Intermittent and extended fasting
- Thread starter Toshi
- Start date
I ate 3 meals yesterday (including some delicious non-low-carb Italian!) and will eat 3 today (including some to-be-delicious non-low-carb Thai!).
Then onto the planned open-ended extended fast after dinner tonight.
:rockon:
Usually on even my daily 16-18 hour fasts I deplete my glycogen after that first night’s sleep by riding into work, but our extended snowstorm will keep me off any sort of bike until at least Thursday, and that's optimistic. Glycogen depletion will therefore be more gradual.
Then onto the planned open-ended extended fast after dinner tonight.
:rockon:
Usually on even my daily 16-18 hour fasts I deplete my glycogen after that first night’s sleep by riding into work, but our extended snowstorm will keep me off any sort of bike until at least Thursday, and that's optimistic. Glycogen depletion will therefore be more gradual.
I've been doing IF for a while now. It was a big change from the 6 meal a day lifestyle I did when bodybuilding and powerlifting. With longevity and health being more important and plenty of evidence for DNA repair in a fasted state. I started with the 16:8 and moved it to one meal a day most the time. This year I added keto to the fasting. And the results are awesome.
25 hours in. Today I’ve had black coffee, water, electrolyte drops, and some green tea. I feel surprisingly full.
On to day 2!
Out of curiousity.
What are your main reasons for doing these extended fasts?
I dont really need to lose much weight so Im doing it mainly for the other health benefits.
But going that long without food would simply not be worth it to me unless my health would be in some sort of substantial trouble.
Eating is also a sort of pleasure for me (i do always eat super healthy) but I just dont want to give that up because its beneficial to my state of mind.
Reasons:
1) Defattening
2) Kicking the insulin resistance trait, which leads to metabolic syndrome and all that entails
3) Longevity/health reasons in general via autophagy
I'm at 39 hours or so in now. Feels good, man(.gif). Re state of mind I think it is beneficial to see that hunger (in the context of well-fed first world types) is merely a transient hormonal signal, not a sign that something is imminently wrong. Switch over to ketosis and one could live for weeks if not months just fine on fat...
Have you ever had a formal body composition test? BodPod, water immersion, DEXA, InBody? I'd guess you're in the neighborhood of 10% body fat... so you could live for at least 3 weeks on fat.Some of us ain't got that much fat, man(.gif)...
Yeah, I'm not making it to lunchtime without some breakfastSome of us ain't got that much fat, man(.gif)...
Even you could probably make it 2 weeks.Yeah, I'm not making it to lunchtime without some breakfast
/me could still make it about 2 months at this point, I reckon. 46 hours into the fast!
2 months? I'd wager you could make it nearly a year!
That's not physically possible given my level of fatness. I probably burn 3000-3500 calories per day. Thus when not eating anything that's burning about 1 lb of fat, as 1 lb fat == 3500 calories.2 months? I'd wager you could make it nearly a year!
The astute reader will note that I do not have 365 lbs of fat.
QED.
I made it to 66 hours in this week's fast. Felt fine the whole time, slightly buzzed throughout day 2 as with my prior long fast. Hunger not bad at all, especially since I played around with coffee and green tea during it to distract me.
Now back to a regular-ish two meal a day schedule for a bit.
Now back to a regular-ish two meal a day schedule for a bit.
I have a lot of fat I could loose, but fasting kind of sucks, I shake if I have not eaten anything and makes my job difficult.
If you ramp down carbs a bit while eating on a normal schedule then fasting becomes much less hard. Once fat-adapted, as they say, then swapping over to fat burning/ketosis is not even noticeable. I don't get shakes/hypoglycemic or anything like that, and day 3 has no more hunger at mealtimes than day 1.
Nice. I could go for some sushi...
I had unagi, yellowtail belly, and softshell crab/spicy salmon roll.Nice. I could go for some sushi...
Mmmm... spider roll? That's one of my go-to rolls whenever I get sushi.
I dug up this old paper from 1969:
Pi-Sunyer,Hashim, Van Itallie. Insulin and ketone response to ingestion of medium and long-chain triglycerides in man. Diabetes 1969; 18(2): 96-100.
Key figure:
My interpretation: MCT induces a rise in ketones, which is pretty well accepted. It also induces a rise in insulin!
