trainee at the Sam and Ann Barshop Institute for Longevity and Aging Studies at UT Health San Antonio under a National Institutes of Health Institutional National Research Service Award (T32) grant, is first and corresponding author of the study, titled, “Non-linear associations between sleep duration and plasma p-tau181 in the Framingham Heart Study,” published May 19 in Alzheimer’s & Dementia, the Journal of the Alzheimer’s Association.
She is a graduate of the Translational Science PhD Program in the Graduate School of Biomedical Sciences at UT Health San Antonio, and at the Biggs Institute is under the mentorship of founding director Sudha Seshadri, MD.
The study’s participants from the Framingham Heart Study – an ongoing community-based cohort study of residents in Framingham, Massachusetts, under the direction of the National Heart, Lung, and Blood Institute of the National Institutes of Health – had an average age of 70, plus/minus 8.45 years, with 55.2% female.
The effort follows a similar one in 2025 in which sleeping nine hours or more per night was associated with worse cognitive performance, particularly for those with depression. The latest study, with several of the same researchers, went further and accounted for multiple health factors and examined blood-based biomarkers implicated with Alzheimer’s disease and neurodegeneration in relation to participants’ self-reported sleep duration.
The long and the short of it
Both short and long sleep duration have been associated with Alzheimer’s disease risk, yet the nature of the “non-linear” associations between sleep quantity and blood-based biomarkers of the disease and neurodegeneration has been understudied. Non-linear, in this case, means the relationship between sleep and the biomarker is not a straight line, but rather it changes in strength depending on how much someone sleeps.
The study notes that dementia affects 57 million individuals globally, with Alzheimer’s disease accounting for 60% to 70% of cases. Even with recent advances in disease-modifying therapies, Alzheimer’s remains a profound medical and societal challenge.
Thus, identifying modifiable risk factors that could enable diagnosis, prevention or delay of disease onset is a critical research priority. Sleep is a promising modifiable risk factor linked to the disease, but existing evidence has remained limited and inconclusive.
In clinical studies, Alzheimer’s biomarkers are measured as surrogates of the disease’s pathology in the brain, and the recent expansion in easily obtained plasma biomarkers has opened the way for more thorough examination of the relationship between sleep and Alzheimer’s.
Researchers at the Biggs Institute recognized that leveraging the large, well-characterized Framingham Heart Study cohorts offered an opportunity to address gaps by simultaneously analyzing up to four blood-based biomarkers using flexible modeling approaches and systematic covariate assessment.
They used non-linear modeling, which – unlike linear models that assume a constant, straight-line rate of change – uses curves, exponentials and logarithms to map complex relationships in data. On a scatter-plot graph of dots representing values on horizontal and vertical axes, for instance, non-linear points form a curved pattern, meaning the relationship changes in strength or direction across the range of data.
Specifically, the scientists used restricted cubic splines, or RCS, a common technique used in regression analysis to model non-linear relationships between a continuous predictor (a factor influencing an outcome) and a target variable (the outcome).
The RCS revealed a robust non-linear association between sleep duration and p-tau181, with higher levels at eight-and-a-half hours or more, after adjusting for factors including age, sex, sleep apnea, depression, kidney function and apolipoprotein E ε4 genotype, a common genetic variant linked to an increased risk of developing late-onset Alzheimer’s and cardiovascular disease.
Young, who previously led a systematic review on sleep duration and Alzheimer’s fluid biomarkers, said the non-linear approach was essential to uncovering the finding.
“When you force a straight line through data that curves, you miss the story entirely,” she said. “We knew from the existing literature that the relationship between sleep and these biomarkers was unlikely to be simple, and that’s exactly what we found.”
The results weren’t the same for all biomarkers tested. Researchers also examined three other proteins in the blood linked to brain cell damage and neurodegeneration. For those, the sleep association disappeared once kidney function was accounted for. Only p-tau181 held up, suggesting the link between long sleep and this particular protein might be specific to Alzheimer’s-related processes, but more research is needed.
The researchers concluded that the findings “identify long sleep as a potential behavioral marker of elevated p-tau181, with implications for risk assessment that warrant prospective validation.”
“In plain terms, if you regularly find yourself sleeping nine to 10 hours or more a night, it may be worth mentioning to your doctor as a useful conversation starter about your sleep quality and overall brain health,” Young suggests.
Other authors of the study are with the Framingham Heart Study; Hôpital du Sacré-Coeur de Montréal, Montreal; University of Montreal; Monash University, Australia; Universitat Internacional de Catalunya (UIC), Barcelona; National Institute of Health Carlos III, Madrid; The University of Texas at Austin; and Boston University.
