Are carbs stored as fat?

Hint: “no”

DNL proper, Op. 144

Lots of metabolism talk below, but first a brief intro.  My “muse,” if you will.

Taubes’ recent article in the BMJ (Taubes, 2013full text) generated some interesting feedback.

In the original article, Taubes basically re-states his philosophy on obesity.  Nothing new.  But one rebuttal by Cottrell got under my skin (Cottrell, 2013), and Taubes’ response was woefully inadequate.

Cottrell [sic]: “A third incorrect assertion is that obesity can be attributed to the conversion of carbohydrate to fat. This is an unsatisfactory explanation of obesity, because this route is a minor pathway to depot fat in humans, even under conditions of substantial overfeeding of sugars to obese subjects.  An unproved assumption is that the hypothetical diversion of carbohydrate energy into fat storage leaves the subject hungry, thus stimulating overeating.”


Cottrell set up a straw man and handily took it down.  The primary mechanism whereby excess carbs contribute to obesity is via insulin’s effects on adipose tissue.  Even if you’re eating very little fat, insulin will cause it to get stored.  Insulin is very good at this – it is actually far more potent at stimulating fat storage than it is at stimulating glucose uptake (eg, Insulin vs. fat metabolism FTW).  Cottrell’s straw man is that excess carbs themselves are stored as fat.  This does not occur to any appreciable extent in humans.  Here is why I believe that to be true, from one of most insightful and informative studies on the topic IMHO.

Glycogen storage capacity and de novo lipogenesis during massive carbohydrate overfeeding in man (Acheson et al., 1988)

The “massive carbohydrate overfeeding” entailed a 5000 kcal diet with over 1000 grams of carbs.  Per day.  Do not try this at home.

If anything would stimulate carb-to-fat de novo lipogenesis proper, 1000 grams of carbs would do it.  After one week of this protocol, the participants gained roughly 10 pounds.  The goal was to keep them in an energy surplus of 1500 kcal, so on day one, 1500 additional kcal were fed.  But energy expenditure cleverly responded with a 370 kcal boost, so the researchers retaliated by feeding 1870 additional kcal on day two.  Energy expenditure responded yet again with +100 kcal, so the researchers adapted.  And so on, and so forth.

Graphically, it looked like this (overfeeding took place from days 3 through 10):Acheson TEE

Lipolysis was completely shut down, and although they weren’t eating very much fat, you can bet all of it was getting stored [as body fat].  A lot of the excess carbs were burned, in 2 ways: 1) they stopped the burning of all other fuels and thus carbs comprised the only fuel contributing to metabolic rate; and 2) energy expenditure increased by 35%! so the excess carbs were used for regular metabolism plus the extra 35%.  Glycogen storage was maxed out – much of the 10 pounds they gained was water that accompanied said glycogen (not fat from DNL).Acheson DNL

It was only then, after energy expenditure was induced by record levels and glycogen stores were superdupersaturated, after about 5 days of 1000 grams of carbs/d, did appreciable amounts of dietary carbs start to show up in fat.  Insulin levels tripled.  The participants could barely do it.  They were cramming bagels down their throats while the researchers were begging them to persevere.  In other words, the participants could hardly eat enough carbs per day to get any of those carbs stored as fat.

I’m not saying none of the fat mass in a morbidly obese patient with >50% body fat didn’t come from carbs – no, DNL is constantly occurring, but so is fat oxidation, “turnover,” such that fat balance is maintained.  But most of the body fat in the morbidly obese comes from dietary fat.  Excess carbs are burned in preference to other fuels and stored as glycogen long before they are converted into body fat.  Obesity does not occur due to massive carbohydrate overfeeding; it occurs due the storage of dietary fat, no matter how much or little is consumed.


If you like what I do and want to support it, consider becoming a Patron! Five bucks a month for full access. It’s ad-free and you can cancel if it sucks ????

Affiliate discounts: if you’re still looking for a pair of hot blue blockers, Carbonshade is offering 15% off with the coupon code LAGAKOS and Spectra479 is offering 15% off HERETrueDark is running a pretty big sale HERE. If you have no idea what I’m talking about, read this then this.

20% off some delish stocks and broths from Kettle and Fire HERE.

