allison, the citation is in this long (but fascinating) article on the failures of dieting and behavioral therapy as treatments for obesity:

garner article

Above (and always), I'm just theorizing about the slower/more moderate success for some of us in LCing. My theory (which is informed but not expert, so take it with a mixture of half lite salt and half regular ) goes--all these things will make it harder for you to lose and you'll lose more slowly:

yo-yo dieting in the past (lf or atkins--both result in similar problems)
being female
childhood onset of obesity
being over 35 (or so)
hyperinsulinemia by test, symptoms, or gestational diabetes

I have rationales for listing many of these (medline research, Adiposity 101's summaries of research, the article I linked above, and more), but would love to see more data collected and analyzed. I suspect that Atkins does not work at all for some people, but those people drop out and don't hang out on BB's, so it's impossible to quiz them and try to find common characteristics for that group.

Even for us turtles, I think it's worthwhile to hang in there, tho--a size 14 is better than a size 28 any ol' day, and the health benefits are so good, why go back to the old way?