Handy dandy carb index

There are a number of ways to gauge your dietary carbohydrate exposure and its physiologic consequences.

One of my favorite ways is to do fingerstick blood sugars for a one-hour postprandial glucose. I like this because it provides real-time feedback on the glucose consequences of your last meal. This can pinpoint problem areas in your diet.

Another way is to measure small LDL particles. Because small LDL particles are created through a cascade that begins with carbohydrate consumption, measuring them provides an index of both carbohydrate exposure and sensitivity. Drawback: Getting access to the test.

For many people, the most practical and widely available gauge of carbohydrate intake and sensitivity is your hemoglobin A1c, or HbA1c.

HbA1c reflects the previous 60 to 90 days blood sugar fluctuations, since hemoglobin is irreversibly glycated by blood glucose. (Glycation is also the phenomenon responsible for formation of cataracts from glycation of lens proteins, kidney disease, arthritis from glycation of cartilage proteins, atherosclerosis from LDL glycation and components of the arterial wall, and many other conditions.)

HbA1c of a primitive hunter-gatherer foraging for leaves, roots, berries, and hunting for elk, ibex, wild boar, reptiles, and fish: 4.5% or less.

HbA1c of an average American: 5.2% (In the population I see, however, it is typically 5.6%, with many 6.0% and higher.)

HbA1c of diabetics: 6.5% or greater.

Don't be falsely reassured by not having a HbA1c that meets "official" criteria for diabetes. A HbA1c of 5.8%, for example, means that many of the complications suffered by diabetics--kidney disease, heightened risk for atherosclerosis, osteoarthritis, cataracts--are experienced at nearly the same rate as diabetics.

With our wheat-free, cornstarch-free, sugar-free diet, we have been aiming to reduce HbA1c to 4.8% or less, much as if you spent your days tracking wild boar.

Comments (21) -

  • Anonymous

    3/25/2011 3:19:26 AM |

    Dr. Davis,

    Can someone have a good HbA1c but still have an undesirable amount of small particle LDL? ..Like perhaps someone with FHC that has their LDL particles floating around longer in the bloodstream and hence exposed longer to oxidants.

    Thank you.

    John M.

  • Tyler

    3/25/2011 3:51:56 AM |

    I love your blog but I have to clarify on this point. Check out the post by chris kresser: http://chriskresser.com/blog/why-hemoglobin-a1c-is-not-a-reliable-marker/

    a1c is not reliable for many people because of the variation in RBC life length. healthy people's red blood cells may live as over 4 months whereas diabetic's live only as 60 days. This results in vast discrepancies.

    For example my fasting BG averages 77 and postprandial peak is 85-90, but my hemoglobin A1c is 5.7

    This doesn't make sense unless you account for differences in RBC lifetime.

  • Kris @ Health Blog

    3/25/2011 11:45:32 AM |

    I'm wondering what your opinion is of glycation and aging.

    I've been reading that a major part of the aging process might be caused by glycation of proteins in the body, mostly caused by elevated blood sugar.

    Do you believe that practically, one could expect a longer life expectancy to correlate with lower blood sugar levels?

  • Larry

    3/25/2011 12:09:46 PM |

    The other day on the tv show, "The Doctors", they profiled a young woman concerned about her FBG.
    She said that Diabetes ran in her family.
    They did a bloodtest and announced that her FBG was 111.
    The scary part was when they told her that reading was okay.
    With that FBG, one can assume that everytime she eats, her post-prandial FBG is heading into dangerous territory.
    But they told her not to worry.
    She was right about her concern...as Diabetes will continue to run in her family.
    Especially with that advice.

  • Jonathan

    3/25/2011 2:44:03 PM |

    I found Walmart carries a Bayer at home A1c test kit that gives results in 5 minutes.  It came with two test cartridges so I was able to take one when I started lowcarb and another one 4 months later to see how much it came down.  (I came down from 8.3 to 5.2 in 4 months)

  • revelo

    3/25/2011 4:45:55 PM |

    What is HbA1c for those long-lived okinawans with their rice-based diet, or those long-lived cretans with their wheat-based diet?

