January 11, 2008

How Much Protein Do You Need on A Lower Carb Diet?

I've replaced my admittedly confusing web page about how to calculate your protein needs on a low carb diet which is found on my Low Carb Facts and Figures site with a new automatic on-line calculator. So now you can easily figure out how much protein you need to eat on your own low carb diet based on your own daily carb intake, your weight, and how long you have been eating a low carb diet.

The assumptions on which it is based are documented on the calculator pages. They are drawn from the research Lyle MacDonald published years ago in his book, The Ketogenic Diet.

You can try out the calculator HERE

The most important thing I learned from Lyle's book is something few nutritionists seem to understand. While it is true that the brain needs some glucose to run properly, since unlike all the other organs in the body it cannot run entirely on ketones or free fatty acids, 58% of the protein grams you eat can be converted into glucose in the liver. So there is no need to eat any carbohydrate at all.

But even more importantly, as Lyle documents in his book, after three weeks in a ketogenic state, the brain adapts and needs much less glucose than before. So while it is true that you need about 100 grams of glucose when you start eating a ketogenic low carb diet--one that provides less than 100 grams of carbohydrate a day--after three weeks on a ketogenic diet your brain needs only 40 grams of glucose.

Dr. Bernstein advises in his book that if you are eating his 6-12-12 grams of carb per meal regimen and are not losing weight, you should cut your protein intake. This should only be done after you have been on the diet long enough to have dropped your protein need, and you are trying this approach do double check that the amount of protein you are eating is enough to keep your muscles repaired.

If you are not eating a ketogenic diet, you will need about 100 grams of glucose to run your brain. You can use the calculator to calculate your protein need in that case, just select that length of diet option for "less than 3 weeks" on a low carb diet.


ItsTheWooo said...

Ah, ok... so that explains why I may have had hypoglycemia and hunger yesterday without pro.

When I discovered how much protein affected my blood sugar, I started cutting back more and more, and yesterday I took in both very low carb and low protein (40 g) by the end of the night I was shaking a lot and fantasizing about meat LOL (I don't crave carbs when I'm low on sugar, I crave protein like crazy!)

It makes sense, considering my protein need is actually more like 100 grams since I do non-ketogenic low carb. On day one of less protein I was getting signs of ketosis, but on day three (when I had the hypoglycemia) I had done a few things which I know are against an easy ketosis.

So basically without enough carbs or protein my sugar tanked and I wound up driving to a supermarket to devour about half of a chicken LOL. I'm paying for it now, though, my blood sugar is totally out of control again today (as was common for me before discovering how protein in excess is BAD)... urge to snack and eat to the point where I'm limited only by self restraint (or food availability if self restraint is lacking)... all tell tale signs.

Red Sphynx said...

Jenny -

Cool new calculator. Thanks!

The body can make glucose from protein or fat. Each molecule of triglyceride fat is broken into three acids and a glycerol. The body can turn glycerol into glucose. By my calcs, that means that every 10g of fat yields 2g of glucose. If you're metabolizing 1500 kcal/d and 60% of your energy is from fat (9kcal/g) then that's 100g/d of fat. Should yield 20g of glucose.

Shouldn't that offset some of the protein need?

Adam Becker Sr

Jenny said...


I have seen varying estimates of how much fat can be turned into glucose ranging from 10% to 0%. Since Dr. Bernstein claims that dietary fat cannot be converted to glucose, based on his many years of observing the effects of small amounts of food on people with no insulin production at all, I tend to believe him.

I'm not 100% certain that the very high protein amounts that the calculator prescribes for new dieters are entirely necessary. I don't myself eat that much protein ever, but the 3 week values certainly work for me. And over the year when I did my major weight loss, I found it necessary to eat to that amount of protein to lose.

But I try to base everything I pass on to others on good science, and Lyle's book is full of citations to decent studies and they are the ones that nutritionists misrepresent to clueless diabetes newbies, too. So that is why I go with that data in my calculator.

To me, the most important take-away fact about protein is that once involved in a ketogenic diet we don't need nearly as much of it as people think they do, and this misunderstanding is to blame for a lot of low carb diet stalling.

Red Sphynx said...

I've been hearing Lyle MacDonald's name bruited on USENET for years. You praise 'his book'. I see three books by him on Amazon. Each runs $40 to $50. Before I rush out and buy all three - which one do you mean?

Jenny said...


I cite Lyle's book, "The Ketogenic Diet." I bought mine back when it was first published in the late 90s . It cites dietary research he collected in the mid 90s, and has not been updated as far as I know.

