Your Sleep May Hold an Early Clue to Alzheimer’s Risk

Sleep quality in midlife may carry an early hint about Alzheimer’s risk — specifically, how often the brain briefly stirs during the night.

Researchers at the University of Liège studied sleep in over 500 healthy participants, mostly aged 18–31 plus an older group aged 50–69, comparing sleep patterns against polygenic risk scores for Alzheimer’s. Frequent micro-awakenings were associated with higher genetic risk in the middle-aged group but not the younger one. The work appeared in the journal Sleep.

What a micro-awakening is

These are brief bursts of brain activity that interrupt sleep without producing full wakefulness. They register on an EEG as a shift toward faster activity lasting seconds, and the sleeper typically has no memory of them — someone can have dozens per night and report sleeping straight through.

They are normal in small numbers. Frequent micro-arousals fragment sleep architecture, reducing time in the deeper stages where much of sleep’s restorative work occurs, even when total sleep duration looks entirely adequate.

The brainstem link

The researchers point to the locus coeruleus, a rice-grain-sized brainstem region regulating wakefulness and attention. Sleep-quality measures correlated with this region’s condition and its role in REM sleep, which supports memory.

The choice of structure is not arbitrary, and it is what makes this more than a correlation. The locus coeruleus is the brain’s principal source of noradrenaline, and it is among the very earliest sites where Alzheimer’s-related tau pathology appears — often decades before symptoms, and before the cortical regions typically associated with the disease.

A structure that both controls arousal and is damaged early offers a coherent explanation for why sleep changes might precede everything else: the disease could be disrupting sleep because it has reached the region governing it.

The proposed mechanism, and its direction problem

The researchers suggest poor locus coeruleus function could promote the buildup of Alzheimer’s-related proteins. That is plausible, because sleep is when the brain clears metabolic waste including the proteins implicated in Alzheimer’s, and fragmented sleep would impair that clearance.

But the causal arrow is genuinely ambiguous, and this is the study’s central limitation. Early pathology in the locus coeruleus could disrupt sleep, making poor sleep an early symptom. Disrupted sleep could impair clearance and accelerate pathology, making it a contributing cause. Or both could operate as a feedback loop.

The distinction determines everything practical. If sleep disruption is a symptom, it is a potential early marker. If it is a cause, it is a target for intervention. Cross-sectional data cannot separate them.

What a polygenic risk score does and does not tell you

This matters for interpreting the finding, and the researchers are appropriately cautious.

A polygenic risk score aggregates the small contributions of many genetic variants into a single estimate of relative risk. It is useful for comparing groups statistically and close to useless for predicting individuals. Most people with high scores never develop the disease; many who do had unremarkable scores.

Critically, the study did not measure Alzheimer’s, or brain pathology, or cognitive decline. It measured a statistical association between a sleep pattern and a genetic risk estimate in healthy people, none of whom have the disease.

Why only the older group

The association appeared in participants aged 50–69 and not in those aged 18–31, and there are two competing readings.

The interesting one is that the effect is real and emerges with age — early pathological processes beginning to affect sleep in midlife, decades before symptoms, which fits what is known about Alzheimer’s long preclinical phase.

The deflationary one is that the finding is a chance result. When a study tests an association across multiple groups, finding it in one is exactly what noise produces, and a genuine effect should ideally be demonstrated in an independent sample rather than in one subgroup of one study.

Why the topic still matters

Sleep is modifiable, which is what makes this line of research appealing. Almost every established Alzheimer’s risk factor is either fixed — age, genetics — or difficult to change substantially. Sleep can be measured, and in some cases improved, including through treating conditions such as sleep apnea that fragment it severely.

If disrupted sleep contributes causally to pathology even modestly, it would be one of the few available levers.

How to read it

As a hypothesis worth pursuing, not as guidance. This shows statistical associations, not proof; the genetic risk measured was low and cannot predict whether any individual will develop Alzheimer’s; and the association appeared in only one age group.

Why sleep keeps appearing in dementia research

This study is one entry in a rapidly growing literature, and the reason the field keeps returning to sleep is a specific piece of biology.

The brain has no conventional lymphatic drainage, and evidence accumulated over the past decade indicates that clearance of metabolic waste from brain tissue is substantially more active during sleep than wakefulness — with the spaces around blood vessels widening to allow fluid exchange that flushes out accumulated products, including the proteins implicated in Alzheimer’s.

If clearance depends on sleep, then chronically disrupted sleep would mean chronically incomplete clearance, and over decades that could plausibly influence how much protein accumulates. It is a mechanistically satisfying story, which is precisely why it warrants scepticism: satisfying stories attract confirmatory interpretation, and much of the supporting work is in animals or in short-term human experiments rather than in the decades-long timescale the hypothesis actually concerns.

Nobody should conclude that their sleep is telling them something about their dementia risk, and the practical advice — treat sleep disorders, prioritise sleep — is well justified on grounds that have nothing to do with Alzheimer’s. This summarises early research and is not medical advice.