Many aspects of the low-carb lifestyle are surprising. For example, the successful low-carber soon learns that he or she must eat fat to lose fat. Why would that be?
A possible explanation comes from a couple of studies published last summer in Cell Metabolism, one by Inagaki et al and one by Badman et al. For those who are interested in the specifics, see the PDFs here and here. These studies were performed in mice but were quite exhaustive and appear to have application to humans as well.
The requirement for eating fat to lose fat begins with a cellular receptor called peroxisome proliferator-activated receptor-alpha or PPAR-alpha for short. PPAR-alpha is a protein found inside liver cells. When dietary fat diffuses into the liver cell as fatty acids, the fatty acids are able to bind to PPAR-alpha and activate it. Activated PPAR-alpha then binds to another protein called the retinoid X receptor or RXR, and these dimerized proteins in turn are able to bind to the cell's DNA. In so doing, they enhance the production of a third protein--fibroblast growth factor-21 or FGF-21.
FGF-21 is secreted by the liver and produces several effects. In white adipose tissue, it stimulates lipid breakdown. The breakdown of stored lipids allows them to be used as fuel. In the liver, FGF-21 upregulates ketone body production. Ketone bodies provide another source of fuel. The two studies showed that production of FGF-21 was greatly enhanced when the mice were fed a low-carb/high-fat (ketogenic) diet. When few carbohydrates are provided in the diet, but the diet does contain fat, mice are able to switch to an efficient mode that allows them to consume stored fat for energy. Mice are not people, and the usual admonition applies--more research is required. But the observation that a ketogenic (low-carb/high-fat) diet allows mice to produce lots of FGF-21, mobilize fat stores and upregulate ketone production suggests an explanation for what low-carbers know by experience--you have to eat fat to lose fat.