Much of the book discusses cyclical ketogenic dieting which is something of interest only to body builders. So because of high cost, and the book's age, I would hesitate to recommend some buy it now.

Because of the lack of published research on the physiology very low carb diets, the research he cites in "The Ketogenic diet" is mostly research on the physiology of fasting and starvation. But because these are also ketogenic states, the findings are relevant.

That said, I found it the very best book for explaining the physiology of ketogenic dieting, and it made me understand that there was no "magic" to what happens in a ketogenic state, the way that Atkins would have you believe there is.

OTOH, there is still a decade later almost no good research on the physiology of ketogenic dieting.

Just this month some imbecile published a study that has been much cited in the press which was supposedly about "low carb dieting"--where the "low carb" diet was one providing 200 grams of carbs.

Another recent much cited "low carb" study lasted only for two weeks, which isn't long enough for the body to fully go into a ketogenic state.

It makes you wonder who these people are who claim to be nutritional researchers. And what mail order "school" gave them their Ph.D.s.

Lyle continues to write, but his focus is on bodybuilding and hence his writings are not all that helpful to those of us whose main dietary concern is blood sugar control.

The book was self-published a decade ago (when self-publishing was a very expensive undertaking) and aimed at a tiny market, which explains the price.

R. Francis said...


Thanks for the calculator page. Much, much easier, even for a math-head like me, to deal with. As one would expect, the results are pretty much what I'd worked out, but I prefer a machine to do it. :)

My protein need's also about 100g, and I have to tell you, I often find it difficult to meet it. I have three eggs for breakfast; that's about 18g. Occasionally for a late night snack I'll a couple of pieces of string cheese: 12g. That leaves 60g, or about 10oz of meat, at the other two meals. Phew, that's quite a bit if I'm not going out for a steak or whatever. I may be going back to a whey protein drink or two a day, I guess.

MAC said...

In "Life w/o Bread", Allan and Lutz say that your brain uses 150-200 grams of ENERGY daily mainly from glucose. Dr. Eades of Protein Power says your TOTAL daily need is 200 grams for all metabolic processes. What is Lyle's source for 100 grams of glucose daily? These figures I would assume are prior to be in a ketonogenic state.

Jenny said...


I just hauled out my copy of The Ketogenic Diet, what Lyle reports, based on studies is that the brain actually can run partially on ketones and that over time it increases the amount of ketones it burns.

He state, "IN a non-ketotic state, the brain utilitzes roughly 100 grams of glucose per day." And this is footnoted to Cahill, G, "Starvation in man" in NEJM 1970, 282-:668-675 and Sokoloff L, "Metabolism of ketone bodies by the brain. Ann Rev Med 1973, 24:271-280.

He further states that unless you drop below 100 grams of carbs you won't induce ketosis. Once ketosis is induced, the body starts to preserve protein and breaks down less of it to produce glucose.

I also see that the research he cites claims that 10% of fat is also turned into glucose. If this is true, perhaps we would do better to use less protein to induce fat breakdown.

A chart he gives explains hat to supply 100 grams of glucose, 180 grams of triglyceride and 75 grams of protein are broken down, which produce 93 grams of glucose.

He then states that by 3 weeks only 40 grams of glucose are needed. If these come partially from fat, perhaps that is another reason to go lighter on protein intake.

All this is in page 41-49 with several more journal citations.

My personal take on this is that to be on the safe side we should probably start out with a higher protein intake, but after 3 weeks cutting protein down should increase fat loss.

This backs up Bernstein's suggestion to cut protein to lose fat.

I don't thin Eades cites anything scholarly in his books to back up anything he states, so I can't say where he gets his "facts."

MAC said...

Thanks for your answer Jenny. Lyle seems to have more facts at hand than some of the sources I sited. Probably due to the need to keep their books somewhat simplified I assume.

In defense of Dr. Eades, I believe somewhere on his website, he has mentioned that the publishers left out his sources by design.

The statement about TAGs and protein being used to yield glucose jives with Lutz in that he says fat metabolism is used to fuel (ATP) gluconeogenesis.

Is your protein amount adjusted to reflect lean body mass? Every reference to protein requirements I have seen refers to lean body mass and not weight. Just curious.

BTW. Great website(s). I like someone who looks for the truth and is willing to examine the cons and pros of an approach.