In the authors' discussion they delve in to possibilities for why this might be so:
1) Theory 1: serum ketones in and of themselves cause an insulin response. They cite authors that did not find that direct beta-hydroxybutyrate injection causes insulin secretion, but note that at that time there was some conflicting evidence that ocanoate may lead to insulin secretion. The verdict for this one is therefore weak but possible.
2) Theory 2: there's some gastrointestinal ingested fat receptor that leads to insulin secretion. This is why they included corn oil, a long chain triglyceride as well. There was a tiny bump in insulin with the corn oil but this wasn't very impressive. Verdict similarly weak but possible.
Pi-Sunyer,Hashim, Van Itallie. Insulin and ketone response to ingestion of medium and long-chain triglycerides in man. Diabetes 1969; 18(2): 96-100.
Key figure:
My interpretation: MCT induces a rise in ketones, which is pretty well accepted. It also induces a rise in insulin!
In the authors' discussion they delve in to possibilities for why this might be so:
1) Theory 1: serum ketones in and of themselves cause an insulin response. They cite authors that did not find that direct beta-hydroxybutyrate injection causes insulin secretion, but note that at that time there was some conflicting evidence that ocanoate may lead to insulin secretion. The verdict for this one is therefore weak but possible.
2) Theory 2: there's some gastrointestinal ingested fat receptor that leads to insulin secretion. This is why they included corn oil, a long chain triglyceride as well. There was a tiny bump in insulin with the corn oil but this wasn't very impressive. Verdict similarly weak but possible.
No. Mainly because IDGAF. But I'd wager you're probably in the ballpark... but that's all just guesswork since IDGAF.
I guess 10% on the low end, 20% on the high end. You could be secretly skinny-plumpish, to use the skinny-fat analogy.
But the world will never know…
Let's go with 10, shall we?
Google Scholar
Every other day fasting in mice protects against myocardial infarction—reduced infarction size with an ischemic insult.
System-wide Benefits of Intermeal Fasting by Autophagy
Autophagy failure is associated with metabolic insufficiency. Although caloric restriction (CR) extends healthspan, its adherence in humans is poor. W…
www.sciencedirect.com
Here’s a really interesting paper on autophagy. It’s an immense paper with many experiments within it on specific knockouts and pathways.
The important things to note are these, in my opinion:
1) They fed the mice in two windows, 8-10 am, 5-7 pm. So longest stretch was 13 hours. These mice are also nocturnal, mind you.
2) Autophagy was upregulated with a peak at 11 am (their “night”) and followed a circadian rhythm, lower in the twice per day fed mice during their “day”.
3) The mice fed twice per day showed the expected insulin sensitivity, body composition etc results that we would expect—this strategy worked.
This suggests that the YouTube level conventional wisdom on autophagy is horribly incomplete. It’s circadian! It is upregulated soon after a “nighttime” meal in these fed twice per day mice, and not even during or at the end of their longest fasted period.
Much to learn yet. This supports the 16:8 approach, or even a 14:10... we need more research and less people parroting third hand crap on YouTube for clicks.
#madlad
Lipids have caught my attention recently. This is logical because one of the frequent criticisms of a ketogenic or at least low carb high fat diet, both often used in conjunction with fasting, is that it potentially could increase heart disease risk.
Rather than blindly put my hands over my eyes as Peter Attia basically accused Dave Feldman of doing in the latter's appearance on the former's podcast ("this way of eating has many other benefits so it must not be bad for my heart!") I'm going to try to understand this a bit. It happens to be that Peter Attia's own series of posts on cholesterol on his website are a great resource to use to wrap one's head around this topic.
1. What is measured in a standard lipid panel is LDL-C, the cholesterol level in LDL apolipoproteins. This is a shame, because LDL-P, the particle count of said LDL apolipoproteins, is a better measure when it comes to cardiovascular risk.
2. Aren't these two alternate measures of the same thing? Actually, not always... and they are discordant more often in the setting of metabolic syndrome.
Here is another data set, looking at people without and with metabolic syndrome who all have LDL-C from 100-118 mg/dl. X-axis units are of LDL-P:
So in non-metabolic syndrome people 22% with a on-paper fine level of LDL-C per conventional wisdom have high LDL-P, and thus are truly high risk. But with people who have metabolic syndrome then 63% of "fine" LDL-C people have high risk as they have high LDL-P! That's kind of shocking, frankly.
3. This discordance between LDL-C and LDL-P is not just of academic interest:
This is a bit busy but the message here is important. Incidence on the y-axis is of heart disease.
