Equal calories, different effects

A great study was just published in the Journal of the American College of Cardiology:

Metabolic effects of weight loss on a very-low-carbohydrate diet compared with an isocaloric high-carbohydrate diet in abdominally obese subjects.

88 obese adults with metabolic syndrome were placed on either of two diets:

1) A very low-carbohydrate, high-fat diet (VLCHF): 4% calories from carbohydrates (truly low-carb); 35% protein; 61% fat, of which 20% were saturated. In the first 8 weeks, carbohydrate intake was severely limited to <20 grams per day, then <40 grams per day thereafter.

2) A high-carbohydrate, low-fat diet (HCLF): 46% calories from carbohydrates; 24% protein; 30% total fat, of which <8% were saturated.

Both diets were equal in calories (around 1400 calories per day--rather restrictive) and participants were maintained on the program for six months.

At the end of the six month period, participants on the VLCHF diet lost 26.4 lb, those on the HCLF diet 22.2 lbs (though the difference did not reach statistical significance). Thus, both approaches were spectacularly successful at weight loss.

Surprisingly, blood pressure, blood sugar, insulin and insulin sensitivity (a measure called HOMA) were all improved with both diets equally. Thus, these measures seemed to respond more to weight loss and less to the food composition.

Lipids differed between the two diets, however:

Total cholesterol: initial 208.4 mg/dl final 207.7 mg/dl

LDL: initial 125 mg/dl final 123 mg/dl

HDL: initial 55 mg/dl final 64.5 mg/dl

Triglycerides: initial 144 mg/dl final 74 mg/dl

Apoprotein B: initial 98 mg/dl final 96 mg/dl

Total cholesterol: initial 208.4 mg/dl final 187.5 mg/dl

LDL: initial 126 mg/dl final 108 mg/dl

HDL: initial 51 mg/dl final 54.5 mg/dl

Triglycerides: initial 157.6 mg/dl final 111 mg/dl

Apoprotein B: initial 100 mg/dl final 95 mg/dl

Some interesting differences became apparent:
--The VLCHF diet more effectively reduced triglycerides and raised HDL.
--The HCLF diet more effectively reduced total and LDL.
--There was no difference in Apo B (no statistical difference).

The investigators also made the observation that individual responsiveness to the diets differed substantially. They concluded that both diets appeared to exert no adverse effect on any of the parameters measured, both were approximately equally effective in weight loss with slight advantage with the carbohydrate restricted diet, and that lipid effects were indeed somewhat different.

What lessons can we learn from this study? I would propose/extrapolate several:

When calories are severely restricted, the composition of diet may be less important. However, when calories are not so severely restricted, then composition may assume a larger role. When calories are unrestricted, I would propose that the carbohydrate restriction approach may yield larger effects on weight loss and on lipids when compared to a low-fat diet.

The changes in total cholesterol are virtually meaningless. Part of the reason that it didn't drop with the VLCHF diet is that HDL cholesterol increased. In other words, total cholesterol = LDL + HDL + trig/5. A rise in HDL raises total cholesterol.

Despite no change in Apo B, if NMR lipoprotein analysis had been performed (or other assessment of LDL particle size made), then there would almost certainly have seen a dramatic shift from undesirable small LDL to less harmful large LDL particles on the VLCHF diet, less change on the HCLF diet.

The lack of restriction of saturated fat in the VLCHF that failed to yield adverse effects is interesting. It would be conssistent with the re-analysis of saturated fat as not-the-villain-we thought-it-was put forward by people like Gary Taubes (Good Calories, Bad Calories).

In the Track Your Plaque experience, small LDL is among the most important measures of all for coronary plaque reversal and control. Unfortunately, although this study was well designed and does add to the developing scientific exploration of diet, it doesn't add to our insight into small LDL effects. But if I had to make a choice, I'd choose the low-carbohydrate, high-fat approach for overall benefit.

Comments (15) -

  • MAC

    1/9/2008 1:55:00 AM |

    Dr. Davis,
        You have to check out Dr. Eades blog on the same study. You both had different takes but came to the same conclusion. http://www.proteinpower.com/drmike/

  • Dr. Davis

    1/9/2008 3:22:00 AM |

    Hi, MAC--

    Thanks for pointing out Dr. Eades post.

    I've lately come to read his posts regularly, as I have been thoroughly impressed with his insights.

    It's good to know there's some real thinkers out there!

  • rick

    1/9/2008 4:37:00 AM |

    Had the HCLF group enjoyed the same nearly 50% drop in TG as the VLCHF group, their calculated LDL would have dropped to 117 rather than 108.  So part of their comparative advantage is not a benefit at all.

  • Anonymous

    1/9/2008 5:33:00 AM |


    Dr. Davis, what is your viewpoint on saturated fat intake and arterial damage? Although perhaps saturated fat doesn't affect lipid values too negatively (in fact, it seems to raise HDL), maybe it can eventually lead to heart disease by other means -- inflammation/damage to arteries, leading to plaque build-up?

    There is a reference to a study here, which you've probably seen:


    I'm curious on your viewpoints, as we know not all fats are bad, but it seems a little muddy as to if certain saturated fats are bad or not.