If you want the benefits of  ‘shrooms but don’t like eating them, Real Mushrooms makes great extracts. 10% off with coupon code LAGAKOS. (I recommend Lion’s Mane for the brain and Reishi for everything else). 


calories proper


Become a Patron!



Be Sociable, Share!
  • jake

    how are you getting bogged down in the insulin idea here? the overall amount of energy into the system, vs. out of the system, will govern adipose storage. i’m not suggesting that we can tightly control the in:out levels, as the experiment shows, but the insulin thing is a side show. insulin would suppress lipolysis to the extent of glucose availability. as soon as the body isn’t meeting its energy needs through glucose, even if you’re eating nothing but carbs, insulin levels would drop to allow the body to use stored adipose. this is why you can lose weight eating nothing but white rice (and it would suck).

    if you don’t accept that, you’re basically arguing that insulin could prevent ultimate starvation, by locking away adipose, even in the context of an energy deficit!

    • George Henderson

      Or insulin could cause penultimate starvation, by locking away adipose, so that an energy deficit could be replaced by eating, far more easily than it could from, and despite, masses of stored energy?

      • “Penultimate starvation”
        George, you’d better trademark that! I think this is what Lustig was getting at in his The Skinny on Obesity series (Ep. 3?

    • andrew

      GOOD POINT!!!!!!

    • Calogero Fiore

      My daughter is a VERY fussy eater all she eats is white rice, steamed veggies (green beans, baby sweetcorn) tomatoes, sometimes rice noodle or wheat noodles, she doesnt like oily foods at all or anything mixed like lasagna or stir fried veggies and she is skinny, my other kids are plump and listen to this they have allergies, egg, wheat but my rice daughter doesnt get colds or have allergies, baffles me cause Im always nagging her to eat meat, dairy, salads and oil but she feels sick if she even tries but she seems healthier than all of us.

  • tess

    Bill, I’m surprised at you: concluding that “carbs can’t be converted to fat” by this one study with obesity-resistant individuals and over feeding…. Talk about over-simplification! Just because Taubes doesn’t have the right answer at his fingertips doesn’t mean that Phinney and Volek haven’t blasted this thing out of the water!

    • Hi Tess,
      Thanks for the comment, but that wasn’t my conclusion at all! Carbs absolutely CAN be converted into fat, even in “obesity-resistant” people.

    • Hi again Tess,

      I apologize for not fully, or at least more directly, addressing your point. The fatty acid composition of adipose tissue largely reflects dietary fat intake – someone who eats a lot of olive oil will have a lot of oleate in their adipose, someone who eats a lot of soybean oil will have a lot of linoleate, etc.

  • donny

    I look at it as glucose metabolism displacing fat metabolism, rather than locking fat away, so you can’t burn it. I always liked Pennington’s balloon thingy–where fat cells increase in size until compromised free fatty acid release is corrected for–but then someone like carbsane will show up and point out that free fatty acids are elevated in the obese–why don’t they get to the point where free fatty acids equal those in the lean, why don’t they reach homeostasis at that point? Elevated free fatty acids might be a sign of of compromised fatty acid metabolism–if you were low in some enzyme or other, increase substrate might be necessary to achieve a certain level of fatty acid oxidation.

    One thing I find interesting in this study was the effect on thyroid hormone–it goes up with the ultra-high-carb feeding. Maybe we could go woohoo! at this–or maybe increased thyroid is necessary for damage control when eating all that sugar.

    • @ Donny: The addition of significant amounts of carbohydrate to diets leads to a significant increase in serum T3 and a decrease in rT3 levels. These alterations in hormone metabolism are associated with enhanced thyroid hormone production. However, more thyroid
      does not equal ‘better’. It’s generally observed that hypermetabolic states in animals lead to shorter lifespans. Low T3, lower metabolism = potentially greater longevity. As our ancestors subsisted on a very low-carbohydrate diet throughout human evolution – a diet that is associated
      with significantly lower T3 levels – a mildly hypothyroidic state might actually be physiologically normal, evolutionarily appropriate, and optimal for longevity.

      • 1000 grams carbs/d, 50% calorie surplus, total energy expenditure increased by 35%… this can’t possibly end well.