    Wouldn't a lean healthy body (especially if there is occasional fasting) eventually clean up glycated and otherwise damaged proteins?

  • Might-o'chondri-AL

    3/25/2011 6:22:18 PM |

    Glycation picks on the amino acid valine "wing" on the molecule of haemoglobin's B-chain portion. Aldehydes, both glucose aldehydes and non-glucose ones can become bound to that valine.

    This can occur several ways. Glucose oxidation yields a byproduct, called gly-oxal; this is what most people monitor. In the glyco-lytic pathway called Embden-Meyerhof triose-phosphate drives gly-oxal into the molecule methyl-glyoxal (MG).

    Type 1 diabetics have circulating methyl-glyoxal (MG) levels that are +/- 6 times greater normal. MG is a glycation end product.

    Tyler's comment links to a discussion of fructosamine monitoring. This is from a non-enzyme driven reaction, called Amadori, where fructo-selysine and the fructos-amine 3 kinase cascade generates 3 De-oxy-glucos-ane (3DG); another glycation end product.

    Enzymatic glycation occurs in pathological states. Macrophage activity spins off  the enzyme myelo-peroxidase; this generates hypo-chlorite. Hypo-chlorite pulls in the amino acid serine and then together they cause the formation of certain advanced glycation end-products; namely glyco-aldehyde and glycer-aldehyde.

    Yet another non-enzyme chain of events can generate advanced glycation end products. This is when the molecule per-oxy-nitrite (ONOO-)gets stalled inside the cell and it induces the formation of gly-oxal/gluco-sone/aldehyde molecules that can contribute to glycation.

    ONOO- normally is part of healthy cell signaling. When a metabolic processes is under sustained "stress" it (ONOO-) can't shift the cell function over to what it (the cell) needs to do (in order to adapt and cope). Instead of briefly signalling, signing off and going away ONOO-
    lingers in the cell; a situation that may also be related to ageing.

  • Anonymous

    3/25/2011 7:10:22 PM |

    I wonder if Dr. Davis can comment on situations where carb intake is reasonable and the patient has a decent HBA1c, yet still has higher than normal triglycerides and small LDL?

    My own HBA1c has been in the 4.5-4.6 range, yet my trigs hover around 140-150, and I still have more small LDL than I'd like.

    If restricting carbs doesn't work, D levels normalized, etc. what else could be the cause of higher than optimal triglycerides?

    I know people with HBA1cs in the 5.4+ range, eat many more carbs than I do, yet still have lower trig numbers.

  • Might-o'chondri-AL

    3/25/2011 8:58:10 PM |

    Hi Revelo,
    Vitis vinifera leaf inhibits advanced glycation end product (AGE) formation. That is what many cultures, like Crete, eat wrapped around their cereal grain; we call it Grape Leaves in English (ex: stuffed grape leaves, a.k.a. Dolma in Greek).

    Japan researchers (2009?) took 1 kilogram of dried grape leaves in 20 liters of water and stirred it for 3 hours at 80*Celcius. They administered the decoction in various dosages and found it can reduce the AGE of 3DG (3 de-oxy-gluco-sone) and also a marker of AGE in kidney disease, pentosidine, down to 1/5th the level from that study's AGE control levels.

    The same study experimented with Anthemis nobilis using the same extraction technique detailed above. They propose the active ingredient responsible for the AGE inhibition is the compound called chamaemoliside.

    Chamomile is the name of this plant in English; I suspect it is drunk as a tea in Crete. In the range of AGE inhibitors that they tested Chamomile was better acting than any other; grape leaves efficacy came in second.

    Plants studied that inhibit AGE forming, in no particular order of effectiveness may interest you. These are: Crataegus oxyacantha (English = Hawthorn berry), Houttuynia cordata (English = Chameleon plant) and Astragalus membranaceous (English = Astragalus). Chameleon plant is a regular condiment used in Vietnamese and some south-east asian food; it smells kind of "fishy".