The lowest risk group, the blue line, has high LDL-C (again, what we usually test) but low LDL-P. (This group would be flagged as high risk per conventional test interpretation!)
The highest risk group, the red line, has low LDL-C and high LDL-P! (This group would be flagged as low risk per conventional wisdom...)
Again, recall that this latter high risk pattern, of having higher LDL-P for a given LDL-C, is associated with the metabolic syndrome as per the figures in point 2 above. These are the group of people for whom their overall metabolism is screwed up, including their lipid metabolism.
4. From these above 3 points it's clear to me at least that we should be assessing LDL-P. It cannot be inferred from LDL-C because it is discordant so often, especially with metabolic syndrome. It must be measured via a test like the NMR Lipoprofile (which I have thus ordered for myself). That link will let anyone order it for themselves, getting the draw done at a local LabCorp or Quest lab.
5) What about HDL(-C)? Turns out there's a similar phenomenon seen here, wherein HDL-P, the particle count, is more predictive. It also turns out that small size HDL particles are more protective (due to antioxidant and reverse cholesterol transport capabilities)... but when HDL-C levels are high it's typically because HDL particles got bigger.
Therefore, high HDL-C in and of itself is not a great predictor of much at all, and there have been numerous drug trial failures that have shown that artificially raising HDL-C may be at best a wash and at worst harmful.
I mentioned Dave Feldman in the top of this post, and have mentioned him before as the inspiration to order up my home lipid testing setup (which will only measure LDL-C and HDL-C, sadly--no assays short of the NMR Lipoprofile offer particle count and certainly not home ones). My setup will still be useful... but only once I have the data from my own NMR Lipoprofile test to see whether my LDL-C and LDL-P are concordant or discordant.
Anyway, one of Dave Feldman's ongoing things is a challenge where he'll pay out $1,000 who can point out a study meeting these criteria:
This actually ties right back in with points 1-4 from above. First, it refers to LDL-C by "LDL", HDL-C by "HDL" in its text so it's missing the whole particle count significance already. The second is that it stipulates having low triglycerides and high HDL(-C).
Why is low trigylcerides/high HDL-C bit significant in this setting? I theorize two reasons:
1. Metabolic syndrome in and of itself is defined by high triglycerides/low HDL-C. So we eliminate the group of people with metabolic syndrome right then and there, and thus eliminate a large amount of risk.
2. Having low triglycerides actually turns out to be a decent proxy for having a low LDL-P! (Whereas low LDL-C doesn't imply low LDL-P as we've seen.) This may be just a restatement of point 1 here, because if one has low triglycerides then one doesn't have metabolic syndrome basically by definition but bears reinforcing. Low triglycerides implies low LDL-P as shown here:
Ergo, if one has high HDL and low triglycerides then one doesn't have metabolic syndrome and is likely to have a low LDL-P, which in turn implies a low risk of heart disease... with the lowest risk per the one figure I inlined actually seen when LDL-C is high given these other prerequisites!
So what does this mean for the average person fasting +/- low fat high carb or frankly ketogenic diets?
Get your bloodwork done once you've stabilized in weight and are stable in your diet choices for at least a few weeks. (Weight changes can alter lipid profiles in and of themselves.) And when you get that bloodwork done get the NMR Lipoprofile so as to get the particle counts (LDL-P, HDL-P), not just a standard lipid profile that gives you LDL-C and HDL-C. I'd personally also check a Lp(a) level.
Plug these TC, TG, LDL-C, and HDL-C values into the risk calculator at cholesterolcode.com/report , which will give you your Framingham Offspring risk among other things. But also pay attention to the LDL-P and check to see whether it's concordant or discordant with your LDL-C. If it's discordant then that's an indicator you have an underlying metabolic problem that's increasing your risk. (But the cure for that metabolic problem is the topic of this very thread: high fats in the diet, low carbs, and intermittent and/or extended fasting.)
Rather than blindly put my hands over my eyes as Peter Attia basically accused Dave Feldman of doing in the latter's appearance on the former's podcast ("this way of eating has many other benefits so it must not be bad for my heart!") I'm going to try to understand this a bit. It happens to be that Peter Attia's own series of posts on cholesterol on his website are a great resource to use to wrap one's head around this topic.
1. What is measured in a standard lipid panel is LDL-C, the cholesterol level in LDL apolipoproteins. This is a shame, because LDL-P, the particle count of said LDL apolipoproteins, is a better measure when it comes to cardiovascular risk.