        • Just reading “total energy expenditure increased by 35%” makes me feel
          uncomfortable. I wonder what increase in energy expenditure it takes to
          cause fatal hyperthermia (e.g. 2, 4, Dinitrophenol o/d)?

  • @ Donny & Anna,
    perhaps one goal of the T3 surge is to [further] activate BAT. This degree of glucotoxicity is going to wreak havoc on most cell types, so why not try to localize the damage to a tissue that is metabolically designed to handle it?
    high T3 levels might be there for your protection on HC and not necessary on LC.
    *end speculation*

  • Kindke

    I posted this on carbsane’s blog but the question remains, if the conversion of carbs to fat is not significant, how is it people got fat following the 30BaD diet? Which durianrider advocates as 0% fat.

    • carbs-to-fat DNL 😉
      plus deposition of the very little fat in 30 bananas (30g?).

      IF (and that’s a big “IF”) this is a bona fide 30BaD diet, and *only* 30 bananas per day.

      1. That’s a LOT of kcals (3600?) &
      2. carbs (1000g?) – similar to Acheson.
      And 3. very little protein (30g? much much less than Acheson [~136g]).
      So even just considering these variables, in light of the Acheson study, I’d say it’s possible, especially given the very low protein intake (eg, in Bray’s overfeeding study, the low protein group gained the most fat, discussed here: I’d also bet the 30BaD dieters wouldn’t get a huge boost in energy expenditure… it might even decline on such a weird diet :/

      P.S. I watched that entire video. you owe me big!

      • George Henderson

        Durainrider needs about 6,000 banana calories a day, Don Matesz a bit less, because raw vegan food calories are poorly absorbed (the almond effect again) and perhaps because DR at least seems to be v. (hyper)active.
        So much for veganism saving the world by more efficient conversion of land to energy. If you need twice as much food, those arguments might need some adjustment.

        • Harry

          Animals require land, which could have been used to grow crops, and they also require even more land to grow crops for them to eat, meaning veganism requires about 60 times less land for the same amount of food for humans…

          • George

            I read somewhere that it takes 15 grams of indigestible plant protein to make 1 gram of digestible animal protein.
            Pasture land which raises healthy animals is not crop land and can’t be converted to crop land without environmental destruction, including the CO2 emissions of the equipment used, the provision of artificial or fish-based fertilizers, and destruction of wildlife that shares pasture land with farm animals. As well as protein farm animals raise large amounts of energy as fat, which can replace carbohydrate in metabolism. Calculating this as food, as we should, greatly improves the energy yield from raising animals.
            It is easy to find vegan-approved foods that are far more wasteful of resources than plain meat and fat. Veganism is not about using the planet economically, but about indulging the squeamish or puritanical aspects of one’s character. Economic arguments can be developed to support this, but there doesn’t seem to be much incentive to make these accurate or realistic.

        • AcaciaG

          Do you not realise how much more land it takes to produce a single pound of beef? A vegan stuffing their faces every day still has far less of an impact on the planet than a calorie-restrictive omnivore…

    • Calories in Context

      Anyone who is “passionate” about eating 30 bananas a day probably isn’t worth listening to to begin with.

  • George Henderson
    At the bottom of page 11 there is an interesting discussion of differential DNL rates in hyperinsulinaemia and in NAFLD. We can see that insulin makes a huge difference to proportion of carbs > DNL. X2, X3, X5.
    Also the interesting claim that Oleate is the major hepatic DNL fat. There is definitely a shift towards oleate in liver disease.
    It seems that what is happening is that carbon from carbs is being added to dietary (or ex-adipose) palmitic acid to form oleate.
    So DNL might not always mean DNL. More a kind of post-facto lipogenesis. An add-on. And this adding-on increases when the liver is inflamed or IR in order to tame the lipotoxicity or ROS generating powers of palmitic acid.

    • I can get pages 10 & 12, but not 11 :/

      However, I think I can still address your point: I wouldn’t be surprised if DNL rates varied > 10X in people on different diets, with different levels of metabolic health, etc., etc. My point is not “DNL rates,” but rather more of what Kindke is getting at – ie, carb-to-fat DNL proper *which results in net fat deposition,* like what started happening around day 5 in Acheson’s subjects… as opposed to DNL that is balanced by oxidation.

