But my bias is probably showing... and when it comes down to it, I try not to care about epidemiology, whether it supports my views or no...

But pretty random with the qualifier "to me" isn't the same as random, and it certainly isn't "vanished."