2. Aren't these two alternate measures of the same thing? Actually, not always... and they are discordant more often in the setting of metabolic syndrome.
Here is another data set, looking at people without and with metabolic syndrome who all have LDL-C from 100-118 mg/dl. X-axis units are of LDL-P:
So in non-metabolic syndrome people 22% with a on-paper fine level of LDL-C per conventional wisdom have high LDL-P, and thus are truly high risk. But with people who have metabolic syndrome then 63% of "fine" LDL-C people have high risk as they have high LDL-P! That's kind of shocking, frankly.
3. This discordance between LDL-C and LDL-P is not just of academic interest:
This is a bit busy but the message here is important. Incidence on the y-axis is of heart disease.
The lowest risk group, the blue line, has high LDL-C (again, what we usually test) but low LDL-P. (This group would be flagged as high risk per conventional test interpretation!)
The highest risk group, the red line, has low LDL-C and high LDL-P! (This group would be flagged as low risk per conventional wisdom...)
Again, recall that this latter high risk pattern, of having higher LDL-P for a given LDL-C, is associated with the metabolic syndrome as per the figures in point 2 above. These are the group of people for whom their overall metabolism is screwed up, including their lipid metabolism.
4. From these above 3 points it's clear to me at least that we should be assessing LDL-P. It cannot be inferred from LDL-C because it is discordant so often, especially with metabolic syndrome. It must be measured via a test like the NMR Lipoprofile (which I have thus ordered for myself). That link will let anyone order it for themselves, getting the draw done at a local LabCorp or Quest lab.
5) What about HDL(-C)? Turns out there's a similar phenomenon seen here, wherein HDL-P, the particle count, is more predictive. It also turns out that small size HDL particles are more protective (due to antioxidant and reverse cholesterol transport capabilities)... but when HDL-C levels are high it's typically because HDL particles got bigger.
Therefore, high HDL-C in and of itself is not a great predictor of much at all, and there have been numerous drug trial failures that have shown that artificially raising HDL-C may be at best a wash and at worst harmful.
I mentioned Dave Feldman in the top of this post, and have mentioned him before as the inspiration to order up my home lipid testing setup (which will only measure LDL-C and HDL-C, sadly--no assays short of the NMR Lipoprofile offer particle count and certainly not home ones). My setup will still be useful... but only once I have the data from my own NMR Lipoprofile test to see whether my LDL-C and LDL-P are concordant or discordant.
Anyway, one of Dave Feldman's ongoing things is a challenge where he'll pay out $1,000 who can point out a study meeting these criteria:
This actually ties right back in with points 1-4 from above. First, it refers to LDL-C by "LDL", HDL-C by "HDL" in its text so it's missing the whole particle count significance already. The second is that it stipulates having low triglycerides and high HDL(-C).
Why is low trigylcerides/high HDL-C bit significant in this setting? I theorize two reasons:
1. Metabolic syndrome in and of itself is defined by high triglycerides/low HDL-C. So we eliminate the group of people with metabolic syndrome right then and there, and thus eliminate a large amount of risk.
2. Having low triglycerides actually turns out to be a decent proxy for having a low LDL-P! (Whereas low LDL-C doesn't imply low LDL-P as we've seen.) This may be just a restatement of point 1 here, because if one has low triglycerides then one doesn't have metabolic syndrome basically by definition but bears reinforcing. Low triglycerides implies low LDL-P as shown here:
Ergo, if one has high HDL and low triglycerides then one doesn't have metabolic syndrome and is likely to have a low LDL-P, which in turn implies a low risk of heart disease... with the lowest risk per the one figure I inlined actually seen when LDL-C is high given these other prerequisites!
So what does this mean for the average person fasting +/- low fat high carb or frankly ketogenic diets?
Get your bloodwork done once you've stabilized in weight and are stable in your diet choices for at least a few weeks. (Weight changes can alter lipid profiles in and of themselves.) And when you get that bloodwork done get the NMR Lipoprofile so as to get the particle counts (LDL-P, HDL-P), not just a standard lipid profile that gives you LDL-C and HDL-C. I'd personally also check a Lp(a) level.
Plug these TC, TG, LDL-C, and HDL-C values into the risk calculator at cholesterolcode.com/report , which will give you your Framingham Offspring risk among other things. But also pay attention to the LDL-P and check to see whether it's concordant or discordant with your LDL-C. If it's discordant then that's an indicator you have an underlying metabolic problem that's increasing your risk. (But the cure for that metabolic problem is the topic of this very thread: high fats in the diet, low carbs, and intermittent and/or extended fasting.)
