Kurt G. Harris MD

PāNu means paleonutrition. The "paleo" here signifies "old" and not necessarily paleolithic. The PāNu approach to nutrition is grounded on clinical medicine and basic sciences disciplined by knowledge of evolutionary biology and paleoanthropology. The best evidence from multiple disciplines supports eating a pastoral (animal-based) diet rather than a grain-based agricultural one, while avoiding what I call the neolithic agents of disease - wheat, excess fructose and excess linoleic acid.

Support PāNu

PāNu is ad-free, completely independent and has no outside sponsorship. If you value PāNu, now you can support it. Read this for more information.

In addition to buying from the book list, you can also support PāNu by making all of your Amazon purchases for any item through the Amazon Portal below

Amazon Portal

« Intra-subject variability in Serum 25-D | Main | Vitamin D via Insolation – the natural route in the North »

Vitamin D Test Accuracy and Variation

Prompted by reader posts, where there seems to be quite a bit of variation and perhaps inconsistency in 25-D levels in the field, I have done some further digging into laboratory testing. Here are some musings with references listed at the end of the post.

What is the best test?

Good question. Most of the testing where levels have been related to outcomes and health measures has been done with Radioimmunoassay (RIA)- developed for Vitamin D by Hollis, whose name you may have seen on various D related papers I have referenced previously. RIA includes a brand-name test called Diasorin, and there is a related chemiluminescence immunoassay called Liaison. Liaison is what my local laboratory, ACL out of Chicago, uses. One drawback of these immunoassay test is that they lump 25 (OH) D2 and D3 together. If you are following most advice here and on other paleo blogs, you are not taking any plant (D2) vitamin D as you are not a plant. So there will be no D2 metabolites floating around in you to worry about.

Sidebar: Why is D2 (ergocalciferol) available at all? Guess what. If you insist on a prescription from a pharmacy for Vitamin D, not only will you pay a lot, you will get D2 and not D3 (cholecalciferol). This is purely an artifact of government regulation and history. D2 was the original medicinal form of D and remains the only form FDA approved as a drug – so it is the only form that can be prescribed from a pharmacy. If you are severely deficient and need a megadose, be sure your Doctor knows you are not interested in Vitamin D for plants. The D3 you can buy as a supplement rather than a drug is not only cheaper and more appropriate for us animals, it is as much as 3 times more effective metabolically.

I believe the two gold standard tests are high pressure liquid chromatography or HPLC and Liquid chromatography- tandem mass spectrometry or LC-MS/MS. It’s been too long since I did much “liquid science” to give a good explanation of these lab techniques, but I would note by way of endorsement that the Mayo Clinic uses LC-MSMS and the University of Wisconsin uses HPLC. I am sure wikipedia would be a good source of further info if you are interested in technical details. These tests can report D2 and D3 metabolites separately.

Sidebar: When you shoot at a target, precision is how close the group of bullets are together. Accuracy is the average of the bullet holes and how far the center of mass of the group is from the bullseye. A test can be accurate but not precise, or precise but not accurate. Example: your rifle shoots a 1.5’’ group at 100 yards. However, the center-of-mass of 3 shots is 4 inches from the center of the target. Precise but not accurate. Adjust your riflescope, or consistently miss by the same amount.

HPLC and LC-MSMS are precise, with interassay precision of about 8%. This means if they measure the same sample twice it will vary by no more than 8% most of the time. Example: 60 ng/ml, then 65 ng/ml, then 55 ng/ml - true value is 60.

The RIA method (Diasorin) and non-radioactive variant (Liaison) are less precise, but correlate fairly well with the other methods. They are used because they are automated and presumably easier and cheaper to run.

The Competitive Protein Binding Assay (CPBA) in this paper overestimated the D3 species by up to almost 60%. Not reliably accurate, not recommended.

In general, the most important thing is to have a lab that calibrates to a standard. Quest Laboratory was using a gold standard technique, but the calibration was off. (Precision without accuracy) They have a voluntary program now that helps with quality control for D testing known as DEQAS – if your lab uses it, that is a good thing. My impression is that the two gold standard methods, and the diasorin and liaison immunoassays are all good (precise) enough if calibrated to ensure accuracy. Inaccuracy due to poor calibration seems to be the bugbear of D testing.

ZRT labs actually uses LC-MSMS, a gold standard technique. Because they are reconstituting dried blood from a fingerstick, they have to use this more precise method. According to their literature, they calibrate versus a Diasorin reference to ensure accuracy (not precision) versus the clinical standard that was first established. Their literature claims interassay precision of 13%. I suppose this is higher than the 8% often quoted for standard LC-MSMS because they are reconstituting a tiny amount of dried blood. That seems a pretty acceptable decrement in precision. This, plus the knowledge of endorsement by Dr. Cannell’s vitamin D council, makes me think this should be a pretty reliable test. I am going to do some calibration of ZRT’s test soon against my own lab.