  • revelo

    3/25/2011 9:02:19 PM |

    According to Steven Gundry MD, it is MEAT which is the primary cause of AGE's. (He doesn't cite any references for this in his "Diet Evolution" book.) He recommends Atkin's style low-carb/high-protein to lose weight, then low-fat (15% of calories from fat) as the maintenance diet. He is not too keen on grains, tubers or fruit, but rather emphasizes green vegetables.

  • Tyler

    3/25/2011 9:50:41 PM |

    Thanks for the nice explanations Might-o'chondri-AL

  • Might-o'chondri-AL

    3/25/2011 10:46:23 PM |

    Diabetic nephro-pathy (ie: kidney complication), and kidney disease have elevated AGE. These are monitored as pento-sidine, gly-oxal, methyl-gly-oxal and 3 de-oxy-gluco-sane; which the body tries to excrete as carbonyly compounds.

    Carbonyl compounds are hard to get through the kidney filters and cause an increase in uric uremia, which can be toxic. Too many carbonyls can cause, the so called, "carbonyl stress" of diabetic nephro-pathy.

    Diabetic patients' kidneys eventually can't excrete enough sodium (Na); and that contributes to the high blood pressure (hyper-tension) diabetics tend to suffer from.

    Ketones merit mentioning too. One of the markers for AGE in the kidneys is N-carb-oxy-ethl-lysine; which may (or may not) be a side effect of ketones. Type 1 diabetics do show elevated ketone levels incidently.

    I am not able to offer any perspective on ketogenic diets and AGE however. However, vitamin C is known to decrease ketone bodies. (In the previous post, "Battery acid ...", more
    diabetic responses to vitamin C appears among the comments.)

  • Anonymous

    3/26/2011 4:46:17 AM |

    I've been eating low-carb (basically paleo) for the last 4-5 mo and just got my lipid panel results.  They sky-rocketed.

    Cholesterol 300
    Triglyceride 150  
    HDL          33
    LDL             237

    Every number got worse.  The part that really sucks, is that the diet makes me feel great and nearly all my body fat is gone.  I'm 37, 5'11, 180 lbs and probably about 9% body fat.  Now I'm wondering what kind of trade-off I'm making.  Any thoughts, doc?

  • Peter

    3/26/2011 12:55:22 PM |

    Testing different foods one hour after meals, it seems like a good rule of thumb for me is that each ounce of carbs raises my blood sugar about 10 mg,and that the kind of carb doesn't matter nearly as much as the quantity.

  • Might-o'chondri-AL

    3/26/2011 6:31:09 PM |

    Paradoxical low carb yet relatively high HbA1c & higher carb but relatively lower HbA1c is reported by Annon. Doc assuredly deals with cases like these and has to resolve their enigma one by one.  

    The gene HFE (human hemochromatosis protein, nicknamed High Fe  where iron = Fe)can have a variation (reference code = HFE rs1800562). This variation is seen in +/- 5% of Caucasians, but is not found in East Asian nor African genes.

    More hemoglobin is in circulation for those having this HFE genetic variation. In this case, the same amount of blood sugar that can contribute to glycation of hemoglobin has more hemoglobin surfaces to glycate. Think of it as the glycation has to spread itself thin; the dilution of it's effect makes the % of Hb1Ac less (ie: lower Hb1Ac % measured in the blood sample).

    On the other hand, genetic variation rs855791 of the gene TMPRSS6 (trans-membrane protease, serine 6)is implicated in anemia. In these individuals Hb1Ac readings range higher; there is less hemoglobin relative to the glycation potential in their blood stream. Think of it as the relatively low proportion of hemoglobin which has to bear all the glycation burden
    (ie: Hb1Ac % is higher in their blood sample).

    Anemic (hemolytic) tendency is also driven by variation of gene HK1 (hexo-kinase 1). This enzyme modulates how glucose inside the cell goes through  it's processing pathways.

    This gene (HK1) codes for the unique iso-form of erythrocytes; erythrocyte configuration can figure in to low hemoglobin. In other words it is also a factor in high Hb1Ac readings; glycation potential in the blood over burdens the limited amount of hemoglobin around.