52.5 hours into my 72 hour fast. Still feel pretty good.
Checked my blood glucose at 48 hours just for kicks, to see what‘s going on there since I’m clearly full blown into ketosis since I’m not dying. At 48 hours I was at 71 mg/dL, and at 52.5 hours just now I was at 73 mg/dL.
These are quite low for me—I typically am in the 85-low 90s range, roughly. This is a normal phase of extended fasts, dropping down to a low, stable sugar level.
Checked my blood glucose at 48 hours just for kicks, to see what‘s going on there since I’m clearly full blown into ketosis since I’m not dying. At 48 hours I was at 71 mg/dL, and at 52.5 hours just now I was at 73 mg/dL.
These are quite low for me—I typically am in the 85-low 90s range, roughly. This is a normal phase of extended fasts, dropping down to a low, stable sugar level.
I tweeted a thing! (I had never tweeted anything of substance before.) Anyway, it refers to Dave Feldman's Low Carb Cholesterol Challenge.
Edit: first point should refer to LDL-C < LDL-P.
Edit: first point should refer to LDL-C < LDL-P.
Last edited:
Deep thought of the day that I’ve possibly shared before since this occurred to me a few weeks ago:
When I’m into a longer fast and therefore firmly in ketosis with the source of triglycerides endogenous (i.e. my own) fat, said fat will likely have a similar ratio of omega-3 to omega-6 as was in my diet when said fats were deposited initially, months or years ago.
(It wouldn’t be exactly the same since there’s probably a constant small flux in and out of any given fat cell.)
Eventually my fat deposited and retrieved will be consistently higher omega-3 in the ratio, which speaks to the possibility of a long-term slow benefit from this.
When I’m into a longer fast and therefore firmly in ketosis with the source of triglycerides endogenous (i.e. my own) fat, said fat will likely have a similar ratio of omega-3 to omega-6 as was in my diet when said fats were deposited initially, months or years ago.
(It wouldn’t be exactly the same since there’s probably a constant small flux in and out of any given fat cell.)
Eventually my fat deposited and retrieved will be consistently higher omega-3 in the ratio, which speaks to the possibility of a long-term slow benefit from this.
Turns out the super cheap used-via-eBay machine is properly broken. Back to the seller at their expense it shall go. All that was wasted on my part was $2.48 between a test cartridge and a pipette/plunger combo. But I do not know of my lipid status at this moment… : sadpanda:
Edit: Return queued up, no cost to me as it should be. "New"/used machine acquired from eBay for $11 more net cost or so. Here's to hoping this one actually works.
Edit: Return queued up, no cost to me as it should be. "New"/used machine acquired from eBay for $11 more net cost or so. Here's to hoping this one actually works.
Last edited:
I just ate a shitload of beef Burgundy, so I'd wager my lipid status is high. If that's how that shit works...
Your triglycerides (really chylomicrons) will be high for a short period. How short? Minutes? Hours? Not really clear. But this doesn’t matter.
If you do get a lipid panel formally for life insurance or the like make sure to be fasted for at least 14 hours, though, for low measured triglycerides.
Next frontier for me may be continuous glucose monitoring, yes, as a non-diabetic. This is actually a really good article on it and also the same food strategy I’m adhering to. This dovetails with the carb tests that I did the first two weeks of this month.
lilynicholsrdn.com
Edit: and a highly relevant podcast:
dranthonygustin.com
CGM Experiment: What I learned as a non-diabetic from wearing a continuous glucose monitor - Lily Nichols RDN
Many non diabetics are wearing CGMs to better understand their blood sugar patterns. Here's what I learned from wearing a continuous glucose monitor.
lilynicholsrdn.com
Edit: and a highly relevant podcast:
043: Dr. Peter Attia - How to Increase Longevity & Improve Your Quality of Life | Dr. Anthony Gustin | Farmer, Founder, Investor
Is there a secret to longevity that only centenarians know? Is a glass of wine or whiskey a day really the common denominator to living longer? As you may have guessed, these anecdotal stories of people living past 100 years of age and claiming their drink-a-day habit is at the root of their...
dranthonygustin.com
Last edited:
It still won’t matter, but I’d shed some blood and burn some test strips to find the trajectory of triglyceride levels after eating a ribeye. For science. Once I get a working unit, that is.