Your D level is not what you expected. Is it real?

Let’s say you get a level of 60 ng/ml. How far away from your true level could this be?

Interassay variation – this can be up to about 6 ng/ml as discussed above.

Laboratory variation – inaccuracy due to miscalibration. This apparently can range up to 5 ng/ml or so (not counting huge errors or the CPBA method) – a systematic bias that is usually to the high side according to this paper.

This abstract describes a study following women and hormone replacement therapy. I usually don’t reference anything where I have not read the full paper, but I don’t have access to it. Anyway, this study is cited by these authors who presumably read the whole manuscript, as showing up to 20% variation in D levels in the same individual on different dates. If we take this as accurate, a level of 60 ng/ml one day could be followed by a level of 72 or 48 a month later, with no change in intake.

So with static intake (and consumption) levels, assuming equilibrium, and adding up our sources of error (not statistically kosher but close enough) a measure of 60 ng/dl could represent a biological level of as low as 40 ng/dl or as high as about 75 ng/dl, with nothing being wrong with the test. This is a good argument for aiming for the high side, say a minimum of 60 ng/dl instead of 50 ng/dl, in order to maximize non-classical effects of D. One paper noted that the labs all showed better precision at low levels than high, so someone could measure in the 80s on one date and then in the low 100s with some combination of statistical variation, and intraindividual biological variation accounting for the change.

Quite obviously, levels can also vary due to changes in D intake, whether from the sun or from oral D3. How fast can you change your 25 (OH) D level by changing your D intake?

Not as fast as you might think.

Remember that Vitamin D and 25 (OH) D3 are two different molecules. D is made naturally in the skin from 7 –dehydrocholesterol. We can say that for young light-skinned people this amount can be up to 20,000 iu/day but is at least 10,000 iu/day for most people. The D then must go through the liver to be hydroxylated into the 25- OH form. Remember that this is the only form we measure. We don’t measure the vitamin form that represents what we ingest or that gets made by the skin. Why do we measure the 25 –D only? Several reasons: It seems to be the principal storage form of D, even though the Vitamin form can be stored in fat as well. The active 1,25 OH form (calcitriol) has a static serum concentration that is normal even when we are deficient. Finally, it was what we measured first and we wouldn’t know what it meant to measure Vitamin D itself without repeating decades of research! I just think it is important to understand that the actual levels of the vitamin form of D are opaque to us even if we are measuring the 25-OH form with great accuracy.

The neat thing about the natural solar method (see last post) is that it is basically impossible to get too much of either the vitamin (not measured) or the 25-OH form (measured). This is because the enzyme system in your skin actually runs in reverse and degrades the vitamin form if your serum levels of D get too high. So we can think of dosing with insolation as having a top speed of as much as 20,000 iu/day with a speed governor that cuts the engine if we start to get into trouble.

Sidebar: An entirely theoretical advantage of solar D is that the Vitamin level (which, again we don’t measure) cannot ever spike into levels that are outside of our evolutionary experience, as could be the case with oral D3. I have found no reason to believe that might be harmful or in any way outweigh the benefits of oral supplementation, but it is something to think about if we are to maintain our rigorously critical stance. We do not have proof of any harm of excess levels of the vitamin form, but it is possible, as it has never really been looked for.

This paper and this paper describe what happens if you give a large single dose of 100,000 iu and measure vitamin D levels and 25-D levels over 3 months.

You can see that 25-D levels peak at about 7 days with a maximum increment of about 14 ng/ml.

The pharmacokinetics of the liver enzyme(s) that form 25-D are exponential  at lower levels of serum 25-D and then become zero order at serum levels above 35 ng/ml. The enzyme gets saturated at a value about where parathyroid hormone is maximally suppressed (another reason to regard 35 ng/ml as a bare minimum value for health), and thereafter the level of 25-D is linearly related to the D level.

The important point here is that 100,000 units produced no signs of toxicity, no change in calcium levels and only bumped the serum 25-D levels 14 ng/ml over a 7 day period.

It would therefore be pretty hard to explain a high 25-level as due to the timing of a recent dose of vitamin D of say, 10,000 units or less.

The other take-home point is that it is pretty unlikely that the effects of D toxicity are mediated by the vitamin form, given that serum levels of the vitamin in this experiment peaked at levels  two orders of magnitude higher than basal levels. So the theoretical concern in my last sidebar should remain theoretical.

A final point would be that your body should be very good at smoothing out variation in 25-D levels even with erratic or irregular dosing with no danger. So if you miss a dose one day, go ahead and double up the next. It won’t hurt you.

Vitamin D toxicity is basically unheard of below 300 ng/ml, but leave yourself a margin of safety.

