On: Semaglutide linked to lower predicted dementia risk

[brookst]

Or, less conspiratorially, if correlation can be proven and causation is plausible but unproven, there may be benefits.

Proving correlation but not causation does NOT rule out causation.