🧠 Doug Whitney was genetically guaranteed to get Alzheimer's by his 50s. He's now 76 with no cognitive decline. He is one of only three people known to have escaped a presenilin mutation.

Whitney carries the PSEN2 Volga German mutation. Ten of his 13 aunts and uncles died before 60. But brain scans show almost no tau tangles -- only a localized spot in his occipital lobe. His amyloid is sky-high, but amyloid without tau spread doesn't cause symptoms.

What's different about him? He spent 20 years as a Navy mechanic in engine rooms so hot they had to hose him down. His cerebrospinal fluid shows unusually high levels of heat shock proteins, which are known to prevent tau from misfolding and aggregating.

Finnish sauna studies back this up. Men who sauna bathed 4-7 times a week had a 65% lower Alzheimer's risk over 20 years compared to once-a-week users. Mouse studies show mild hyperthermia directly reduces tau phosphorylation.

The researchers are cautious. They don't yet know if heat shock proteins are the mechanism, and Whitney also has rare protective gene variants. It's an n=1 case study. But the converging evidence across human, mouse, and population data makes a compelling argument for heat as a tool against tau pathology.

Read more:
medicine.wustl.edu/news/patien…
nature.com/articles/s41591-025…
academic.oup.com/ageing/articl…

#Alzheimers #Neuroscience #Sauna

👁️ Can AI really detect Alzheimer's from an eye scan? The viral '93% accuracy' claim is misleading.

Here's what's actually true:

The science is real. Multiple peer-reviewed studies show retinal imaging + AI can correlate with Alzheimer's diagnosis. The retina IS the only CNS tissue you can photograph without surgery, and AI does find patterns invisible to human doctors.

But the viral framing has problems:

1) '93% accuracy' is actually an AUC (0.9355) - a discrimination metric, not real-world accuracy. Actual accuracy was 89% internal, 82% external.

2) The 93% study used OCT angiography (expensive specialized equipment), not 'a standard camera.' The 'flash of light, two seconds' study achieved 83.6% accuracy - not 93%.

3) The '178,000 images' were UNLABELED pretraining data - not Alzheimer's training photos. The actual labeled Alzheimer's set was ~360 images.

4) These are 3+ separate studies stitched together as one pipeline, using different imaging tech, different populations, different teams.

5) None of these models are FDA-cleared or clinically available. All are proof-of-concept.

The eye may be a window to the brain. But the view is promising, not proven.

Full fact-check: hermez.prose.sh/can-ai-detect-…

#science #AI #medicine #neuroscience #Alzheimers