Monday, April 4

Extra: Glycogen, de novo lipogenesis and carbohydrates

This is an unusual post as I dont like discussing about weight loss or dietary carbs vs. fat superiority in terms of dieting. 

The metabolic advantage is a never ending discussion in the blogosphere. Recently, a more "scientific" debate has started on Peter's blog. While some prefer to argue using scientific references and basic biochemistry concepts, others use their own experience mixed with their opinions. This happens in both who defend insulin's central role in weight loss/gain and in those who state that its calories in-calories out all the way. This last group of people, when trying to dismiss the carbohydrate drives insulin drives fat storage hypothesis, use an old study as their bible. It is not surprising that the one who "popularized" this study was Carb Sane, with her famous post Nutrient fates after absorption. I wrote a quick answer some months ago, which was posted but strangely has disappeared. 

The study which basically dismisses the carbohydrate-insulin obesity hypothesis is one by Eric Jequier, called "Nutrient effects: post-absorptive interactions". I will only focus on the "Metabolic Fate of Dietary Carbohydrate" part. 

The author states:
"Although  the  glycogen  stores  are  normally maintained  within  a  relatively  narrow  range, the capacity for storing large amounts of  dietary CHO by conversion to glycogen is relatively large (Acheson et al. 1982, 1984, 1985, 1988). Fig. 1 shows that a large load of CHO (500 g dextrin-maltose  given as three meals over 5 h) to healthy subjects induces a marked stimulation of CHO oxidation over the 14 h after the first meal (240 g oxidized,  260 g stored; Acheson et al. 1985). Net lipogenesis occurred at a low rate from 5 to 10 h  after  the  first  meal, but  this lipid accumulation was  offset  by  a greater rate  of  lipid  oxidation over the next 4 h. The fat balance calculated over 14 h was negative, indicating  that the large load of  CHO did not induce a gain in  fat."
Lets review the studies used to support his statements. 

1. Glycogen synthesis versus lipogenesis after a 500 gram carbohydrate meal in man

Subjects participating in the study were 6 healthy male volunteers. They tested the metabolic effect of consuming 480g of carbohydrates from bread, jam and fruit juice, with little butter added. The results were as expected, no net lipogenesis occured, just an increase in muscle glycogen capacity. Ok, one big high carbohydrate/low fat meal in healthy subjects does not induce fat gain in the short term (10h), rather increases energy expenditure, glucose storage and oxidation. Nothing new.

2. Glycogen storage capacity and de novo lipogenesis during massive carbohydrate overfeeding in man.
Lets look at the subjects and methods used in the study:
"Three healthy young men, one of whom was a competition swimmer at university level, (21-22 y, 62-72 kg, 174-180 cm, and 11-14% body fat) with no family history of diabetes or obesity and who were not taking any medication, participated in this study."
Ok, we are talking about healthy, young and lean men. Is everyone healthy, young and lean?  I think most of the regular blog readers are people with weight/metabolic issues. The subjects included a competitive swimmer, a fact that in my opinion is very relevant, besides the fact that all of the subjects were lean. This somehow escapes from the people referencing this study. 
"The experiment lasted 14 consecutive days. During the first 3 d the subjects consumed a restricted diet, high in fat and low in carbohydrate, and followed an exercise program. Halfway through this period the subjects were admitted into a respiration chamber in which respiratory exchange measurements were to be continued for 10 d. After 36 h in the chamber the diet was changed to a high-carbohydrate, low-fat diet that was ingested for the following 7 d. During the last 2 d while still in the chamber, the subjects received limited amounts of a high protein diet (protein-sparing modified-fast [PSMF], ‘ı2.5 MJ or 600 kcal) essentially devoid of carbohydrate. The subjects then left the respiration chamber but continued to consume the high-fat, low-carbohydrate diet in restricted amounts for a further 2 d."
I wonder, if a massive carbohydrate intake does not promotes fat storage as the previous study suggested, why using a glycogen depleting protocol before an extended massive carbohydrate intake? Hmm...