I will add to my supplement if below 60ng/ml and cut back if I am over 120 ng/ml. A level of 120 is not completely arbitrary – lifeguards “in nature” can have such levels. So I’ll consider levels above that unnatural.

A final note on oral dosing. I found the Carlson’s gelcaps get stuck in my throat which is really annoying. I have switched to Carlson liquid vitamin D drops - the vehicle is coconut oil but it is pretty much flavorless. One drop is 2000 iu. I put 3 drops in about 5 tablespoons of whole cream, then add my morning coffee.

Much easier than the pills. 



Correlation among 25-Hydroxy-Vitamin D Assays

HPLC Method for 25-Hydroxyvitamin D Measurement: Comparison with Contemporary Assays

Pharmacokinetics of vitamin D toxicity

25-Hydroxylation of vitamin D3: relation to circulating vitamin D3 under various input conditions

Assay Variation Confounds the Diagnosis of Hypovitaminosis D: A Call for Standardization

Pharmacokinetics of a single, large dose of cholecalciferol


Note: The firearm is a Dakota model 10 in 25-06 with Swarovski scope. It's both accurate and precise.


Reader Comments (42)

I have used Labcorp and ZRT labs within three months of each other. One came out 76 the other was 78.

October 29, 2009 | Unregistered CommenterJake

Thank you for this really interesting article. But I'm confused because 18 months ago when I first started searching for answers to my ever changing vitamin D levels, varying from 384 nmol/L (154 ng/dl) to 125 nmol/L (50 ng/dl), with 2,000 IU D3 per day constant, I wrote to several experts including Dr Reinhold Vieth and Dr James Dowd. I began to understand that the Diasorin method was the most accurate which is not your conclusion. The Diasorin method is hardly used in the UK where I live, I was alerted to just one private laboratory in London that uses it. I get tested at an ordinary NHS hospital laboratory and I've no idea what method they use but I understand that it will not be the Diasorin method as that is too expensive, but my 25(OH)D levels vary as ever.

I'm beginning to think that there is far more to vitamin D and 25(OH)D than anyone currently knows.


October 29, 2009 | Unregistered CommenterAnne


Regarding the Carlson gel caps (and any fish caps as well), I find that if I swish them around with a bit of water in my mouth to get them wet, then they go right down.

Of course, you mileage may vary. I don't like the drops because sometimes it takes like 20 seconds between each drop.

October 29, 2009 | Unregistered CommenterRichard Nikoley

I use carlson drops also. Do you put it in cream for better absorption?


October 29, 2009 | Unregistered CommenterMarc Feel Good Eating


I tried the water trick but it was not reliable for me. Also, my drops actually come out too fast! Oh well.


I thought it would be miscible in the fat in the cream but microdroplets float on top when I add the coffee. there is enough cream in there that the solution is not hot enough to chemically alter the vitamin D (I hope).


I do not think Diasorin is inaccurate, I think it is imprecise. For the tests that I "endorsed" , accuracy is mostly a function of calibration. The HPLC and LC-MSMS methods are likely most precise. Why not use ZRT as a check against your local lab and see if it is in the ballpark? You can "drop it in the post" to the US if they will ship it to you. Shouldn't cost too much I would think.

October 29, 2009 | Unregistered CommenterKurt G. Harris MD


I might just use ZRT as a check against the NHS lab yes, good idea.

Regarding the Carlson's gel caps - I use their Solar D soft gels and you can chew them...they taste a bit of lemon and are pretty yummy. But I'm going to have to go to drops soon as I'm about to run out of the gel caps, not Carlson's drops because they're too expensive to ship over to the UK, but I've found another firm called Biotics who do drops of 2,000 IU D3 (Bio-D-Mulsion Forte).


October 29, 2009 | Unregistered CommenterAnne

Hey Kurt,
Coming from more of an orthodox Paleo perspective, what are your thoughts on the insulin spiking properties of dairy? I know Cordain references a study of whole milk vs skim milk (both spiked insulin pretty high) so I think milk would in theory have more or a spike than heavy cream but I wanted to check with you on that. I agree that as long as you're not eating wheat/legumes then you don't have to worry about auto-immune effects of dairy but from an insulin management point of view, maybe it's important? It would be nice to add heavy cream to my coffee as I like doing low carb and higher fat (moderate protein, usually end up around .85 g/lb of bdwt) but it's hard to find some good sources of fat that don't have carbs too without going to heavy on protein (I sometimes get sleepy after eating too many nuts for example). Since going Paleo, I have leaned out some and I'm not sure if its due to watching calorie intake or the removal of grains and dairy. A common trend across Crossfit is to remove dairy if someone is looking to lean out, does that make any sense?

