Contamination is possible, but it's hard to explain the results that way. Definitely needs replication, of course.

I don't see the point of the samples all being from older individuals. Even if it exclusively occurs in older individuals (unlikely), it would still mean that the blood-brain barrier is not 100% effective (which we already know to be true).