Scientists Restore Sleep in Alzheimer's Models by Targeting Brain's Immune Cells

Sleep disturbances are among the most debilitating and earliest symptoms of Alzheimer's disease, often emerging years before significant memory decline. For decades, the medical consensus assumed this insomnia was a permanent consequence of the disease's physical toll—specifically, the gradual death of neurons or the physical clutter of sticky amyloid plaques building up in the brain. Because these structural changes are incredibly difficult to reverse, treating Alzheimer's-related sleep loss has largely been limited to managing symptoms with sedatives that often leave patients groggy and prone to falls.[2]

A new study from the University of Kentucky has fundamentally flipped this understanding, revealing that the primary driver of sleep loss is not the physical damage itself, but the brain's own immune system overreacting to it. Published in the journal Alzheimer's & Dementia, the research demonstrates that the brain's resident immune cells, known as microglia, are the true culprits keeping the brain awake.[1]

Microglia act as the first line of defense in the central nervous system, constantly scavenging for damaged neurons and infectious agents. When amyloid plaques begin to form, the microglia recognize them as a threat and attempt to clear them. However, in the context of Alzheimer's, this protective mechanism goes into overdrive.

Researchers liken the phenomenon to a localized kitchen fire triggering a whole-house sprinkler system. The amyloid plaques represent the small, contained fire, while the microglia act as the sprinklers. In their aggressive attempt to address the plaques, the microglia kick off an elaborate, widespread cascade of inflammation. This inflammatory storm keeps the brain's neural circuits highly active, effectively preventing the brain from settling into deep, restorative sleep.

To test this "whole-house response" theory, the research team, led by physiologist Dr. Shannon Macauley, utilized mouse models genetically predisposed to develop amyloid plaques. They fitted the mice with miniature electroencephalography (EEG) and electromyography (EMG) monitors to track their brain waves and muscle movements, allowing the team to precisely measure periods of wakefulness, light sleep, and deep dreaming sleep.

The researchers then introduced a drug called pexidartinib, originally developed for cancer treatment, which blocks a specific survival signal that microglia rely on. By administering this drug for 14 days, the team successfully depleted approximately 87 percent of the microglia in the brains of the Alzheimer's mouse models.[1][3]

The results were dramatic. With the vast majority of the overactive immune cells removed, the mice regained more than two hours of sleep per day. Crucially, this restored time was concentrated in non-rapid eye movement (NREM) sleep—the deep, restorative phase essential for tissue repair, clearing metabolic waste, and consolidating new memories.[3]

Perhaps the most significant finding of the study was what did not change: the amyloid plaques. The mice recovered their sleep despite the fact that the physical plaque burden in their brains remained entirely unaltered. This proves that the inflammatory response is a reversible, independent cause of sleep loss that can be treated without needing to solve the monumental challenge of clearing the plaques themselves.[1]

The study also uncovered a "ceiling effect" regarding sleep disruption. The researchers observed that the sleep deficit emerged early, exactly when the plaques first appeared and triggered the initial immune storm. However, as the disease progressed and the plaque levels more than doubled, the sleep loss did not worsen.[1]

This ceiling effect positions sleep architecture—and the specific EEG signatures associated with it—as a highly sensitive, early biomarker for Alzheimer's disease. Because the sleep disruption begins at the onset of pathology, portable EEG monitoring could eventually be used in local clinics to screen patients for early-stage Alzheimer's years before cognitive decline becomes apparent, bypassing the immediate need for expensive brain scans.

However, translating this breakthrough into human treatments requires navigating a delicate biological balance. While the Kentucky study proves that hyperactive microglia drive sleep loss, completely wiping out the brain's immune system is not a viable long-term therapy for humans. Microglia perform essential housekeeping roles that keep neural networks stable.

A separate study published in July 2026 by researchers at Trinity College Dublin and Swansea University highlighted this exact danger. When that team used a different drug to indiscriminately suppress microglia in Alzheimer's models, they found that while it protected some neural connections, it failed to improve memory and actually increased abnormal, epilepsy-like electrical activity in the brain.

"The main finding, that cutting down microglia can make brain activity less stable rather than more, is a useful caution for the development of treatments that target these cells," noted Dr. James Murray of Swansea University. "It isn't as simple as fewer microglia being better."

Recognizing this complexity, the University of Kentucky team is now pivoting their focus toward modulation rather than depletion. The goal is to find ways to calm the microglia down—stopping them from "partying all night"—without removing them entirely.

The lab is currently exploring whether existing, FDA-approved medications, such as the diabetes drug metformin or the anti-seizure medication stiripentol, can be repurposed to alter how microglia metabolize energy. By changing their fuel source, researchers hope to safely dial back the cells' hyperactivity. If successful, this approach could restore restorative sleep, improve daily cognition, and significantly elevate the quality of life for millions of patients navigating the early stages of Alzheimer's.[3]