Small LDL: Perfect index of carbohydrate intake

8. December 2009 William Davis (15)
Measuring the number of small LDL particles is the best index of carbohydrate intake I know of, better than even blood sugar and triglycerides.

In other words, increase carbohydrate intake and small LDL particles increase. Decrease carbohydrates and small LDL particles decrease.

Why?

Carbohydrates increase small LDL via a multistep process:

First step: Increased fatty acid and apoprotein B production in the liver, which leads to increased VLDL production. (Apoprotein B is the principal protein of VLDL and LDL)

Second step: Greater VLDL availability causes triglyceride-rich VLDL to interact with other particles, namely LDL and HDL, enriching them in triglycerides (via the action of cholesteryl-ester transfer protein, or CETP). Much VLDL is converted to LDL.

Third step: Triglyceride-rich LDL is "remodeled" by enzymes like hepatic lipase, which create small LDL.


Carbohydrates, especially if they contain fructose, also prolong the period of time that triglyceride-rich VLDL particles persist in the blood, allowing more time for VLDL to interact with LDL.

Many people are confused by this. "You mean to tell me that reducing carbohydrates reduces LDL cholesterol?" Yes, absolutely. While the world talks about cutting saturated fats and taking statin drugs, cutting carbohydrates, especially wheat (the most offensive of all), cornstarch, and sugars, is the real key to dropping LDL.

However, the effect will not be fully evident if you just look at the crude conventional calculated (Friedewald) LDL cholesterol. This is because restricting carbohydrates not only reduces small LDL, it also increases LDL particle size. This make the calculated Friedewald go up, or it blunts its decrease. Conventional calculated LDL will therefore either underestimate or even conceal the real LDL-reducing effect.

The reduction in LDL is readily apparent if you look at the superior measures, LDL particle number (by NMR) or apoprotein B. Dramatic reductions will be apparent with a reduction in carbohydrates.

Small LDL therefore serves as a sensitive index of carbohydrate intake, one that responds literally within hours of a change in food choices. Anyone following the crude Friedewald calculated LDL will likely not see this. This includes the thousands of clinical studies that rely on this unreliable measure and come to the conclusion that a low-fat diet reduces LDL cholesterol.
Tags :

Comments (15) -

  • nitrile exam gloves

    12/8/2009 3:58:05 AM |

    Thanks for the knowledge sharing...it helps to be healthy.

  • x.ds

    12/8/2009 12:02:31 PM |

    Here is a link showing the atherogenicity in mice of different saturated fats in diets with 1% cholesterol. Look at page 1416 of the free full report that can be downloaded here:

    http://www.ncbi.nlm.nih.gov/pubmed/8409772

    On the opposite you can see bread not being atherogenic in baboons here:

    http://www.ajcn.org/cgi/content/abstract/33/8/1869

    By the way you can notice the effect of 0.1% cholesterol at the end of the article = 1 gm cholesterol per kg of food = 5 egg yolks.

    Does it look a lot ? "There is evidence from animal experiments showing that if atherogenic dietary factors are reduced to levels comparable to man's intake, the same vessel changes occur as with higher levels, but more slowly."

    Download the free full report here:

    http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1938976/

    In this report you can also see butter to be highly atherogenic to swine without additional cholesterol to their diet.

  • Bobber

    12/8/2009 5:06:03 PM |

    What about rice?  Does it also increase small LDL particles?

  • Nigel Kinbrum BSc(Hons)Eng

    12/8/2009 6:54:52 PM |

    @x.ds: Humans aren't C57BL/6J mice (susceptible to diet-induced fatty streak lesions), baboons or pigs.

  • Future Primitive

    12/8/2009 7:34:20 PM |

    @x.ds

    The strain of mice in the first study are predisposed to diet-induced obesity, type 2 diabetes, and atherosclerosis.

    Likewise, here's an interesting quote from the last paper based on a porcine model: "Whenever there are three
    animals from the same litter, they are divided equally among the three groups (ie, "control", "butter", & "egg yolk" groups)  In the present experiment, this occurred once. These three pigs had the most aortic atherosclerosis in their respective groups. The control pig with the most atherosclerosis was the brother of the pigs with the most disease fed egg and butter."  

    The group of pigs with the highest rate of atherosclerosis was the egg yolk group - yet at most we see a vanishingly positive relationship, if any, in a large number of human epidemiological studies of moderate egg consumption and heart disease (too many to list here - many are recent and easy to locate, though).  Do we even have grounds to formulate a hypothesis of egg induced atherogenesis based on human observational studies? I don't know, really - though a casual glance suggests, "no".

