Kurt G. Harris MD

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 an animal-based diet high in fat, low in cereal grains and relatively low in carbohydrate.

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

« H1N1, Vitamin D3 and Innate Immunity | Main | Raw paleo and food re-enactment »

Bone Density Assessment

Lex Rooker, a reader of PaNu and a blogger in his own right, has asked me to comment on the results of his recent DEXA bone density scan.

I believe Lex had this done for two reasons. Firstly, he was a vegan for 20 years, which to the degree he was relying on “healthy whole grains” and was likely deficient in Vitamin K2 from animal sources, he might legitimately be concerned about his bone density. Secondly, Lex is in the avant-garde of VLC carnivory, following a raw meat/no dairy regimen. He needs to assuage those (not me) who think his diet is deficient relative to the SAD, or those who think (like Loren Cordain, for instance) that one must use a pH meter on food to avoid dissolving your bones due to “acid base imbalance”, as if our kidneys are not smart enough to do that for us. Unless Lex eats butter, there might still be some concern about K2, but a priori I am generally not concerned that a Steffanson- like diet that otherwise has enough D3 and K2 is bad for your bones.

He has posted his results here.

Here are Lex’s comments and the abstracted results:

“I finally got the results of my DEXA scan.  Considering that I was a devout vegan for 20 years, and dental x-rays from 6 or 7 years ago showed that my deteriorating dental health was due in large part to loss of bone density, I’m pretty happy with the results.” 

Age: 58   Gender: M   Height: 73 inches   Weight: 165 pounds


Region              BMD    Young Adult%   T Score     Age Matched%     Z Score


L1-L4              1.212             99%              -0.1              104%            +0.4

Neck                0.976             91%              -0.7              103%            +0.2

Ward               0.805             84%              -1.2              103%            +0.2

Trochanter     0.858             92%              -0.7                98%            -0.2

Total               1.009             93%              -0.3              101%              0.0


Studies were performed using a Lunar DPX IQ.  Technical quality of the scans were excellent with no artifacts. According to the World Health Organization guidelines, the patient is classified as NORMAL.  Based on these results, a followup exam is recommended in two to three years.

“This is my first DEXA scan so this becomes the base line.  At this point I really don’t know much about them and the best my doctor could do was read the report which indicates that everything is normal – whatever that means.  If anyone can provide a more comprehensive analysis it would be greatly appreciated.”

Lex had a dual-photon x-ray absorptiometry scan or DEXA scan. This study uses low dose x-rays at two different energies to acquire data used to calculate body density. A thin beam is passed through the body part with two energies. How much the body attenuates (diminishes) the x-ray flux and the difference in attenuation between the two energies allows calculation of the density of bone in the scanned area.

Geeks can go here.

The first thing to understand is the reason we might be interested in bone density. We are interested in it to the degree it predicts bone strength, and we care about bone strength to the degree it predicts fractures. However, in the same way that the density of a pile of bricks and a mortared brick wall might be the same, but they obviously differ in strength, we use density of bones as a proxy for bone strength with the critical assumption that the structure of the bones is normal. This may or may not be the case in the pathologic state, but with healthy bone of normal structure, it’s a fair assumption that is clinically accurate.

The density of your bones, properly considered, is in units of mass per unit volume, for example grams per cubic centimeter or g/cm3. If you have bone density measured with quantitative CT (an accurate technique not used today as it has a fairly high radiation dose) the results are reported in g/cm3. Confusingly, DEXA scans give a measure that is clinically accurate but sounds like it comes from “flatland” – the units of density are in grams per centimeter squared or g/cm2. This is because the data are acquired from a planar (two dimensional) projection of a three-dimensional volume – your body. A density reported in g/cm2 is disturbing to those of us with training in the physical sciences, but it correlates with true density measurements closely, and is the basis for all the clinical data acquired on populations.

So what do the numbers mean?

In interpreting these numbers, it’s good to know what we know as well as what we don’t know. We know from clinical studies that there is a certain risk of fracture for white women that can be estimated from their age and the bone density measured at the proximal femur (hip) or the spine.

Note that age and bone density are independent predictors of fracture risk. A 65 year old with t-score (explained below) of -3.5 has the same fracture risk as a 80 year old with a better t-score of -2.5. Not all fracture risk relates to bone fragility. Weakness due to age-related sarcopenia (muscle loss) and neurological degenerative diseases also affect fracture risk.