  • Dr. William Davis

    3/26/2011 6:34:23 PM |

    In response to several questions about the potential disconnect between small LDL/triglycerides and HbA1c: Yes, there are people in which one measure is more resistant. It varies based on the mix of underlying genetic predispositions, so it's hard to generalize.


    Great discussion. Thanks, as always. You bring an incredibly sophisticated perspective!

  • Dr. William Davis

    3/26/2011 6:35:48 PM |


    Spectacular! And within an unusually brief timeline for HbA1c.


    Might-o'chondri-AL is referring to endogenous glycation. You are citing a discussion about exogenous glycation, two separate phenomena.

  • Might-o'chondri-AL

    3/27/2011 1:31:59 AM |

    Might Jenny's observation and Nigel's study reference be reconciled somewhat ? I'll tag on my disclaimer of being unqualified to judge low carb or specific diets; since I've never struggled with weight or diabetes, and am not a doctor.

    The study Nigel linked was done with all Kuwaiti subjects. In that country co-sanguinity in marriage is practised by +/- 54.3 % of Kuwaitis. And 1 in 5 are reported to be diabetic.

    The data is very admirable; my suggestion is that the data trend may not exactly transfer to a modern Caucasian population; which is essentially interbred from migration and war (rape). This may be why Jenny sees a +/- 6 month plateau among her respondents and the co-sanguine Kuwaitis saw changes continue for a year +.

    Genetic poly-morphisms influence fasting glucose (GCK, G6PC2 and MTNR1B), are implicated in Hb1Ac, triglyceride levels, HDL levels & so on. That said, I personally would try the low carb approach if I was diabetic.

  • Might-o'chondri-AL

    3/27/2011 1:32:47 AM |

    oops posted this in wrong thread

  • Anonymous

    3/27/2011 3:12:44 AM |

    Re: Anonymous with Cholesterol 300,  Triglycerides 150,  HDL 33 ...

    Suggest you try a technique many dieabetics find helpful to understand food consumption influence on their blood sugar profile,"eating to your meter".
    For a few days, record your blood sugar level immediately before eating a "normal" meal, and then after the meal get 1-hour and 2-hour post-meal blood sugar readings. Separate meals by at least 4 hours. Concentrate on monitoring your main meals and ignore snacking for the first go around. Better however, if you can actually avoid all snaking during period of the testing. Also you will want to add to your journal the foods, ammount consumed, and time it was consumed. If post-meal blood sugar values are high, then to determine a pattern folllowing a meal do a series of hourly post-meal readings until you reach 85 mg/dL or so. As a graph, these results should be helpful to you. Expect that the results will be revealing to you with unexpected high blood sugar values even after following a paleo diet. And if so, it does mean that paleo is not for you, only that you need to more discriminating in what and how much you actually consume.

    I would be interested in hearing about your findings. By the way, you did not mention the blood glucose or HbA1c results of your recent lab tests.

    My regards and good luck ... spo

    BTW: practice good technique with the finger sticks. Do a quick but good hand wash using soap and a warm water rinse prior to a stick. Dry hands well. Dont squeeze hard at the site to encourage blood flow. The original stick should be sufficent to raise a drop of blood for the test strip. Using alcohol swabs and changing out lancets is not necessry when only working on youtself. Keep the test strip vial tightly closed other then when removing the current test strip. If you encounter an "extreme" value, retest for confirmation but clean hands again prior to the retest. My experiences regarding unexpected readings seems to usually invovle hand and finger contamination of some form.

    Finally, on Amazon.com I am able to purchase unexpired test strips in 50 strip lots for my old AcuCheK Confort Curve meter for less than $0.16 or so a strip and often with free shipping. You just have to broswe around a bit.

  • Jonathan

    3/30/2011 2:47:32 PM |

    @ Anonymous with 300 TC
    I would say it could possibly be your liver cleaning itself out (it could have been getting fatty).  The higher Trig might be a sign you are getting too many carbs from somewhere (at least till your sugar stores empty some and insulin sensitivity goes back up) but it could be the liver cleaning out as well.  I think HyperLipid posted something about this once.