    Looking forward to reading the other study you pointed us to when I get the time.

  • Anonymous

    12/8/2009 7:42:17 PM |

    x.ds:

    Often times these animal studies don't translate well when applied to humans.

    Also, like Dr. Davis pointed out in his blog, most (LDL) cholesterol lowering research doesn't use advanced lipoprotein testing like NMR so the data is misleading to say the least.

  • Anonymous

    12/8/2009 9:16:54 PM |

    x.ds:

    It would be nice for a change to see experiments on other than herbivorous (mice) or mostly vegetarian animals (both pigs and baboons on the wild).

  • Dr. William Davis

    12/8/2009 11:48:34 PM |

    Bobber--

    While all carbohydrates increase small LDL, the effect of wheat is the most extravagant.

  • Anon X

    12/9/2009 3:18:19 AM |

    I generally agree with those who deny the universal applicability of experimental results in mice and rats to men. However, I do so with this one caveat; there is one fact I cannot deny: Many women are convinced that most men are rats.

  • LynP

    12/9/2009 4:42:51 AM |

    Doc, does this mean that I ca sorta log my particle size by getting my apoB checked in quarterly labs? I'll never be able to convince primary to do the outright particle size test...yet (working on that). Thnx.

  • Anonymous

    12/9/2009 3:08:55 PM |

    I admit I do not understand VLDL-C. What I would really like to find is a simple range scale. Mine shows up VLDL-C..7  Non HDL ..70
    I do not know if that is good or bad. Is there such a chart showing like,  1 good ... 100 bad ??
    LP(a) shows up by itself and I understand that because there is a range showing bad ..over 30.
    Any help understanding appreciated

  • David

    12/10/2009 2:39:45 AM |

    Dr. Davis,

    I honestly think you are doing God's work and have learned to appreciate the value of heart scans through your site.  However, it still bugs me that so many people in Asia can live off of large quantities of white rice with minimal atherosclerosis.  For instance, this study shows that American Whites have much higher atherosclerosis than Japanese despite the Japanese having much higher LDL-cholesterol, blood pressure, fasting glucose, and smoking rates:

    http://ije.oxfordjournals.org/cgi/content/full/34/1/173


    This other study (below) tries to explain this difference based on the Japanese consumption of fish. However, if you look at the data, those Americans who consume the most fish oil consume about as much fish oil as those Japanese that consume the least fish oil, and yet between these comparable groups in terms of fish consumption, the Japanese still have vastly lower atherosclerosis on heart scans.  And the Japanese American group consumes more fish than the White American group and has more coronary calcium.

    http://content.onlinejacc.org/cgi/content/full/52/6/417


    And here is another study showing much lower CAC in Japanese than in Japanese-Hawaiians even after controlling for a bunch of risk factors including fish intake:

    http://aje.oxfordjournals.org/cgi/content/full/166/11/1280


    Although in this other study looking at only at Americans, the incidence of CAC appears to be similar to that in Japanese - so maybe there was something unusual about the US samples in the other studies?

    http://content.onlinejacc.org/cgi/content/full/49/20/2013

  • Anonymous

    12/10/2009 8:23:46 PM |

    Dr. Davis,
    What is your opinion of the LP-Pla2 test for arterial plaque?  If you've used the test, do any elements of your program reduce levels of this enzyme?

  • buy jeans

    11/3/2010 9:44:57 PM |

    However, the effect will not be fully evident if you just look at the crude conventional calculated (Friedewald) LDL cholesterol. This is because restricting carbohydrates not only reduces small LDL, it also increases LDL particle size. This make the calculated Friedewald go up, or it blunts its decrease. Conventional calculated LDL will therefore either underestimate or even conceal the real LDL-reducing effect.

Add comment

Loading
Lipoprotein(a): Surprising Poll Results

Lipoprotein(a): Surprising Poll Results

6. May 2008 William Davis (5)
No doubt, our little informal poll asking readers whether they have lipoprotein(a), is skewed towards people inclined to respond because they have this genetic trait.

Nonetheless, the response is telling. Of 82 respondents:

--40 (48%) said they did have Lp(a)

--16 (19%) said that they did not have Lp(a)

--26 (31%) said that they did not know whether or not they had Lp(a)


Though admittedly an informal analysis, I'd draw several conclusions from this simple "experiment".