Lastly, if someone has been following the PaNu approach consistently say for a year and we can assume their gut is healed/healthy, do you think a dosing of wheat here and there (every 3 to 4 days) would be consistent enough to break down that gut lining? Sorry for the obscure question but I read that the gut lining regenerates every 3 days and therefore eating wheat once or twice a week may not cause problems (again, assuming a healthy gut to start with). I'm not looking for an excuse to eat wheat regularly, just theorizing. Thanks for your perspective Kurt!

October 29, 2009 | Unregistered CommenterMark

Just chew the caps. They don't taste like anything except gelcap. I'm currently chewing NOW brand.

October 29, 2009 | Unregistered CommenterGrok

Hi Mark, I was just about to ask Dr. Harris a similar question about insulin. Yes it appears to be well accepted that whey stimulates insulin secretion . I had carefully considered this before deciding to include whey protein as the primary protein source of my diet. From what I've read, I figure the benefits of whey outweigh this drawback, for example:

Research so far makes a strong case for (1) whey protein exhibiting anticancer activity.
The authors of this article hypothesize that (2):

Milk evolved under continuous Darwinian selection pressure to nourish mammalian neonates. Evolutionary pressure appears to have led to the elaboration of a complex food that contains proteins, peptides, complex lipids, and oligosaccharides that act as growth factors, toxin-binding factors, antimicrobial peptides, prebiotics, and immune regulatory factors within the mammalian intestine.

Also (3), whey is relatively low in methionine.
Sounds somewhat convincing to me, though I'd wonder how Dr. Harris would evaluate that statement.

I am still not sure whether insulin secretion per se is harmful, or whether dietary carbs are a necessary requirement to develop IR. This still worries me because the diet I follow, while it mostly qualifies as paleo, is probably rather insulinogenic for a < 50g carb diet. I substitute MCT oil (hence the name) for most dietary fat, and while they make ketosis easy, some studies show that they too raise insulin levels. Suppose you combine this plus whey protein into a single meal per day and who knows how much insulin is generated under these strange conditions.

I'll likely spam some more articles on my blog as I find them.

October 29, 2009 | Unregistered CommenterMCT

Hi Mark,

You wrote: "it's hard to find some good sources of fat that don't have carbs too without going to heavy on protein "

I eat a more orthodox Paleo diet - neither grains, legumes nor dairy except a drop of milk in my cuppa (I'm a Brit). I'm diabetic so it's essential that I eat very low carb but I'm a thin diabetic and need to eat a lot of calories so I increase my fats by always cooking my low carb veggies, for example kale, cabbage, asparagus, okra etc in coconut oil. I cook lots in coconut oil or goosefat to increase my fats, frying meat or fish in it too. Of my calories approximately 20% are from protein, 10% from carbs and 70% from fats.

You're worried about milk spiking insulin but I'd be more worried about wheat spiking it in response to the glucose spike....yikes !


October 30, 2009 | Unregistered CommenterAnne

Hi Anne,
Thanks for your response, I love cooking my eggs in coconut oil. After doing some reading yesterday, I'm not even sure if I should be worried about insulin spikes. There are plenty of cultures and tribes that live on a high carb diet yet never develop insulin resistance/metabolic syndrome. So if we look at it from that perspective, are we just to avoid wheat, legumes, industrial oils, and excess sugar (meaning that fruit is ok)? While high fat/low carb is easy to do, is it the only way to achieve long term health? I'm not ready to demonize insulin entirely, just foods that are foreign to our evolution because it is those foods that lay the foundation for metabolic syndrome? If you could chime in Kurt, I'm curious to see your opinion on this too. Thanks!

October 30, 2009 | Unregistered CommenterMark

I dose my boy's with drops, but place them in a tablespoon of cream rather than in a cup of it. As cream is a thick substance that leaves a coating on the sides of a cup, I would be fearful of the D drop getting washed down the sink rather than going down in the gut where it belongs. I would think that the same thing could happen by mixing it with water! By using a spoon, one would not have to worry about the heat from coffee possibly degrading it either!

Great posts Kurt!

October 30, 2009 | Unregistered CommenterJenny Light


Thank you

It's nice to know that normal people can be as anal-retentive as doctors :)
I worried about that too, but I can see the tiny droplets floating in the center of the coffee cup - they are mostly gone after the first couple of swallows. It may work best if you take cod liver oil to add it to a tablespoon of that, too. Even cream is still mostly water, unlike CLO.

October 30, 2009 | Unregistered CommenterKurt G Harris MD

I just put the Carlson drops straight on my tongue, usually right after a meal.

I'm supplementing with 4000IU/day. My test results came back form ZRT labs, and I'm at 64 ng/dl. My doctor is also going to test my levels, it will be interesting to see how the 2 compare.

October 30, 2009 | Unregistered CommenterCavePainter


"Coming from more of an orthodox Paleo perspective, what are your thoughts on the insulin spiking properties of dairy?"

