Every night, healthy sleepers briefly rouse dozens of times without ever knowing it — split-second shifts in brain activity called micro-awakenings that don't reach conscious awareness and leave no memory behind. A new study from the University of Liège's GIGA Neurosciences institute, published in the journal Sleep, finds that how often this happens tracks with a person's genetic risk for Alzheimer's disease, but only starting in middle age.
Tracking Sleep in More Than 500 Healthy Adults
The research team, led by senior author Gilles Vandewalle, recorded brain activity overnight in more than 500 cognitively healthy participants split across two age groups: young adults in their 20s and older adults in their 50s and 60s. Using each participant's polygenic risk score — a composite measure of how many Alzheimer's-associated genetic variants a person carries — the researchers compared genetic risk against the frequency of micro-awakenings detected in each person's sleep recordings.
Among the middle-aged group, participants with higher polygenic risk scores had significantly more frequent micro-awakenings during the night. Among the young adults, there was no measurable association at all. The genetic risk was present in both groups from birth; the sleep signature associated with it only became detectable decades later.
Why the Brainstem Is the Prime Suspect
Vandewalle's team points to the locus coeruleus, a structure in the brainstem barely the size of a grain of rice, as the likely mechanism. The locus coeruleus is one of the brain's principal regulators of arousal, attention, and the transitions between sleep stages, and it is also known to be among the earliest structures affected by the tau protein pathology that characterizes Alzheimer's disease, sometimes years before amyloid plaques become detectable. The researchers hypothesize that subtle, early degeneration in this structure destabilizes sleep continuity long before any cognitive symptoms appear, producing the extra micro-awakenings observed in genetically at-risk, middle-aged participants.
That timeline is what makes the age-dependent finding notable rather than incidental. If micro-awakening frequency simply reflected fixed genetic risk, the association should have shown up equally in both age groups. Instead, it emerged only once participants reached an age when early neurodegenerative changes plausibly begin, suggesting the sleep disruption tracks the biological process itself rather than the genetic risk in the abstract.
An Association, Not a Diagnosis
The authors are explicit that this is a correlational finding, not proof that disrupted sleep causes Alzheimer's or that Alzheimer's causes the sleep disruption first. Polygenic risk scores also don't predict who will develop the disease with certainty; they estimate relative risk across a population, not an individual diagnosis. Larger, longitudinal studies that track the same individuals over years — following whether elevated micro-awakening frequency actually predicts later cognitive decline — would be needed before this could function as a validated screening tool.
What This Means for Patients
For now, there's no test a patient can request and no action to take based on this study alone. Its significance is what it points toward: a potential low-cost, non-invasive marker — ordinary sleep recording — that could flag elevated Alzheimer's risk long before memory symptoms appear, at a stage when interventions like lifestyle changes, cardiovascular risk management, or future disease-modifying therapies might have more room to work. Vandewalle has described the goal as making sleep "an accessible marker for the early identification of vulnerable individuals." Turning that into something clinically usable will require years of further validation, but it adds sleep architecture to a growing list of early biological signals researchers are racing to catch before Alzheimer's disease becomes symptomatic.
