I was recently listening to a "Livin' la Vida Low-Carb" podcast where Jimmy Moore was interviewing Dave Feldman from the Cholesterol Code blog. Dave is testing a hypothesis that his LDL-C measurement is inversely correlated to the amount of fat he consumes in his diet during a few days prior to his lab work. He believes that the liver upregulates the production of vLDL in response to the drop of chylomicrons carrying dietary fat in the blood (@ 14m 57s). During the podcast the idea of the the LDL concentrations in the blood increasing as the length of the fast done before a test is done increases is also discussed. Jimmy questioned if longer fasts would result in higher LDL concentrations (@ 32m 28s). Dave speculates that LDL concentration increases would still be seen with fasting periods around 72 hours but that he has not done any testing to support this hypothesis yet.
Coincidentally, a few weeks before I had heard this podcast, I had decided to do a 24 hour fast before my routine lab work to see if that affected any of my tests. I was shocked to find that my LDL-C had shot up 99 mg/dl in my June testing after it had been testing exactly the same on my two previous rounds of lab work. After hearing the podcast with Jimmy and Dave, I paid for another lipid profile out of pocket to see if my LDL-C results would come back down after a 14 hour fast. My LDL-C did drop back down by 67 mg/dl with a 13 hour fast. It would be interesting to see how the LDL-C changed day to day during a continuous 7 day water fast. For now though, I will stick with the 12-14 hour fasting period prior to my lab work.
Here are the results from my blood lipid tests discussed above. Note that all other measurement stay relatively consistent compared to the large jump in LDL in June after the 24 hour fast:
I should note all my numbers from cholesterolcode.com ultimately come from blood draws following a 14 hour fast since last food consumed. This includes all results high and low. This is due to the required fasting period being 12-14hr. In other words, that particular window doesn't appear to affect the DFI.
I think to truly test a fast against the DFI, it will need to go 3 or more days. I'll eventually test this myself, probably later this year. But I wouldn't mind if you beat me to it and shared the data. :D