Good point below, Travis. In fact, that is the point. It is not caloric restriction per se that mediates longevity; it is a change in, specifically down regulating, particular metabolic pathways that elicits the shift in genetic expression towards a longevity phenotype. These pathways must be intact as they are in non-genetically modified people for CR to have significant benefit. The latest science is showing that these pathways likely pertain to insulin and mTOR that are conserved in virtually all animal species, and I suspect the leptin pathway will be shown to be very significant in mammals including humans. This places (non-fiber) carbohydrates and excessive protein in the middle of the fray.
The calorie restriction protocol has been shown to extend lifespan in the vast majority of species studied. In many of these species, obesity is a non issue such as flies, worms, spiders, and many others. DBA2 mice are being studied as in the paper cited by Travis precisely because they are a known exception to the CR extending lifespan rule. But for every exception, you could likely find hundreds of studies since the 1930s that show caloric restriction extends lifespan in other strains and species. Research is being done to try to find out why exceptions exist.
The DBA 2 mice are quite weird aside from their well known non response to caloric restriction. CR increases their blood glucose rather than decreasing it as in almost all other species, including humans. This can hardly be called an insignificant difference. They also die of unusual diseases and also have some strange genetic defects affecting their brain manifested as hearing impairments and epilepsy I believe.
It is known that if you dietary restrict any animal too much it will tip the scale beyond hormesis and decrease lifespan. A 30% reduction in calories may be too much for this strain.
Attributing the lack of effect of caloric restriction on lifespan extension in this strain of mice to their being thin and non obese is far from conclusive and a hugh scientific leap, as these mice have many other genetic and metabolically relevant differences other than being thin. I would not sing the death knell to CR just yet.
It is quite possible and I would say probable that the lack of lifespan extension in this mouse strain is due to its inability to burn fat as opposed to glucose, secondary to the strange genetic abnormalities that the strain possesses including the maintenance of elevated glucose and therefore the likelihood and necessity of constantly burning it.
To summarize, the advantage to health and lifespan that caloric restriction elicits is not due to CR per se, but to changes in metabolic hormone signaling. Part and parcel of that change is the increased ability to effectively burn fat.
To quote from p 18 of my book, "I am often asked to briefly summarize what it is that establishes health. I can do this in a single sentence. 'Health and life span is determined by the proportion of fat versus sugar people burn throughout their lifetime'. The more fat you burn as fuel, the healthier you will be. The more sugar you burn as fuel, the more disease-ridden you will be, and the shorter will likely be your life."