      From another angle: fatty acids in adipose tissue reflect the fats your eating (eg, PMID 11684525, 17903320, & refs from 12821880, and there’s a really good long-term prisoner feeding study which I can’t seem to find).

      But yeah, I agree: “DNL might not always mean DNL,” and I was referring only to one specific type of DNL.

      • George Henderson

        It’s DNL Bill, but not as we know it.
        You know how better-informed people who still believe in the lipid hypothesis claim that stearate is an OK SaFA because it can be converted to oleate on demand? Well so can palmitate, apparently, by the magic of DNL. So much for the modified lipid hypothesis.

  • Pingback: [BLOCKED BY STBV] tn requin pas cher()

  • That’s not how I read the study. The 35% increase in energy expenditure is, they assume, down to nutrient storage. That is, it costs energy to store energy. Once the glycogen stores are full, then there is DNL, but this is expensive, requiring about 25% of the energy. Therefore the body would rather use the excess carbohydrates, before storing them either as glycogen and then as fat.

    That said, I’m not sure this study is of much use to normal people. The diet is so odd that it is hard to draw conclusions from it. In addition, the study was done on fit young men, one of whom was an athlete.

    For the average Joe, who lives in an environment where the body is being fed a continual calorie excess, the cost of storage as fat is low (as the next super-sized value meal will be following shortly), so the there is no reason not to store it as fat. It’s not like the excess carbs are going to be used in exercise.

    All in all, the study doesn’t tell us much new. Don’t we all know that if you eat too much you will get fat.

  • Pingback: [BLOCKED BY STBV] chaussure nike tn()

  • donny


    “as soon as the body isn’t meeting its energy needs through glucose, even if you’re eating nothing but carbs, insulin levels would drop to allow the body to use stored adipose.”

    This is pretty much the same argument that Peter D. at Hyperlipid used to explain how somebody might lose body fat on the all-potato diet–and he was defending insulin’s role in fat balance in doing so. Maybe this is true even on a mixed-diet when everything’s working right. I don’t think we can start with the assumption that everything’s working right when considering obesity and overweight.

  • Pingback: [BLOCKED BY STBV] louis vuitton store()

  • Pingback: [BLOCKED BY STBV] ralph lauren pas cher()

  • Pingback: [BLOCKED BY STBV] maillot france()

  • Sir Harrington
  • Anon

    Would you say that 450 gram intake every day is approriate for a female who exercises 3-4 days a week and also has a low body fat? Along with 2800 caloires. History of eating restriction of calories in the past. I have been eating this much for quite some time, curious as to your thoughts. Thanks for the article!

    • “low body fat”
      That could be your answer right there. If you have good energy levels & no major health complaints, then sure!

  • Pingback: Test 02 | Intermittent Feasting()

  • sylvie

    Pardon my ignorance, but does this suggest that a HCLF diet is, for a non-diabetic, preferable for achieving fat loss in the short-term than a ketogenic diet? And the implication here is that a metabolically normal person couldn’t get fat by eating a HCLF diet (e.g. tons of pasta with low fat sauces, plain rice, potatoes, bread, etc.)?

    • Hi Silvie,
      Not necessarily – on HCLF, insulin ensures that any dietary fat, even if it is present at very low levels, will be stored. You’d need to actively restrict food intake induce weight loss on this type of diet.

      A metabolically normal person…? can probably get fat on a lot of different diets. But then they might be more likely lose the weight, since they’re “metabolically normal” ?

    • Todd

      I don’t know if it’s preferable, but I’ve been following “spud fit” on YouTube with great interest. This man has lost a lot of weight rather quickly eating just potatoes. It’s essentially a more restrictive form of the McDougall diet. It makes sense that it would work in the virtual absence of dietary fat: hardly any ingested fat to store, and relatively low rate of DNL.

      I decided to experiment with this myself, and indeed I’ve been losing weight at a pretty fair clip. Of course, another factor is the sheer blandness of fat-free starches. They are very low in terms of “food reward.” I find potatoes even more satiating that high protein/high fat foods. That is, it’s just hard to eat enough spuds to achieve caloric balance, even with seasonings on them. If I could add butter or cheese, it would be a different story, both in terms of how much I could consume and weight loss.