In short, the protocol lasted 14 days in which:

3 first days: HFLC + exercise.
7 days: HCLF
2 days: PSMF (LF, LC)
2 last days: HFLC, restricted amounts.

So we have a scenario in which glycogen is depleted by diet plus exercise and then there is a massive high carbohydrate low fat refeed for seven days. Glycogen storage increases until the 4th day, in which glycogen stores had become saturated and de novo lipogenesis starts to increase. This shows that you need 4 days of overfeeding carbohydrates to start getting fat? No. This shows that in LEAN HEALTHY YOUNG SUBJECTS, AFTER A GLYCOGEN DEPLETING PROTOCOL, a relatively well known phenomenon called GLYCOGEN SUPERCOMPENSATION occurs. 

I dont think we can extrapolate these results to everyone nor this study shows that you can eat 500g everyday and not get fat. Specially if you are not lean and/or healthy.

I find odd the use of this study to support the author's hypothesis. Subjects consumed either a high fat, mixed or high carbohydrate diet during 3-6 days preceeding the test. From the abstract (unfortunately I dont have access to the full-text):
"CHO oxidation and conversion to fat was significantly less in the high-fat diet group (222 +/- 5 g) than in the mixed (300 +/- 13 g) or high-CHO diet (331 +/- 7 g) groups, resulting in a greater glycogen storage in the high-fat (278 +/- 6 g) than in the other two groups (197 +/- 11 and 170 +/- 2 g). Net lipogenesis occurred sooner and lasted longer in the high-CHO group, amounting to 0.8 +/- 0.5, 3.4 +/- 0.6, and 9 +/- 1 g of lipid synthesized in the high-fat, mixed, and high-CHO groups, respectively. The thermic effect of the CHO load was 5.2 +/- 0.5% on the high-fat, 6.5 +/- 0.4% on the mixed diet, and 8.6 +/- 0.4% on the high-CHO diet."
This study shows what is an already known fact about high fat diets and carbohydrate loading. HFLC reduce glucose oxidation and increase non-oxidative glucose disposal, ie. glycogen stores. Although net lipogenesis was small (9g on the high carbohydrate diet), measurements were done for 24 hours. While not significant for body composition in the short term, the metabolic effect might be significant, specially considering the effect a high carbohydrate meal has on gene expression (1, 2)*. This study shows the difference between eating a 500g carbohydrate meal after a HFLC and a LFHC diet. 

This study is a follow up on the previous study. Nothing new to add, really. From the study:

"After the load, carbohydrate oxidation increased in each group with a corresponding decrease in fat oxidation. However, in the high carbohydrate (Fig 3C) and mixed (Fig 3B) groups, fat oxidation decreased to zero at 2.5 h and 4 h respectively after which energy expenditure was principally due to carbohydrate oxidation and de novo lipogenesis for a further 9 h and 6.5 h respectively. The negative values of lipid oxidation represent an equivalent amount of carbohydrate energy converted into lipid, ie net de novo lipogenesis. During the evening fat oxidation increased progressively and became the major energy source throughout the night in each group."

The authors concluded:

"The results imply that the more the composition of the everyday diet is rich in carbohydrate, the more full the glycogen stores become, but at the same time more carbohydrate is oxidized to provide energy. When both storage and oxidation become saturated, de novo lipogenesis can occur but not at a rate sufficient to prevent increasing blood glucose and insulin concentrations."

Summing up

It is noteworthy to mention that the author of the review, Eric Jequier participated in all of the studies shown above. The conclusion he gets from his review is the following:

"The  practical consequence  which  results from  the  different fates  of  the  ingested 
nutrients  is  that  body-weight  regulation primarily  depends on fat intake. A chronic 
excess of  fat in the everyday diet is stored and contributes to increase the adipose tissue 
mass. (...) The fat content of the habitual diet could influence the steady-state of weight maintenance. When a high dietary fat intake is chronically ingested, body-weight gain occurs."