We know also fairly well how some drugs can mitigate this risk of fracture in white women. In those with a history of previous fracture or with t-scores worse than -2.5, bisphosphonate drugs can reduce fracture risk by as much as half.

We don’t really know as well how data from other ethnic groups or from white men predict fractures.

From Lex’s study:

Region              BMD    Young Adult%   T Score     Age Matched%     Z Score


L1-L4              1.212             99%              -0.1              104%            +0.4


L1-L4 means the first through fourth lumbar vertebrae in the spine (low back) were measured. BMD of 1.212 means the areal (flatland) bone density at these four levels averaged together is 1.212 g/cm2.

Young adult % means Lex’s BMD at L1-4 is 99% of the mean for a male of age 30.

One caveat about BMD in the spine. Measuring BMD at the spine can be “false negative”. If there is enough arthritis, bone spurs can falsely elevate the bone mineral density measured, and give a reading that seems normal when the central structure of the bone is actually weak. The interpreting radiologist has the responsibility to look out for this situation by looking at a low-resolution image obtained at the same time as the DEXA scan.

The t-score is the same statistic normalized mathematically, and expressed as standard deviations above or below the mean for a 30 year old, presumed to be the peak of bone mass. A t-score of zero means you have the same BMD as a 30 year old of your sex that anatomic site, regardless of your age.

T-score is the clinically used measure and the one on which fracture risk estimates and treatment decisions are based. A good rule of thumb is that each decrement of t-score of -1.0 doubles your lifetime risk of fracture. So a t-score of -2.0 is about 4 times the risk as 0.0. For hip and spine BMD, a t-score of -1.0 equates to having BMD about 15% lower than those with t-score of 0.0 (normal 30 year old). A woman will typically lose about 1% of bone density per year after age 60, so if starting at -1.0 it would take about 10 years to get to -2.0.

The WHO (world health organization) arbitrarily defines t-scores of 0 to -1.0 as normal, -1.0 to -2.5 as osteopenia. (More than half of post-menopausal women are “osteopenic” by this definition.) The WHO defines t-scores lower than -2.5 to be osteoporosis. Scientifically, the risk of fracture is a continuum. These WHO definitions are more about enabling comparative epidemiology and big pharma marketing of drugs than they are about science.

Lex’s t-score at L1-4 is -0.1 meaning his BMD is just 1/10 of a standard deviation below the mean for 30 year old men. 

The z-score is your bone density relative to others of the same age. In Lex’s case, his L1-L4 score is 4% higher than the average 58-year-old.

The z score is not predictive of fractures, it just tells you how bad or good your numbers are relative to your peers. Being the winner of the nursing home bone density derby among your 90-year old peers does not give you any protection. It’s the t-score that counts.

Here are some charts from a good review article in JAMA. You can look at these along with my short narrative to get a feel for fracture risk and bone density.



Continuing with Lex’s results:


Neck                0.976             91%              -0.7              103%            +0.2

Ward               0.805             84%              -1.2              103%            +0.2

Trochanter     0.858             92%              -0.7                98%            -0.2

Total               1.009             93%              -0.3              101%              0.0


For the rest of Lex’s study, the numbers mean the same, they are just numbers from different body sites. Neck means femoral neck of the proximal femur or thigh bone. Ward means ward’s triangle, a particular small anatomic area. Trochanter is another specific area of the proximal thighbone. Total refers to the whole hip area.

Femoral Neck and Total are the most predictive of future fracture risk. Incidentally, it is hip fractures that are most significant clinically. Spontaneous insufficiency spine fractures are common and painful, but are not as life-limiting and life threatening as hip fractures.

Finally, what can we say about Lex’s bone mineral density?

At the least, we can say that we see no evidence of any significant abnormally at any of the areas surveyed, at least based on the stated numbers and assuming the spine numbers are not falsely elevated.

There is absolutely no way to know if his bone density might be higher now if he had never had a vegan diet, but there is also no real reason to suspect that it would be.

From what we know of the way those with severe nutritional deficiencies are able to correct their bone density once they get proper nutrition, it is reasonable to think that if Lex did have osteopenia from his veganism that it may well be fully corrected by now.

In the context of the concern that his present raw paleo diet might be “leaching the calcium from his bones” due to acid-base imbalance from lack of lettuce or some other nonsense, we also see no evidence for that.

