Learn more about Peter Attia’s digital program
Nothing like fasting to put fasting on one’s mind. I just finished a one-week fast as part of my quarterly “KFK nothingburger,” as Bob Kaplan calls it (1-week of keto, followed by 1-week of fasting, followed by 1-week of keto).
A lot of folks asked me questions this week on social about why I do this quarterly 3-week KFK fasting protocol and, more importantly, if I think this protocol is the “best” fasting protocol. Unfortunately, the answer is: We have no clue what the best protocol is for intermittent fasting (fasting that is done periodically, and not to be confused with time restricted feeding).
Obviously, there are an infinite number of permutations for intermittent fasting (how long you fast x how frequently you fast). Furthermore, if you include hypocaloric (i.e., limited calories beyond water) fasting, it becomes even greater. In trying to simplify the problem I scratched out what I think are the fewest “discrete” cases of fasting or hypocaloric fasting which basically means toggling these 3 variables. I’ve attempted to simplify to the fewest cases possible.
1. How much you reduce by?
Let’s simplify and suggest the following: 50% reduction; 75% reduction; 100% reduction (pure water fast) from baseline.
2. How long you reduce?
Let’s simplify and suggest the following: 1 day, 3 days, 5 days, 7 days, and 10 days.
3. How often you cycle?*
Let’s simplify and suggest the following: every other day; every week; every month, every quarter.
* Note: some of the options need to be subtracted out because 3-, 5-, 7-, 10-day can’t be cycled daily; 7-, 10-day can’t be cycled every week; 5-day 100% reduction can’t really be done weekly either.
Crude calculation = 3 x 5 x 4 = 60, but we need to strip out above (7 cases), so call it 50 to be conservative! Think of how much I simplified this problem (I didn’t even get into what you reduce, although 100% fasting certainly simplifies this one…), and yet there are at least 50 plausible, physiologically “reasonable” approaches to fasting. Want one more wrinkle? I’ve said nothing about what one might eat during the longer non-fasting periods. Surely, that must matter, too?
So what can we take away from this?
Point 1: Anyone who tells you *their* fasting protocol is the best is either lying or does not understand the most basic aspects of finite combinatorial mathematics.
Point 2: We need much better tools/assays to measure the cellular response to fasting (e.g., mTOR activity, PI3K activity, AMPK activity) to have a hope of sorting this out.
And, by the way, who is to say any two people will have a “best” response to any one permutation?
So, while I’ve chosen one model—7 days of complete fasting every quarter (not to mention bookending it with a weekly ketogenic diet)—I have no way of knowing if there are more efficacious models out there. In fact, if you think about it, there almost must be, right?
– Peter