One, while the proportion of people responding that they have Lp(a) may not be accurate, it is a prevalent genetic risk factor that, according to formal studies, is present in 17% of people with coronary or vascular disease, 11% of the broader population. This number may be even higher if the newer particle number assays (measurements) are used (with results expressed in nmol/L), since an occasional person with a "normal" Lp(a) in mg/dl (weight-based) will prove to have increased Lp(a) by nmol/L (particle number-based). (The reason for this phenomenon is not clear. It may be consequent to variation in apo(a) size, with larger apo(a) varieties of Lp(a) occasionally escaping detection .) As our little poll shows, plenty of people have Lp(a).

Two, readers of this blog tend to be highly motivated, sophisticated, and knowledgeable about health and heart disease. Yet a substantial portion--31%--did not know whether they have this crucial risk factor. That shouldn't be. The unnecessary difficulty of getting this simple blood test performed has been driven home to me repeatedly when I identify this factor in someone and then suggest that their grown children and parents, each of whom have a 50% chance of having Lp(a), be tested. It's not uncommon for a 35-year old son, for instance, to say that his doctor refused, claiming it is an unproven risk marker, or to simply say that he/she doesn't know what it is.

No doubt, just knowing whether you have Lp(a) or not is not the end of the story. Reducing Lp(a) and its associated co-factors is no easy matter. With several hundred patients in my practice with Lp(a), it occupies much of my time and energy. Sometimes it leads to enormous successes , but it can also pose a real challenge.

There should no longer be any doubt that Lp(a) is associated with significantly increased risk of cardiovascular disease. This has been demonstrated conclusively across dozens of studies. Risk from Lp(a) is over and above that posed by other risk factors; it also amplifies the risk posed by other factors, e.g., small LDL, inflammatory phenemena, homocysteine, total LDL, low HDL.

In the world of Lp(a), our two most desperate needs for the future are:

1) Better education of physicians and the public, and

2) More effective treatment options.

Thus, our reasons to form The Lipoprotein(a) Research Foundation. Steps to gain tax-exempt status are being pursued as we speak.

I can't help but wonder whether, like vitamin D, a solution is right beneath our noses. An investment in research to fund the trials to better explore both basic science as well as practical treatment options might yield an answer more readily than we think. Wouldn't that be great?
Tags :

Comments (5) -

  • mike V

    5/6/2008 3:53:00 PM |

    Thanks for your work in achieving these goals.

    I am one of the naieve do not know my Lp(a)score.
    As I have mentioned in the past, I am fortunate to have no detectable plaque by recent CTA.
    What tests do you advocate for your patients in this circumstance?
    (I have long followed preventive nutrition similar to your advice.)
    Is age a factor? I am 72.
    Thanks again.
    mikeV

  • Ross

    5/6/2008 7:33:00 PM |

    Well, I didn't answer the poll because my Lp(a) was 16mg/dL in November and is now 12mg/dL.  So it was borderline and is heading down.

    So, do I "have" Lp(a)?  Yes.  There is Lp(a) in my blood.  But not so much that I'm worried about it.  And I do know what my Lp(a) is, so the "don't know" response isn't right.

    None of the responses seemed to fit me.  So I didn't respond.

  • Anonymous

    5/7/2008 3:17:00 AM |

    Similar for me too.  My lp(a) was 6 mg/dl in the first test, 7 mg/dl in the second and 11 mg/dl in the third.  Not quite sure what to make of this so I answered the poll "don't know."

  • Bad_CRC

    5/7/2008 3:08:00 PM |

    Ross,

    Dr. D has said that Lp(a) is not one of the markers where a normal value is 0.  In the TYP book and online library, he says that a desirable score is <30 mg/dL (again, with the caveat about mass vs. particle size).  Superko's book puts the threshold at 20, and the VAP score sheet puts it at 10.  Mine was 7 by VAP, and I took this to mean that I don't "have" Lp(a).  Sounds like you're in the same boat.  See Dr. D's response to me under "Red flags for lipoprotein(a)."

    I didn't respond to the poll simply because I didn't notice it until it was closed.

    Dr. D, out of curiosity (if you have time to respond), what percent of the population scores zero for Lp(a)?

  • Dr. William Davis

    5/8/2008 2:37:00 AM |

    bad_crc--

    Curiously, a Lp(a) of zero is rare.

    Perhaps this provides some insight, though I'm not sure precisely what.

Add comment

Loading