Non-linear associations between sleep duration and plasma p-tau181 in the Framingham Heart Study
Vanessa M. Young, Cystal Wiedner, Andrée-Ann Baril, Matthew P. Pase, Agustin Ruiz, Arash Salardini, Christopher R. Frei, Tiffany Kautz, Rebecca Bernal, Stephanie Yiallourou, Lachlan Cribb, Alexa Beiser, Antonio L. Teixeira, Jayandra Jung Himali, Sudha Seshadri.
Key Questions Answered:
A: In standard statistics, researchers often use linear models, which assume that a relationship is a simple, constant straight line, meaning if you change one factor, the outcome moves at a perfectly predictable rate in one direction. However, biology is rarely that simple. As Dr. Young noted, “When you force a straight line through data that curves, you miss the story entirely.” By using non-linear modeling (Restricted Cubic Splines), the team allowed the data to curve naturally. This revealed that while varying sleep amounts between 6 and 8 hours shows a flat relationship with the Alzheimer’s protein, a critical turning point occurs at 8.5 hours, where the risk line suddenly changes direction and curves sharply upward.
A: No, forcing yourself to set an early alarm will not lower your risk or stop the accumulation of the protein. It is essential to understand that this study does not show that long sleep causes Alzheimer’s. Rather, the current thinking is that excessive sleep is a behavioral byproduct or warning sign of a brain already experiencing early, microscopic changes. As neurodegeneration quietly begins to disrupt internal brain networks, it alters the body’s sleep-wake cycles, forcing the individual to naturally require more hours in bed. Long sleep is the indicator on the dashboard, not the engine failure itself.
A: When proteins like p-tau181 or other cellular waste products build up in the bloodstream, the body relies on the kidneys to filter them out through urine. If an older adult has sluggish or declining kidney function, those proteins will naturally back up and register as abnormally high on a blood test, completely independent of what is happening in their brain. By using rigorous mathematical controls to account for kidney efficiency, the UT Health San Antonio team discovered that three other neurodegenerative proteins vanished from the sleep data, proving their elevation was just a filtration issue. Only p-tau181 stood firm, confirming that its link to long sleep is deeply specific to Alzheimer’s pathology.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- Journal paper reviewed in full.
- Additional context added by our staff.
About this Alzheimer’s disease and sleep research news
Author: Steven Lee
Source: UT Health San Antonio
Contact: Steven Lee – UT Health San Antonio
Image: The image is credited to Neuroscience News
Original Research: Open access.
“Non-linear associations between sleep duration and plasma p-tau181 in the Framingham Heart Study” by Vanessa M. Young, Crystal Wiedner, Andrée-Ann Baril, Matthew P. Pase, Agustin Ruiz, Arash Salardini, Christopher R. Frei, Tiffany Kautz, Rebecca Bernal, Stephanie Yiallourou, Lachlan Cribb, Alexa Beiser, Antonio L. Teixeira, Jayandra Jung Himali, Sudha Seshadri. Alzheimer’s & Dementia
DOI:10.1002/alz.71499
Abstract
Non-linear associations between sleep duration and plasma p-tau181 in the Framingham Heart Study
INTRODUCTION
Both short and long sleep duration have been associated with Alzheimer’s disease (AD) risk, yet the nature of the non-linear associations between sleep quantity and blood-based biomarkers of AD and neurodegeneration remains understudied.
METHODS
Among 2410 Framingham Heart Study participants (mean age: 70.0 ± 8.45 years; 55.2% female), we examined associations between self-reported sleep duration and plasma phosphorylated tau (p-tau)181, total tau, neurofilament light chain, and glial fibrillary acidic protein using restricted cubic splines (RCS) and categorical comparisons (≤ 6, > 6–< 9 reference; ≥ 9 hours).
RESULTS
RCS revealed non-linear associations between sleep duration and p-tau181 (overall p = 0.005; non-linearity p = 0.002), with higher levels at ≥ 8.5 hours, after adjusting for age, sex, apolipoprotein E ε4 genotype, sleep apnea, depression, and kidney function. Categorical comparisons showed no association in adjusted models.
DISCUSSION
Sleep duration exhibits robust non-linear associations with p-tau181. Findings underscore the importance of non-linear modeling in relating sleep to blood-based biomarkers. Longitudinal studies are needed to clarify temporal relationships and mechanistic pathways.