R. Francis said...

I'm not sure where that assertion about Eades comes from, anyway. Protein Power, for example, would recommend 200g of protein... for an athlete who has a lean mass of 222 lbs. For someone like me (moderately active, perhaps 160 lbs of lean mass -- sadly 230 lbs total), it's more like 96g, which is actually a few grams under what the Lyle computation gives for me. So unless he went radically different in his estimates in later works (I've read PPLP but don't have a copy handy to check), then I'm not sure where mac got the 200g figure from.

Anonymous said...

It doesn't come from his book. It comes from one of his posts on his blog. I was just referencing his book in case someone was not familiar with him.

Here is the post in case anyone is interested:

He says in the blog that once ketosis kicks in your TOTAL glucose requirement drops to 60-70 grams.

Jenny said...

The thing none of the other books mentioned which I found to be the most helpful was the finding that the glucose requirement drops to 40 g after a few weeks. I don't have my copy of Protein Powder any more, having lent it to someone who didn't return it, so I can't check what they recommend, but my memory was that they implied you should keep eating at the initial high protein level.

Protein Power was the first low carb diet I used and I lost weight at first but then stalled out pretty fast and very completely.

The next time I did a long term low carb diet I ate a lot less protein (and calories) and that time I reached and exceeded my weight goal, with no loss of lean body mass.

I wonder myself about how applicable those generic statistics are not only in terms of lean body mass but also age and gender. My guess is that they may be high for those of us who deal with age-related metabolic slowing.

And since NO research was done in women until very recently, it is quite likely that the numbers are wrong for women. Our metabolisms, unlike mens' are optimized to retain fat and ensure survival under harsh conditions. So we may have other things going on that are relevant.

The difference in the rate at which men are able to lose weight using a relatively easy diet and the struggles most women go through to lose a pound should teach anyone with half a brain that female bodies fight a lot harder to retain fat.

MAC said...

To r.francis: Eades is talking about 200 grams of glucose. Should have been clearer about what I was stating.

jpatti said...

I think Bernstein is wrong about dietary fat not converting to glucose.

Fat can only theoretically convert at a maximum of 10%. I eat around 1400 kcal at around 60% fat, which means around 30g fat per meal. The most that would convert to is 3g glucose. 3g glucose is going to make such a small difference in bg that it's going to be within the margin of error of most meters.

I don't "count" stevia-sweetened lemonade, coffee with cream, coconut oil/cocoa/peanut butter/DaVinci candies... all of which run 1-3g carb per serving, because they have next to no measurable effect on my bg.

I found when learning to dose insulin that I had to account for carb and protein, but didn't have to account for fat. I don't presume this is because fat doesn't convert, but because the amounts you'd need to eat to see a bg rise are just insane - no one sits down and drinks a cup of oil (a full cup of vegetable oil could theoretically convert to about 21g of glucose).

So even if my lunch were a few lettuce leaves floating in a cup of olive oil, it'd only take a unit of insulin to cover that (even though it's almost a 2000 calorie meal), whereas my normal dose for my normal low-carb meals run around 6-10 units.

Since I've never craved a cup of oil in my life, the amount of glucose from fat is utterly irrelevant just cause it's so tiny.

Also, I doubt fat really does convert much if there's sufficient glucose available, just as I doubt protein converts much if you eat a high-carb diet. These things are very much dependent on the current biochemical balance of your body overall, which varies a lot from person-to-person. 10% for fat and 58% for protein are *theoretical* maximums based on the stoichiometry of the molecules; they're not necessarily what actually happens. Gluceoneogeneis costs a lot, it's a lot easier for the body to use glucose available than to make more if it doesn't have to.

It's very much a YMMV thing. For instance, I've read it's only T1s who have to account for protein, but I'm a T2 and I have to count protein in my insulin dosing. There's just so many variables involved that there's no way to know without testing yourself.

jpatti said...

I forgot to add... this calculator thinks I need around 100g protein per day; the calculations in the PP book come out to around 90g for me. So there isn't a lot of difference between the two.

I generally aim at about 30g protein/meal... the exception being if I have yogurt for breakfast or lunch, it's only about 15g, so I try to have more protein at dinner that day.

This is not a huge amount of protein. A cup of yogurt at breakfast, a cup of cottage cheese at lunch, and a whole chicken breast at dinner covers my protein needs. It's not the huge amounts people unfamiliar with low-carb imagine we eat.

Jenny said...


If you check out the assumptions I used when constructing the calculator, you'll see I threw in an extra 10 grams as a safety measure.