The headline is broadly defensible, but the qualifications matter. Effect sizes vary by population, the strongest claims rest on shorter trials, and credible voices push back on how it's typically framed.
So the really rigorous way to do this would be I'd put a catheter in your arm and I would sample your blood every 30 or 60 minutes over the course of a day while you ate. And I'd integrate that function and there would be an area under the curve of insulin, and that's actually the number I care about. But since I can't do that outside of a research setting, I rely on these other proxies.
Every Sunday: the week’s new conflicts and verdict changes — and nothing else.
Native comments, Twitter mentions, and Reddit threads about this claim — surfaced together so the conversation isn't fragmented across platforms.
Bookmarking — the dossier-vs-overview split is the right call. Most of the time I want overview; sometimes I want receipts.
Would love a "what would change this verdict" RSS feed. Sign me up if it exists.