I originally posted this in its own thread, which I'm not allowed to delete, for some reason.
The initial article does a poor job of communicating the study results, so I wrote the reporter for clarification.
The reporter replied (a little huffy for my nitpicking) and gave me a link to the
full study paper. The graph on p. CR87 tells the basic story, but here's a rough summary:
Maximum Insulin Responses for Test Foods (mean values ±standard error of mean, in pmol-min/L):
White Bread: 9,413.7
High-carb bar: 16,241.1
moderate-carb bar: 12,718.2
low-carb bar: 6,942.9
Chicken: 2,105.5
I found the shape of the response curves interesting, too. All of the energy bars had a rapid rise up to the 30 minutes, followed by a sharp decline. I wonder how that affects the sensation of hunger?
Also noteworthy is the small sample size (20), the age range of the test subjects (21-40), their generally good health, and their high-carb diet (3-day diet records were verified before eating each test meal to ensure consumption of at least 150 g carbohydrate/day.) You have to wonder how that last factor affected the results.
The p values for the study were quite good, IMHO.
If anyone want the full text of the interchange between me and the reporter, send me a PM.
Even after clarifying the results, do they matter? Searching on another low-carb forum, I found some thought-provoking commentary on that question
by Orion.