      I can see how it might be useful to mix “McDougall days” with VLC days, avoiding eating starches and fats on the same days. But that’s just a guess.

      • Interesting. Not sure if I’d volley back-and-forth between VLC and McDougall — seems your metabolism might never adapt to either, but interesting none the less

  • Thomas Hemming Larsen

    Why not just eat something like 25-30% protein, 30% fat and the remainder in carbs? Even if you were to eat slightly more carbs some random day you wouldn’t store it fat. What is the advantage to a low carb diet given this post?

    • “Even if you were to eat slightly more carbs some random day you wouldn’t store it fat.”

      if “slightly more carbs” wasn’t compensated for by eating less of something else, then you would store a bit of that “30% fat” as fat.

      Most body fat comes from dietary fat, not de novo lipogenesis. Even on very low fat diets, if they’re hypercaloric.

      • Thomas Hemming Larsen

        I meant ‘slightly more’ as in one time per month. It seemed like the subjects were able to ‘supersature’ their glycogen. So if you did overeat (not all out binge) carbs that one day it probably wouldn’t to lead DNL – as I read it.

        My point is also that I don’t fully understand the love for low carb diets. If you overeat on a high fat, that would make it even easier to store fat.

        • “So if you did overeat (not all out binge) carbs that one day it probably wouldn’t to lead DNL”

          this is correct, but you still may gain some fat mass (from dietary fat) because CICO. Regardless of which macro you overeat, dietary fat is the most likely macro to be stored (doesn’t matter if LC or LF diet).

          • Thomas Hemming Larsen

            Thanks Bill. I didn’t want to sound like a pain, I was honestly curious about it.
            I guess some of this goes back to metabolic flexibility and why insulin sensitive people do better with more carbs. Insulin resistant people can’t efficiently switch between carbs and fat which is why they should keep the carbs low. I’m rambling, just my two cents.

  • Pingback: Mata la mosca a cañonazos. Deja los palillos para los nutricionistas a sueldo de la industria | No vuelvo a engordar()

  • Thomas Hemming Larsen

    Quick question Bill. I don’t quite understand the abstract. As I read it, it says that 475g/d of carbs were converted into 150g/d of fat. How can the conversion rate be so low? They say its after the glycogen stores were saturated so I would have thought the conversion was closer to 75-80% (355 – 380g/d of fat).

    • daz

      Don’t forget…generally/roughly 1g carb ~4cals, 1g fat ~9cals

      • Thomas Hemming Larsen

        Yeah, I forgot to take the caloric difference into account. This even more enforces the point of the post that it’s dietary fat being stored. I’m just thinking that for the really obese persons it would be hard to eat enough carbs for them to be turned into fat. They have too eat a fair amount of fat in connection with them to gain that much fat.
        The conversion rate in this case would be 71% then.

        • also, at this point in the study, glucose oxidation was through the roof
          (not sure if this is relevant to your q, though)

          • Thomas Hemming Larsen

            You mean that they were burning so many carbs that less than expected would be stored as fat?

    • daz

      so just as a non real world maths example,
      155g fat equates to 337.5g carbs

  • brenda

    ugh … im so confused… so i will more then likely gain weight on a hclf plant based diet ? what about 80/10/10 ? does it matter what you eat? is losing weight really just about calories in calories out ? i keep hearing people say you don’t need to calorie constrict when on a hclf diet…… 🙁

  • Jojo

    I’m IR and fat (15% over ideal weight, lost about 25lbs since Dec). To control my BG and insulin I’m eating LC. sometimes VLC. I’m eating 3 times a day (6am,12pm,5pm) But it seems my weight and BG are fluctuating (+/- 3lbs). I can’t tell which macro is really giving me the problem. Sometimes I think it is the protein so I up my fat intake but my weight seems to go up as well. I would like to control my BG and at the same time loss fat. Bill, any advise? Based on this blog, it seems I need to restrict dietary fat. Maybe no added fat. But what would be my source of energy? Non-starchy vege and meat. I cannot go crazy on meat because I think it would raise my insulin. Thinking of limiting it to a palm size which is I think about 4-5oz. Eating more than a fist size (1-1.5 cup) of vege is not enjoyable to me. Thank you.