I have heard there is something unique about the ability of dairy to "spike" insulin. does the lactose in milk or the protein in dairy lead to an insulin response substantially greater than equimolar amounts of other proteins or disaccharides relative to their glucose component? I personally am not concerned about this but would like to look for more real scientific evidence before commenting further.

"I know Cordain references a study of whole milk vs skim milk (both spiked insulin pretty high) so I think milk would in theory have more or a spike than heavy cream but I wanted to check with you on that."

cream has miniscule lactose and protein compared to whole milk so I think it's silly to worry about the insulin spiking properties of little bits that it does have. whole milk is again full of lactose.

"I agree that as long as you're not eating wheat/legumes then you don't have to worry about auto-immune effects of dairy but from an insulin management point of view, maybe it's important?'

insulin, like glucose and oxygen is not a poison. we want to minimize the area under the curve not eliminate it from our bodies.

" It would be nice to add heavy cream to my coffee as I like doing low carb and higher fat (moderate protein, usually end up around .85 g/lb of bdwt) but it's hard to find some good sources of fat that don't have carbs too without going to heavy on protein (I sometimes get sleepy after eating too many nuts for example)."

I assure you cream does not have too much protein. eat a few nuts if you like them but they are not particularly healthy - another category of "magic food' in my opinion.

"Since going Paleo, I have leaned out some and I'm not sure if its due to watching calorie intake or the removal of grains and dairy. A common trend across Crossfit is to remove dairy if someone is looking to lean out, does that make any sense?"

If "leaning out" means you think you're still fat and want to lose weight, them eliminating milk makes sense. Otherwise I don't really know what "leaning out" would mean - it sounds like some kind of body building concept - preparation for a contest?

"Lastly, if someone has been following the PaNu approach consistently say for a year and we can assume their gut is healed/healthy, do you think a dosing of wheat here and there (every 3 to 4 days) would be consistent enough to break down that gut lining?"

It would be hard to prove statistically that smoking a few cigarettes a month is harmful. I think I would enjoy a cigarette more than a loaf of bread. Seriously, in folks who have actual celiac disease, a piece of bread every three or four days is enough to keep their disease going full bore. If you have only the innate and not the adaptive response and just a subclinically leaky gut ( like most of us) then who knows?

October 30, 2009 | Unregistered CommenterKurt G Harris MD


I'd like to know why some of you are using so much coconut oil. I hope it's for the flavor and not because you think it's magic. I view coconut oil as the only acceptable plant oil, but not preferable to beef fat from a pastured animal or pastured butter or ghee.

If you have neurodegenerative disease, the MCTs in coconut oil can help you achieve ketosis.

Coconut oil is sort of outside our evolutionary experience. Most humans throughout history would not have eaten anything biochemically similar. Of course you know my views on dairy so that is not a deal breaker.

I think coconut oil is fashionable in the paleocommunity as spillover from the mainstream health food movement - because animals do not have to die to make it.

The PaNu approach is based on the central dogma that it is what we avoid that makes us healthy. There are no magic foods.

October 30, 2009 | Unregistered CommenterKurt G Harris MD

Because of its highly magical properties.

;) Seriously though, why not?

I'm under the impression that food reenactment is not necessary. "Why eliminate a food or practice without solid contemporary evidence of its harm. Why presume the only criterion of health is provenance and history."

What is lost by trading butter for coconut oil?

October 30, 2009 | Unregistered CommenterMCT

Hey Kurt,
Great response. I just read some great posts by Matt Stone at 180Health. He hasn't blogged in a bit but from what I can gather, he supports a high fat and mod/low carb/protein approach. He seems to be rooted in digestion as the root of all problems and it makes a lot of sense. As far as carbs go with him, he suggests white rice, peeled potatoes, or veggies that have thoroughly cooked to break down the fiber. When you look at it from this angle, it's easy to see that carbs aren't evil. If your body can digest them (glucose) then it will know how to process them. If your body can't (fructose, sucrose, artificial sweeteners, lactose, oligosaccharides - beans/legumes, and fiber) then the whole insulin signaling process will be thrown off leading to hypoglycemia. http://180degreehealth.blogspot.com/2008/10/diver-tickle-o-sis.html
He has a lot of good posts but I think this one is quite complete. Let me know if you disagree with the approach. I would be curious to see what you have to say.

While I'm not a bodybuilder, I do have a hard time not thinking like one when it comes to food. I'm really trying to work on this and I think approaching it from a digestion standpoint and not calories or macronutrient breakdown is the way to go.

Thanks for your responses.

October 30, 2009 | Unregistered CommenterMark

Hi Kurt,

I use both coconut oil (virgin, organic) and goose fat for cooking - I use those fats mainly because I enjoy the taste of them and because, like other saturated fats, they can be heated without getting damaged, and also because they're easy to buy even if they are a bit expensive. I would use beef fat too if I could get it, but the only beef I get - organically reared and grassfed - is sold only in steaks here in the UK and there's no fat left over which I could use for other dishes.


