Circadian disruption as a measurable variable
In the first study, published in *SLEEP*, researchers exposed Alzheimer's disease models to dim nighttime light calibrated to typical household or street-level exposure. Outputs:
- Rhythm stability decreased; fragmentation increased.
- Amyloid burden rose modestly.
- Microglial activity shifted toward a pro-inflammatory, immune-activated state.
Lead author Adam Bachstetter, Ph.D., associate professor of neuroscience at the UK College of Medicine: "These studies examine why sleep and daily biological rhythms become disrupted in Alzheimer's disease, and whether those disruptions are influenced by both the outside environment and inflammation inside the brain."
An accompanying editorial in *SLEEP* characterized artificial light at night as a modifiable environmental variable acting through circadian and neuroimmune pathways.
Inflammation, not amyloid, as the proximate driver
The second study, published in *Alzheimer's & Dementia*, tracked sleep, activity rhythms, cognition, and inflammatory signaling longitudinally. Critical timing: disrupted sleep and fragmented rhythms surfaced in midlife — before major memory deficits appeared.
The team then administered MW151, a compound developed by Linda Van Eldik, Ph.D., that targets excessive glial inflammatory signaling. Results:
- Sleep architecture improved.
- Daily rhythms normalized.
- Amyloid levels did not change.
Bachstetter: "We now know that sleep can be improved without reducing amyloid. This finding separates sleep disruption from amyloid load and points to neuroinflammatory signaling as a modifiable driver of poor sleep in Alzheimer's-related pathology."
What the data actually licenses
Two variables, both preclinical, now warrant quantification:
- Ambient light exposure during the pre-sleep window. The study's "dim light" condition was matched to real-world sources — TVs, phones, hallway fixtures, streetlights. Replication in human sleep architecture is the missing link.
- Sleep fragmentation as a longitudinal biomarker. In the models, rhythm disruption preceded cognitive decline by a measurable interval. Whether the same lag holds in humans is unknown.
MW151 efficacy data in human populations remains pending. The mechanistic signal is clean; the clinical translation is not.