So the major factor here is generally proposed to be dietary fiber. People on a high-fiber diet are healthy, and that's really the major separating variable bet

So the major factor here is generally proposed to be dietary fiber. People on a high-fiber diet are healthy, and that's really the major separating variable between laboratory diets like the one in this article and real diets that make you fat. Going off the description of dishes in the article, I'd estimate the subjects had a daily fiber consumption of 40-60 grams, most of it soluble. Studies that look at meat consumption generally do not control for fiber, though they usually mention the correlation in many populations between increased meat consumption and less fiber consumption. Of course there are correlations with fiber and CAD and cholesterol. But we already knew that. It's a rather significant confounding factor when anyone uses an extremely high-fiber test diet to argue for the health value of various other characteristics of said diet. This author does it all over the place. Frankly, a mere 12 pound loss over five years may not really be too impressive, depending on where they're starting. I imagine Mark Sisson could post better results than that. Similarly, I imagine the guy who wrote The China Study could post better results than that... most diets will! What they've got statistically significant, but pretty much any diet that cuts out processed foods alone will show statistically significant changes... The sole really impressive thing is the prostate cancer improvements: but there are too many confounding factors to determine what, precisely, is causing the improvements. Maybe there's some anticancer compound in a vegetable emphasized in the diet. Diindolylmethane comes to mind, or beta-caryophyllene.