Talk about bias. Although he mentions that "The practical implication of these findings is  that the proportion of fat and CHO energy plays an important role in body-weight regulation" he goes on saying that "The ad  lib. consumption of high-CHO-low-fat  diets induces a slow rate of  weight loss, with a spontaneous decrease in energy intake" and "Prevention of obesity should benefit from this recent advance in our knowledge of the post-absorptive effects of nutrients".

I did a quick search on more studies done by Jequier and I found these gems:

I encourage you to check the references used in the section "Metabolic Fate of Dietary Fat". 

Studies can show whatever the author wants to. The only way of knowing and interpreting the real data is reading the data source for yourself. Or reading trustful blogs. But hey, after all, fat is what makes us ugly, fat and ultimately kills us, doesnt it?

* Some further reading on ChREBP is recommended.

ResearchBlogging.orgJéquier E (1995). Nutrient effects: post-absorptive interactions. The Proceedings of the Nutrition Society, 54 (1), 253-65 PMID: 7568258


  1. Wow your scared of carbs,,,

    I can assure you no one will get fat eating 500+ grams on a low fat diet just look at the 80/10/10 retards and how skinny they are.

  2. Hello Cliff,

    You can also eat only twinkies or potatoes and still lose weight. I dont believe weight per se is a good indicator of health. The point I am trying to make with this post is that concluding that excess carbohydrates arent turned to fat when eaten chronically extrapolating the results from study number 2 is wrong. Specially when you consider the obvious bias of the author.

  3. I really like your blog and appreciate the analysis you do here. The discussion on Hyperlipid is pretty interesting..

  4. Great post again Lucas,thanks.I think that what is being overlooked by large amounts of the internet insulin debate is a high metabolism...ala Matt Stone of 180deg health.Matt proposes a high carb diet to do this and I always would try it and falter when my hypoglycemia would become crippling.My skin would get all red and I would look like I was aging.Yet many of the followers of the eat more idea are getting great results with high carb.So I said to myself that if your body can adapt to high carb and then thrive then all is good...and I envy plate loads of pasta,potatoes etc. :(

    But for me its a different story and as many probably found weight gain was high.So recently I tried the leanGains approach and strength went thru the roof but fatloss was halted.I started to think that maybe the protein being high(240gms)was causing a very high insulin surge or insulin resistance from not claim to know what is exactly happening.

    So I revert back to Kwasnieski...though with that ever nagging sad feeling that this just stops working also.This time I say duck it and get away from the "diet means lower calorie" mentality and decide to really up my fat intake(idea started by ) to see if my theory is correct....that fat being very low insulinogenic is best to boost the metabolism by staying away from all the negatives that unbalanced BS bring,for the people that suffer BS abnormalities of course. ;)

    Its been eye opening for past 8 days and I feel great and have been losing "bodyfat" while eating 3000cal compared to 1500cal.SO I wonder the case of BS abnormalities...that you want to ramp up fat calories till weightloss really starts to slow and just eat under this level for sustained weightloss over the long haul.3K today and I am sitting here feeling hot(never do)and am still hungry(rarely am even at 1500).

    I am eating 60gms protein/30gms carbs and gobsmack loads of saturated fat in form of eggs,pork sausages,swiss cheese,butter,corn tortillas,veggies and cream w/splenda as calorie topper for the day.I take in a good quality 50gms whey PWO...ala LeanGains to ensure I remain in positive nitrogen retention after a WO and for the insulin boost it causes which is very anabolic.What do you think of this?

  5. @Tyler: Thanks for your comment.

    @Wolfstricked: If I understand well you are doing good on the OD. The thing I dont get is that you are eating 50g of whey protein PWO? This should be almost your daily protein allowance on the OD.

  6. I actually wanna test seeing if just 60gms protein per day can build muscle.I drink shakes for now due to belief that the increased insulin response from the whey is a boost to muscle building and the extra amino acids are there to ensure rebuild...why waste time in gym when you can take a quick whey shake to augment what the weight lifting is trying to achieve.If you think about it,its amazing that I can build muscle on 60gms protein with a 50gm whey shake every few days compared to the minimum of 1gm per pound bodyweight thrown around.That nets to around 80 per day and most bodybuilders will say its suicide.yet here I am gaining strength every WO.