October 30, 2009 | Unregistered CommenterAnne

Hi Mark,

Even though you might not be diabetic, it would be a good idea for you to buy a blood glucose meter - they're ever so cheap though the testing strips aren't - but still get one and do a few tests 30 to 40 minutes after after you've eaten the white rice or peeled potatoes. You'll most likely see your blood glucose shoot up high and then plummet down low later- it is this sea saw effect which is so harmful and can lead to diabetes.


October 30, 2009 | Unregistered CommenterAnne

Hi Anne,
Thank you for your post. While I'm not a doctor, I don't think blood sugar spike's are what cause diabetes. I think it comes down to digestion. I could be way off but this approach makes much more sense to me. I'm hoping Kurt can affirm or set me straight here.

October 30, 2009 | Unregistered CommenterMark


"Coconut oil is sort of outside our evolutionary experience. Most humans throughout history would not have eaten anything biochemically similar. Of course you know my views on dairy so that is not a deal breaker."

No need to quote me back to myself if you read this - I did not advocate re-enactment. The fact remains it is more outside our experience than fat from ruminants, no?

Coconut oil is more expensive than butter and way more expensive than beef fats.

What is lost by not eating coconut oil other than the coconut flavor? I personally like the flavor but not in my eggs.

I am not saying it is unhealthy, I am asking if it has positive health benefits other than being better than PUFA-rich plant oils. Is it better than butter or grass fed beef fat?

I seriously want to know why it is so popular. When people have a handle like "MCT" it suggests to me there is something special going on.

Anne and Mark

Glucose and insulin "spikes" per se do not cause diabetes. It's the area under the curve.

Chronically elevated levels of glucose, fructose damaging insulin sensitivity, WGA binding insulin receptors, visceral fat secreting hormones that cause insulin resistance - chronically elevated glucose levels

The pancreas deals with chronically elevated glucose (not just spikes) and decreased insulin sensitivity by making as much insulin as it can until it can't make anymore - then you have Type II diabetes.

October 30, 2009 | Unregistered CommenterKurt G Harris MD

I must clarify - glucose and insulin spikes don't cause diabetes, I never said they did, but I said that they can lead to it. This is because they do put a great deal of strain on the pancreas with the result that beta cells get burned out and diabetes results.

I have type 2 diabetes but it was not caused by chronically elevated glucose or decreased insulin sensitivity. I am not insulin resistant. I have never been overweight and have never eaten junk or processed foods. My pancreas doesn't produce enough insulin, but it is not like a type 1 diabetic who will not produce any insulin and so I am classed as a type 2 diabetic !


October 30, 2009 | Unregistered CommenterAnne

"I seriously want to know why it is so popular. When people have a handle like "MCT" it suggests to me there is something special going on."

commercial interest.
seems to be a regular character trait, always looking for that special elixir, supplement or esoteric ingredient that will foster 'wonder' health.
bah - nothing against coconut oil, and no doubt many folk have eased into saturated fats with this, but one can create a special need for anything if one tries.
After years of trial and error, no grains, no starchy vegetables, (there are many more vegies still to choose from anyway) plenty of animal rich fats (some fiddling around with ratios between 3's and 6's to bring 3's into a more acceptable ratio, and no doubt many folk who have eaten unsuitable foods for lengthy periods would probably benefit from eating more fish than less), and the occasional piece of fruit. (if at all)
Heeding Dr PaNu's dietary recommendations, is as good a framework as any. It is so easy to get lost in a maze of dietary miasma which empties your pockets, fuels unnecessary thoughts and does little in the way of sustaining and strengthening the organism.
just my two pennies worth.

October 30, 2009 | Unregistered Commenterj

oh yeah, and meats (sorry, left out)

October 30, 2009 | Unregistered Commenterj

Dr. Harris,

Not going to speak for others, but my reason for substituting medium-chain triglycerides for other fats is out of curiosity. MCTs are metabolized differently than LCTs. The coconut oil is only for taste; I mostly use pure C8, C10 MCT oil. How will I respond if nearly all my fats are not LCTs? is such a diet even possible to maintain? MCTs can be insulinogenic, yet potentiate ketosis at the same time? There are many studies on PubMed that exist specifically to compare MCTs to LCTs. It's my own n=1 experiment.

I don't have a difficult time at all controlling my intake of MCTs, as they seem to have a self-limiting effect. I don't know of anyone yet who can overfeed on extra virgin coconut oil, for instance.

