A University of Kentucky study found that microglia, the brain’s resident immune cells, drive an early loss of non-rapid eye movement (NREM) sleep in mice with Alzheimer’s-like amyloid pathology. Temporarily removing most of the cells restored more than two hours of sleep per day without changing the animals’ amyloid plaque burden.
The researchers compared genetically modified APP/PS1 mice with normally aging mice at six and 18 months. Amyloid plaques reduced NREM sleep, while normal aging primarily reduced rapid eye movement (REM) sleep. Sleep loss linked to plaques appeared when the deposits first emerged and did not worsen as plaque levels increased substantially.
Brain recordings showed that the Alzheimer’s-model mice had cortical hyperexcitability, weaker coordination between brain regions and unstable sleep-wake patterns. Whole-brain imaging also found expanded microglial populations in regions involved in sleep regulation, including thalamic areas and white-matter tracts, many of which did not contain plaques.
To test whether the immune cells were causing the sleep disruption, the team fed the mice pexidartinib, also known as PLX3397, for 14 days. The drug blocks a survival signal used by microglia and reduced their numbers by about 87 percent.
After the drug was stopped and microglia were allowed to repopulate for 10 days, the increase in NREM sleep persisted, although the effect was smaller and the REM improvement did not remain at the same level. The drug did not alter sleep in ordinary mice, suggesting that the effect was linked to disease-associated microglial activity rather than cell depletion alone.
The researchers also used EEG analysis to separate rhythmic brain-wave activity from background electrical activity. The resulting signatures distinguished amyloid-associated network dysfunction from changes caused by normal aging, raising the possibility that sleep measures and EEG could help track early disease-related changes. The experiments were conducted in mice, so the findings do not establish that removing or modifying microglia would restore sleep in people with Alzheimer’s disease.
The work, led by Shannon L. Constantino, was published in 2026 in Alzheimer’s & Dementia. The laboratory is now examining whether microglial activity can be moderated without eliminating the cells, including through investigations of existing drugs that may alter their metabolism or activation state.