  7. Frank


    There's rather a lot of review on DNL on pubmed and they all come to the same conclusion, regardless of the authors.

    Beside, if one decide to go the way that "all studies are not methodologically sound, ergo we can't come to any conclusion" or "this researchers is biased, ergo the conclusion are not valid" then it's pretty much useless to argue with scientific references, as good published studies with a strong metho are the exception rather than the norm.

    You do not imply any conclusion in your post as far as I can tell but I get the impression that you believe that the conclusion on DNL are erroneous, whereas everyone in the scientific litterature pretty much agree that DNL is not a significant pathway in humans. Even in the obeses, it seems.

    We could say that you've got a rather big bias regarding high-fat diet, too.

  8. @Wolfstricked:

    Yes, if there are enough calories to support the training sessions there should be no problem building LBM with 1g/kg protein.


    I think its impossible to be bias-free. Opinions arise by differences on interpretation. I dont think carbohydrates by their own promote obesity, nor dietary fat. The problems I have with DNL research is that most of the studies are done in the short term (ie. one meal). If you read study number 2 you will find that as the days of carbohydrate overfeeding pass, DNL starts to increase. I havent looked on any long term study on carbohydrate overfeeding and DNL. That should be interesting.

    Looking at the studies cited, is the evidence enough to say that carbohydrates dont contribute to obesity? The metabolic scenario is equal between a lean-glycogen depleted individual and an obese with glucoregulatory issues? It is important to mention that DNL is not only important to fat mass regulation but also to lipoprotein metabolism and hepatic fatty acid metabolism. Some interesting, updated links:

  9. Frank

    Thanks Lucas for the links. I indeed have some update to do regarding DNL but one thing that pops right away : DNL occur when there is overfeeding only. SO why not just not overfeed in the first place? It's very unlikely that any DNL takes place on a eucaloric or hypo diet, even if it was very high in carbs. But obviously you'll have Taubes et al. telling you that carbs all turn into fat when there is no evidence that this is true.

    I think the way carbs promote fat accumalation is not so much by DNL but more, as the last paper you cited explain, by decreasing fat oxidation. But of course, again, this is a problem only if there's a caloric surplus in the first place...

    You're right about the impossibility of being bias-free. Hence why I don't think we can dismiss Jequier conclusion solely based on the fact that he might be biased toward low-fat diet.

    You're critic of the studies stand correct, but I don't think DNL is the main mechanism by which CHO might promote fat accumulation. And in the end, it's calorie in = calorie out that will determine what really happen to nutriments.

    That's how I interpret things at least.

    Thanks for your time.

  10. malpaz

    hmm, i like this post. So, if one HAS to overfeed(in order to gain BF and restore fertility in my case) would more damage be done overfeeding with carbs, protein or fat? i am more than willing to eat, thats not the problem i just dont know how to be going about it with the least long term damage to my liver/pancreas via glycation, spiking my BS, etc.

  11. It's refreshing to see someone take a holistic approach to this question; to see someone able to handle more than one variable at a time. Kudos to you.

  12. Sorry for the late answers!


    I agree with most of your points. DNL is indeed a very interesting topic and with advancing biochemistry laboratory techniques we should be getting some interesting information.


    I would recommend trying to eat mostly whole foods (not the whole grain type BTW). You can easily gain weight on a low carbohydrate/high fat diet.


    Thanks for the kind words.

  13. Anonymous

    based on my personal n=1 experience: being an avid fitness-junkie (hundreds of push-ups/pull-ups, miles of running/sprinting daily) I noticed that the best way to lean out is: eating LC (but preferably VLC or ZC) for 3-4days, and then have a high carb/low fat (+zero animal protein if possible) + random daily fasts - all this in a 1 meal a day, and no junk whatsoever (in the evening) context

  14. I have cited this analysis and linked to it in the following essay