  • chcikadeenorth

    1/9/2008 5:37:00 AM |

    Hi, you commented once on my high hdl..68...it happened after I low carbed, high calories( plus 1800 a day) and hi fat but under 20 gr of sat fat a day.My Ldl went up but lipoprotein(a) was still within the norm not for TYP but for lab values,
    I hear nothing but good results  with LC going back to Atkins,Bernsteins, Eades, Westman and you Dr D so keep plodding along. Soon everyone will know what a wheat belly is and rather than plod you'll be galloping writing another book SmileThnx for all.

  • Dr. Davis

    1/9/2008 1:33:00 PM |

    Most of the feeding studies like the one you cited are flawed in that they claim to have isolated the effect of saturated fat on some measure, brachial forearm dilatation, in this instance. Unfortunately, they did no such thing. They did not control well for carbohydrate effects. Gary Taubes would point out that they presumed that carbohydrates are good and therefore all adverse effects must be from the saturated fat component.

    We are planning a thorough review of the issue in future.

  • g

    1/9/2008 4:50:00 PM |

    You know... I'd love to see the CAC scores (or even IMT if possible?) for people on Protein Power?  Has that ever been studied?  Mac, have you ever had an EBT/CT scan? (are you > 40 yr?)

    I think that would be very very COOL data Smile   Thanks for sharing -- I've checked DR. Eades out since you started posting...  His wife does a nice 'tablescape' like Sandra Lee!  
    I loved his post on foie gras!  That's what I tell my elevated liver test patients -- they are making their fatty livers into FOIE GRAS (and good think I'm not Hannibal... but I like chianti *ha ahaaa*).


  • andyj

    1/9/2008 5:53:00 PM |

    While I would dearly love to try this myself, I am still (as always) having a mental problem with the high-fat part.  I am currently trying to fine tune a plan to do a calorie-restricted diet (about 1600 calories) but I'm not sure how low I can get the carbohydrate segment to go  -- certainly not under 10%.   Most of the fat will be from nuts and chocolate (and salmon and sardines) and I will certainly do a NMR after the fact.  The biggest problem is crafting something you can actually live with for an extended period of time, not just a couple of months.  Therein lies the real problem.  Maybe we should just stop eating altogether!  I have plans for a blog entry on just this subject --  what if we ate only when we truly had to?  Some days I'd have no problem dropping down to about 1000 calories a day, but of course then exercise would be out of the question.  Yeah, this plan still needs some tinkering before I attempt it.  

  • MAC

    1/9/2008 8:01:00 PM |

    Dr. Davis,
       New paper on Vitamin D and heart diseases. Paper says they cannot recommend testing for Vit D nor recommend supplementaion for those with a known deficiency.

    To g: I am over 40 and only low carbing recently. Previous to that used vegan diet to lose weight successfully. Have not had a heart scan yet but seriously thinking of it as my father was diagnosed with Type 2 at 65 and had quadruple bypass.

    P.S. Went to doc the other day and we decided to do some blood work since it had been while and unbeknownst to me until I looked at the paper work he had ordered a Vit D 25OH test and I got him to order a lipoprotein analysis for the lipid part. No discussion, he agreed. I think he keeps up on  latest research.

  • Dr. Davis

    1/9/2008 8:25:00 PM |

    Hi, MAC--


  • MAC

    1/9/2008 11:18:00 PM |


    Posted by me on Dr. Eades site:

    "You and Dr. Davis both reviewed the same study in your respective blogs on the same day. Bit of different takes but same conclusion. http://heartscanblog.blogspot.com/

    Hi MAC–

    I’m a reader of Dr. Davis’ blog from time to time. I guess today that great minds thought alike.

  • g

    1/10/2008 4:20:00 AM |

    MAC -- it sounds like you have a biochem background too?  Yes, I agree many great thinkers are coming up with vastly similar conclusions!  I think that the best balance betw being fed and 'fasting' maybe key (didn't u discuss earlier?). Where is that? maybe being mildly ketotic? at 5-10? or 0-20?  for CAD who knows yet?

    Here's an example of industry looking for a single drug ligand/target (a $325 million one)...  The answer has already been discovered.  you've found it, dude!  I'm not sure about the relationship betw protein and plaque and CAD yet...  do you have some insights?


    THANKS!! g

  • MAC

    1/10/2008 3:24:00 PM |

    To g: I think you have me confused with another poster. Sorry don't think that was me. Maybe Peter? Minor in chemistry and lots of science courses but no biochem.

    BTW, the great mind in this case was Dr. Davis, and that was Dr. Eades paying him the compliment.

  • g

    1/10/2008 8:49:00 PM |

    Sorry for the confusion -- so many quality post-ers here!  It's great that you're considering starting on vit D -- it improved insulin sensivity in a small trial 60% (that's more than any drug out there like metformin or Actos).

  • chickadeenorth

    1/11/2008 7:45:00 PM |

    g et al  do you have some reading material about Vit D improving insulin resistance I could take to my doc. I am on 4000 units a day, haven't noticed any difference but it is only about 2 weeks now.