Point taken, that coconut oil is indeed outside of the general population's evolutionary experience compared to animal fat. Still, how can coconut oil be any more discordant to evolution than dairy fat, and is there a clear reason why either should have preference over the other? What matters is creating the correct metabolic state, right?

yes the letters "MCT" do have a meaning, anyone can click behind the name to know more.

October 30, 2009 | Unregistered CommenterMCT

sounds like someone's past the chance of trying anything else, looking forward to you sharing your experiences and learning from.

October 30, 2009 | Unregistered CommenterMCT


If by spike you mean a short transient elevation, then I have to disagree. AUC - area under the curve is what counts - think integral calculus if you have math training. Eating a small potato once a day if you have normal insulin sensitivity and pancreatic function will not predispose you to diabetes - farting or intestinal diysbiosis maybe - but not DM type II. It will definitely spike your serum BG for an hour or so, though.

I am going to risk offending you and disagree with either your own diagnosis or your terminology.

If your pancreas does not produce enough insulin and you never were insulin resistant and never had chronically high insulin levels, then you do not have type II. Pancreatic burnout and high serum BG in the face of high insulin levels is the definition of type II DM.

Also, you can indeed still be producing insulin and be a type I diabetic. Many Type I diabetics still produce insulin and on VLC no longer need to inject it as they still make enough to handle minimal carbs and protien intake on VLC. See Dr. Bernstein's books.

From your description, you way actually have LADA - latent autoimmune diabetes

See this link http://www.diabetesmonitor.com/lada.htm

As this is actually late onset type I, total elimination of gluten grains offers some prospect of halting the autoimmune reaction with time.

Optimized vitamin D status may help as well secondary to the effect on B cells.

October 30, 2009 | Registered CommenterKurt G. Harris MD

i recall reading somewhere that medium chain fa's, such as in coconut oil, are directly absorbed and travel via the portal circulation to the liver, while longer chains are of course incorporated into chylomicrons, traveling via the lymphatics to - eventually- wider circulation. the theory was that the direct transport to the liver made for more efficient processing into energy rather than storage. i am not knowledgeable enough to comment on either the veracity or implications of this claim.

October 30, 2009 | Unregistered Commenterjeff klugman


At the risk of saying the same thing for the 3rd time, I am not indicting coconut oil or MCTs. I hope that is clear. So no need to be defensive about eating them. I would, however, disagree that the fatty acid composition of whole cream is as much outside our evolutionary experience as coconut oil. Mammalian milk from a herbivore we evolved to eat versus an extracted plant oil. That doesn't make it bad, but I have to disagree on that.

N=1 experimentation is a righteous reason when done with eyes open and appropriate Popperian skepticism.

Kudos. Carry on and let us know what you find. I wonder, though, what is is the fundamental premise of the advantage of MCT over LCT? Is it the enhancement of ketosis? There has to be a provisional hypothesis to justify experimentation with your own body, I am thinking. What is your provisional hypothesis?

Again, not saying the taste of coconuts isn't reason enough to eat coconut oil - let's stipulate that.

October 30, 2009 | Registered CommenterKurt G. Harris MD


I recall that from one of Peter's posts as well. Sounds like what happens with fructose, where the evolutionary strategy seems to be to keep it from the general circulation! Not sure that's bad, but I don't find that reassuring either.

October 30, 2009 | Registered CommenterKurt G. Harris MD

my hypothesis is a hypothesis of hypotheses; to validate existing hypotheses,

is "just because I can" not sufficient?

October 31, 2009 | Unregistered CommenterMCT

Hi Kurt,

You do not offend me by disagreeing with my diagnosis. For a long time my endocrinologist thought I might be a slow onset type 1, but because three years after diagnosis I have not evolved into type 1 he says that he has to say I am type 2. I should indeed probably be classified as type 1.5 but he doesn't say that and hasn't done the antibodies test. I think the fact that I eliminated wheat and all grains shortly before diagnosis meant I have not gone on harming my pancreas, true. I also began Dr Bernstein's way of eating very soon after diagnosis so that my carbs are very low indeed, I find it entirely consistent with Paleo eating.

My vitamin D status is high.....100 ng/dl at the last test, testing again in two weeks !


October 31, 2009 | Unregistered CommenterAnne


You are either not stating the hypothesis you are testing or you don't have one.

"Just because you can" is a a valid reason to do anything not illegal or unethical.

Most would agree with me, I am betting, that it cannot serve as any kind of scientific or even rational claim to justify self-experimentation.

I honestly thought you might invoke the claim that MCTs can help you lose weight or maybe they enhance your health in some way not found with LCTs.


Sounds like you are dong everything you can do for type I.5 in any case. Bernstein is a great guide. Good luck and let us know how it goes!

October 31, 2009 | Unregistered CommenterKurt G Harris MD

I have a confession to make! Not only do I practice "lacto-paleo" (PaNu) eating, but I am also a self professed "food snob" or what today is known as a "foodie".....a.k.a. I enjoy cooking, and so only use the finest ingredients!

