Anne Trafton's write-up of Li-Huei Tsai's lab reports that light flickering at 40 hertz drove gamma rhythms in mice bred to develop Alzheimer's pathology, and that an hour of it reduced beta-amyloid by roughly half in the visual cortex of early-stage animals — an effect that washed out within 24 hours. Seven days of daily hour-long sessions reduced both plaques and free-floating amyloid, along with abnormally modified tau, and microglia became more active in clearing amyloid.
The important part is the hedge, and Tsai supplies it herself in the piece. These are mice, the stimulation is visual, and the treated region is the one nearest the eyes.
We include it because it is the origin of nearly every "40 Hz" claim now attached to sound-and-light devices, and because reading the original makes clear how narrow the 2016 result actually was.