Paperonium

Diabetes · Diabetes Care · 2023

Summary: more liver fat, slower insulin removal. Its printed numbers are too weak to show that.

1 in 100every time1 in 20the usual barrarer than1 in 20The summary: a real linkabout8 in 20Its own numbers: could be chance
1 in 100every time1 in 20the usual barrarer than1 in 20The summary:a real linkabout8 in 20Its own numbers:could be chance

How often chance alone would give a liver-fat link this strong, if none existed

The paper says

In newly diagnosed type 2 diabetes, the summary says, more liver fat goes with slower insulin removal from the blood, a link it calls too strong for chance. Two tests of removal give link strengths of −0.037 and −0.005 (0 is none; −1 is perfect).

Its own data show

The study’s 489 people with type 2 are the most these links could use; fewer would demand a stronger link. With 489, a link must reach −0.089 or +0.089 for chance to give it under 1 time in 20. Chance gives one as strong as −0.037 about 8 times in 20, and −0.005 about 18.

Chance alone makes small links appear even where none exists, so a link counts only if chance would rarely make one that strong. The summary’s liver-fat conclusion for type 2 draws on these two. Either the printed numbers are wrong (on the paper’s own chart, the −0.037 link looks closer to −0.18) or calling them real is.

“The summary” means the paper’s abstract: the short summary printed at its top, which is the part most readers see.

Read the paper →Full technical write-up →

Reduced Insulin Clearance Differently Relates to Increased Liver Lipid Content and Worse Glycemic Control in Recent-Onset Type 2 and Type 1 Diabetes · Diabetes Care (2023) · doi:10.2337/dc23-1267