Finally, we should be aware that all the reference data, the "normal" values, may be representing pathology which we have made normative by our perverse definition of normal. It seems quite likely that the epidemic of hip fractures in old white women, heretofore blamed on northern european genetics and long lifespans, may be yet another disease of civilization caused by deficiency of Vits D3 and K2 and grain consumption.

So should you have your bone density assessed?

If you have a history of long term steroid use (prednisone, etc.), if you are a woman over age 50 on the SAD, or if you have some concern that prior or current dietary indiscretions (lots of healthy whole grains and lack of D3 and K2, for instance) have affected your bone health, then DEXA is a safe and reasonably effective test to get reassurance.

I have a DEXA machine at my imaging center. We are not necessarily the cheapest, but if you mention the PaNu weblog and want to come in for a scan you will get a small discount and also something you never get anywhere else - a sit down consultation with the radiologist to review your scan. We also can do coronary calcium scoring on our 64-slice CT. I don’t think everyone needs calcium scoring (blog posts on that are coming up eventually), but if you decide you want it we can review those results with you, too.

Finally, the Vitamin D post discusses osteoporosis (the medical definition, not the WHO one) and my recommendation from that post remains:

"Vitamin D, grain avoidance and eating grass-fed butter and hard cheeses (for the K2) are my strong recommendations for avoiding osteoporosis."


Clinical Use of Bone Densitometry - Cummings, Bates and Black

Clinical Challenges in Management of Osteoporosis - Vondracek, Minne, McDermott



Reader Comments (21)

"Unless Lex eats butter, there might still be some concern about K2, but a priori I am generally not concerned that a Steffanson- like diet that otherwise has enough D3 and K2 is bad for your bones."

I read the Stefansson study recently and indeed it seems that an all-meat diet, despite being low in calcium and highly acidic, didn't hurt their kidneys and bones. Do you have any more info on why this is so (e.g. is potassium important in this)? I guess the organ meats they ate provided them with some K2.


September 22, 2009 | Unregistered CommenterJLL

As a person diagnosed with osteoporosis I found your article particularly interesting I don't eat raw Paleo but I do eat low carb Paleo which is high in protein. Although I eat green veggies, I do worry about the high protein affect on my bone density, mostly because whenever I have my six monthly blood tests my serum urea is always high. My endocrinologist says it's because I am dehydrated but I'm not dehydrated ! I make sure I am not ! I drink plenty and my urine is always dilute. My serum urea levels were in the normal range before I went Paleo (three years ago), but since going low carb Paleo they have been high every single time ie last month serum urea was 11.00 mmol/L ref range 2.6 - 6.7 (this is not the BUN test - I live in the UK). I think it is the protein in my diet doing this as my creatinine, sodium and potassium levels are normal. Do you think I should pursue this further ?


September 22, 2009 | Unregistered CommenterAnne


My first comment would be that we have no way of knowing if Stefannson and Anderson had any decrement in bone density on their diet - they certainly had no accurate way of assessing that in the early part of the 20th century. An x-ray only shows low bone density once you are severely osteoporotic or when you have a spontaneous fracture. We have to be honest and say we just don't know.

I agree with you about the organ meats. When in civilization, where did they get their Vit D? The Inuit probably got both K2 and VIt D from organ meats, and obviously, fish livers.


How high is your protein intake? If it is very high and your serum creatinine is normal, then high protien catabolism is indeed one of many explanations for elevated urea or BUN levels. If your doc is not concerned about it and has knowledge of your diet that should reassure you.

But why not eat more fat and less protein?

September 22, 2009 | Unregistered CommenterKurt G. Harris MD

Hey Kurt,
Great to see you back at blogging. I'm looking forward to your Brother-in-Law's guest post, can't wait! Thanks for all that you do. Between you, Robb Wolf, Whole Health Source, Theory to Practice, Free the Animal, Hyperlipid, Mark's Daily Apple, and Primal Wisdom, one can find all the info they need about nutrition. Good stuff.

September 22, 2009 | Unregistered CommenterMark

Also, about the acid-base relationship could you do a simple run-down on that?
My hypothesis is that we are designed to eat an appropriate number of carbs to fuel the brain and such (<20% of diet made up of root veggies (potatoes and other tubers) and some fruit, I don't think we're meant to eat broccoli), an appropriate number of proteins (around 1.4 g/kg/day assuming that one is eating 10% to 20% of carbs, otherwise one would need more protein to fuel gluconeogenesis), and the remainder will be met by fat to meet EFAs.
This approach should, assuming mostly real food/"Paleo", keep one healthy and allow them to have good body composition and performance. What do you think?
Thank you again.