I use expeller pressed coconut oil from Tropical Traditions (it is deoderized and tasteless as I despise the taste and smell of coconut), and use it to saute, bake (good substitute for shortening) or baste when I don't need or want the food to brown excessively. Occasionally I use a high quality extra virgin olive oil , but only as a finish (never heated).

My first choice for cooking though, is always pastured raw butter!!

I never felt that coconut oil was a "magic bulllet" either Kurt, and use it only when the recipe would benefit from its use. I would imagine that it would be considered a godsend for those who are lactose intolerant however! Unfortunately, a good many in the paleo community are still strict anti-dairy, so in their case, coconut oil is the go-to for cooking oil.

BTW: In case any of you wonder, it IS possible to eat full PaNu style, and still put food on the table that would rival any meal in a fine restaurant!

November 1, 2009 | Unregistered CommenterJenny Light

Strike the "lactose-intolerant" comment above. There is only a trace of lactose in butter, so that is no excuse people!

November 1, 2009 | Unregistered CommenterJenny Light

Regarding coconut oil vs. butter, the former is even higher in saturated fatty acids and lower in PUFAs than the latter, which makes it more resistant to oxidation during cooking. Also, butter is not entirely fat, so if cooking at high temperatures, I would choose coconut oil rather than butter to avoid the AGEs. I assume browned butter is higher in AGEs than fresh butter.

November 2, 2009 | Unregistered CommenterJLL

Why do I use coconut oil? Well, first I don't use a lot in cooking because my family doesn't like the taste (we use lard or butter instead). But I take it orally (1 tsp or so a day) because of the anti-microbial properties - ref. Dr. Enig (http://www.westonaprice.org/knowyourfats/coconut-antimicrobial.html). I'm hoping it improves lymph health. My lymph nodes have a tendency, after swelling because of an infection, to take their time about going back down, sometimes on the order of months. (Note the first time it happened I took the precaution of FNA biopsy etc, nothing. It just took 8 months for the node to return to normal, so as long as the node remains soft and does not continue to increase in size I don't worry). I don't know why my body does that, and standard docs are no help, so I'm hoping the coconut oil will use its "magical" anti-microbial properties as it flushes my lymph system. It's just a guess, but I am pretty sure it won't hurt.

November 2, 2009 | Unregistered CommenterLucy


You are correct about the lower total PUFAs, but my pastured butter has a 6:3 ratio of 1:1 so the very slightly higher total PUFAs are in a completely balanced ratio - which to me makes the PUFA content better, not worse

Your own chart shows the major difference to be MUFAs, not PUFA.

Not much Vitamin K and conjugated linoleic acidid in coconut oil.

Pretty trivial to make ghee from butter - nothing but fat, then.

I agree coconut oil is great to cook with high heat, though. Also, it sure tastes good.

November 2, 2009 | Registered CommenterKurt G. Harris MD

Dr. Harris, my apologies. I think my last comment was borderline disrespectful. You are right, those were two of the benefits I would claim, but I thought it's so obvious to mention, it's almost a joke. Maybe not to others.


Webster, I think you thought I was challenging you more than I was. I like coconut oil!
Writing a blog can give you an emotional investment that can make one a bit defensive, you can see it in my responses, too I am sure. I enjoy your contributions here, rest assured.

November 9, 2009 | Unregistered CommenterMCT

You repeat in your blog several times the suppresion of PTH by 25D, an argumentation that comes from the work of Hollis and Holick. Unfortunately, I think this assumption is wrong. The relationship between PTH and 25D is everything else as a single mathematical function (PTH (25D) but depends on other factors. In deed, it is not 25D that transcribes PTH but 1,25D. Thus the complete path is 25D -> 1,25D -> PTH.

You may look to the original data, not to the derived mathematical models that may be wrong. The data shows a tremendous scattering of PTH in dependence of 25D. That is what is measured.

best regards,



I don't recall saying PTH depends on nothing else.

You are saying that PTH does not rise to liberate more calcium with very low D levels?

What is your point?

1,25 D activity locally is driven by 25 D concentrations. Never said that was not the case.
25 D doesn't cause transcription of anything directly (usually)

November 9, 2009 | Unregistered CommenterAlex

I am clearly a victim of lab error on my latest vitamin D results. Approximately six months ago, I was tested and found to be low--the value I was told was "24" but I don't know the units. My doctor put me on 50,000 units of vit D once a week--I was retested last week and the value came back at "4". (Again, I don't know the units.) Of course they asked me if I had been taking the vit D--I have taken it religiously. I'm going for another lab test tomorrow.

November 24, 2009 | Unregistered CommenterMNH
Comments for this entry have been disabled. Additional comments may not be added to this entry at this time.