  • kat

    Great article.. So the high carb vegans were right after all

  • Thomas Hemming Larsen

    Its not until now that I think I’ve fully understood what you meant by this post. Now it makes a lot of sense 🙂

    I have a question. I heard Jason Blaha say that if you eat very low fat (even with sufficient carbs) you can lose fat because the body will have to use some fat for some processes. E.g. a guy eats 30g of fat but burns 50g during a day so those 20g have to come from body fat. Can that really be true? I can’t quite get my head around it if you’re eating maintenance calories.

    • oh man, you are so far down this rabbit hole!

      in brief, yes, it can happen. Just make sure some of those 30g are DHA!

      If you’re burning fat and *losing* weight, you’re below maintenance calories [by definition].

      • Thomas Hemming Larsen

        Haha, stop writing such interesting stuff then. This will be my last post about it 🙂

        His point was that you could ‘force’ the body to lose fat at maintenance (being weight stable) by eating very little fat. And that was the thing I was a little puzzled by.
        I fully agree on the DHA 🙂

  • Warriah

    So if you don’t eat enough carbs to start DNL the only calories you have to worry about are protein and fat?

  • abhishek_nitt

    Hi, I am 4 years too late on this article so I don’t know if you’ll read it or not. I wanted to ask that shouldn’t high glycemic carbs (in a constant hyper-caloric diet) lead to LESSER fat gains as compared to low glycemic? The quicker the glycogen stores fill in, the quicker will DNL kick in storing the excess glucose to fat, thus causing about 25-30% loss of calories. When insulin goes down and fat oxidation increases, body will burn some fat. But the net storage (from the carbs) will be with a loss of calories. If the carbs are low GI, then DNL will not kick in (or may be for lesser amounts if the carbs are too much) and the fat storage from excess calories will be all from dietary fat, with no caloric loss. So shouldn’t having high GI carbs contribute to lesser fat gain as supposed to lower GI? Please reply if you read this. Thanks!

    • no, high GI carbs wouldn’t contribute to lesser fat gain than low GI carbs. It’s not a race; quantity & energy balance are the more important drivers.

      • abhishek_nitt

        Thanks for your prompt reply. As I understood, DNL occurs when glycogen stores are full and there is still glucose to process. Also the energy expenditure is increased. Both these processes would cause calories to be lost. So I am considering a case with a regular hypercaloric diet and a 24 hour period. So glycogen stores are near full. Low GI carbs would cause lesser insulin and thus lesser reduction of fat oxidation. So there will be a lower chance of there being excess glucose to fill the glycogen stores. The net result would be storage of the extra dietary fat. But with high GI carbs, there will be a higher insulin response with higher suppression of fat oxidation and increased carb uptake which will lead to filling the glycogen stores. So excess glucose will cause increased energy expenditure and/or DNL, with both processes losing calories. So even though fat oxidation will again increase after insulin drop and by the end of the day excess calories stored as fat, calories would have been lost in either/both DNL and increased energy expenditure. So the net fat gain should be lower. So what exactly am I missing here.

        • There is a difference between glucose and fructose. As we are not rats DNL from glucose just doesnt contribute a significant part to fat gain as Bill put it.

          Now overeating on fructose on the other hand will lead to greater fat gain, due to the fact of how it is processed.

          • abhishek_nitt

            I am not at all talking about fructose. I am talking about high GI carbs. Neither am I talking about use being similar to rats. The whole premise of what I am saying is that glucose doesn’t contribute significantly to fat gain, hence in the particular condition which I described, there should be a lower fat gain when on high GI carbs as compared to low GI. Please read what I have written.

  • Jeff Yates

    Once again another blogger who uses junk studies to mislead people!! The study you give reference to was performed on 3…3…only 3 participants. No less than 30 subjects is considered suitable for a serious study/experiment. This study is worthless unless you can show multiple studies that followed similar criteria and came up with the same results. Learn what methodological rigor means and THEN start basing your opinions on REAL science!!