Sleeping longer may signal early Alzheimer's-related brain changes

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. 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."

Vanessa M. Young, PhD, MS, postdoctoral research fellow, Glenn Biggs Institute for Alzheimer's and Neurodegenerative Diseases, UT Health San Antonio

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.