Losing as much "lean mass" as fat mass isn't really a failure, so much as it's the expected result. Dexa doesn't measure protein content of the tissues, it just lumps fat and lean into two groups. So you really can't say from a Dexa how much the loss of lean mass is just the normal reduction in water weight.

Even if the lean mass is assumed to be lost in the usual ratio you'd find it in in muscle or lean organs--three grams of water per gram of protein means that each gram of lean mass lost would amount to one calorie per gram of lean mass (four grams of lean mass equals three grams of water plus one of protein, equals four calories). So since fat is 9 calories per gram, a 50:50 fat:lean ratio lost gives a 9:1 ratio of calories lost as fat vs protein during the fast. Some people can hope for a better ratio of fat to lean lost over an extended water fast--but for most people, counting on this is unrealistic. It's already very much skewed towards burning fat for energy.

I think Jimmy is probably disappointed about the results of the refeed part of his study--obviously maintenance is what he needs to work on, rather than weight-loss. I think what can be improved with the refeed is the ratio of lean mass to fat mass gained--unlike weight loss during the fast, this ratio is more amenable to manipulation through working out, perhaps increasing the protein in the diet, etc. A prolonged fast is a protein depleting method of weightloss, so a somewhat higher ratio of protein in the refeed diet, at least for a couple of weeks or so, might be something a person could benefit from. Coming out of a fast, you've got all those anabolic adaptations that Dr. Fung has written about, a person can often achieve a positive nitrogen balance while eating less protein than they would have needed just to maintain their lean mass previous to the fast--but that doesn't mean that they can't benefit from a little extra protein after the fast. In a similar vein--people on steroids don't have a higher protein requirement, it will actually take less dietary protein for them to maintain their lean mass, but it may take more than their usual protein intake for them to get the full benefit of muscle growth.

Jimmy says he just ate normal in February. So what does that mean? During his one year nutritional ketosis experiment, he found that about 80 grams of protein a day, 100 grams on days he worked out, along with 30 grams of total, not net, carbs, kept him in ketosis and losing weight. If you look at Jimmy's lean loss during the fast, it's about 10 pounds over 30 days, so a third of a pound per day. That's about 151 grams, divide by four for protein grams--about 37.5 grams of protein lost per day during the fast.

If you assume these are the protein numbers--this is a pretty unlikely scenario. Protein oxidation doesn't generally go down when you eat protein, it goes up. If Jimmy ate 80 grams of protein during the refeed, and regained 10 pounds of lean mass, 2.5 pounds of it protein, then his protein oxidation rate over the refeed would have had to be 42.5 grams a day--a scant 5 grams a day over what it was during the fast. Which sort of seems unlikely. So--maybe Jimmy is eating more protein than he thinks he is, I know from experience that if you're not weighing out every morsel, keeping fat at 85 percent of calories becomes very difficult. The other possibility is that using Dexa as a measure of lean mass is questionable. A four pound variation in water weight that has little to do with body protein stores is fairly normal. The old water fasting studies, where obese subjects experienced a decrease in protein oxidation down to 20 grams or so a day involved nitrogen balance studies, not perfect, but at least directly related to protein.


http://www.ncbi.nlm.nih.gov/pmc/art...t00645-0194.pdf

This is just a ten day study, comparing low fat and ketogenic 800 calorie diets vs fasting.



What if somebody did a before and after Dexa scan? They'd get a 2:1 ratio of lean to fat being lost. Meanwhile, if you look at fat vs protein--they're burning 291 calories as fat per 26.8 calories as protein. They'd have started with replete glycogen stores, so that changes things a little, but not much averaged over 10 days. Pretty close to that 9:1 fat: protein calorie ratio I calculated for Jimmy--but with a wildly different fat:"lean" ratio of weight loss.