September 22, 2009 | Unregistered CommenterMark

Hi Kurt,

I'm not sure how high my protein intake actually is, but at a guess I'd say between 100g and 120g per day, eg today I've had 3 oz sardines for breakfast, 1/2 lb salmon for lunch and I'll be having 1/2 lb chicken with some bacon for supper, plus 2 oz or 3 oz of almonds and walnuts inbetween, and veggies with each meal and small amount of milk in tea. I think my diet should be quite high in fat too. I always eat a lot of oily fish and meat with the fat on and eggs, plus I always cook my veggies in coconut oil to increase the fat (I need to put on weight, I lose it very easily despite eating extremely well) and I always eat lots of nuts for inbetween meal snacks. Not sure where to get more fat short of simply eating it straight !

Btw, I must comment about your article when you said that a reporting radiologist should look at a low resolution image of the lumbar spine when interpreting the BMD because arthritis could create a false negative. I've had two DEXA scans and on neither occasion did the radiologist do this, yet on a CT scan (of my heart) the reporting radiologist commented that I had degenerative changes in my spine. Now that I've read your article I think I want my DEXA scan to be reinterpreted ! My lumbar spine BMD is reported as being slightly better than my hip density, and usually it is the other way round, so your article made me think. Certainly when I have another DEXA done I shall write a short note for the radiologist !


September 22, 2009 | Unregistered CommenterAnne

Your analysis went far beyond anything I expected. Thanks so much for taking the time. Bone density is so much in the news (and advertisments) these days, yet few, including our family doctors, are able to explain the results of a DEXA scan in layman's terms. My guess is that this post will be widely referenced.


September 22, 2009 | Unregistered CommenterLex Rooker


thanks - My day job(s) make demands that make blogging difficult intermittently, but I don't feel like making "excuse posts" all the time.

Publication will always be somewhat irregular unless I start writing 'filler" posts or posting pictures of food.

I'll add acid/base balance to the list of topics.


thanks - I hope you feel reassured

September 22, 2009 | Unregistered CommenterKurt G. Harris MD

Excellent post Dr. Harris! Thank you for taking the time to cover this topic in such detail.

September 22, 2009 | Unregistered CommenterCavePainter

oh please post some pics of your food! i'ts fun to see what loca carb bloggers eat!
great post too, and glad to see u back!

September 22, 2009 | Unregistered Commenterreamz

Nice to see you back Dr. Kurt and thanks for this excellent post.

September 22, 2009 | Unregistered CommenterRandy Evans


The radiologist will usually comment on it if there is enough bony sclerosis or osteophyte to falsely elevate the bone density. If there is not, they usually don't mention it. Just because there are "degenerative changes" does not mean there is enough osteophyte to throw off the results. Just disc space narrowing is "degenerative change".

September 23, 2009 | Unregistered CommenterKurt G. Harris MD

Hello Dr Harris,
It is a Dairy RA question I guess...I have a swollen sore finger joint after eating cream. I guess this is an autoimmune reaction to the casein- would that be right? I think that is more likey than a reaction to a high saturated fat intake in my evening meals. I also eat Foie gras twice weekly but I do not think it would be that either. I take a small MSM supplement for my growing hair....
Am guessing is the cream - is the only dairy I have and is at 25% fat. But along with aching knees I have a swollen knuckle this morning for the first time in my life- dairy me thinks?

I eat no grains or gluten, just the cream and fish/ meat and added fats.

What a pain- literaly- when this happens, will it go away again and recur each time I digress? Or does it become a chronic condition that never goes even if the dairy is dropped?
Any thoughts please - I am 40 years old and have never had a sore joint before in my fingers.
Thanks, Violeta

September 24, 2009 | Unregistered CommenterVioleta

Hi Kurt,

i worked out the composition of my food yesterday, which was a fairly typical day: proteins 120g, fats 155g and carbs 40g Would that be classed as very high protein ?


September 24, 2009 | Unregistered CommenterAnne

Well, the whole acid-balance thing didn't make sense to me either. When I look at a random biochemistry textbook, like this one, it seems like the body has a whole arsenal of methods to control the pH level of itself. Don't know why it would use calcium.

So I looked if there was some research to support the theory. I found this one, which claims that counteracting the acid load of the diet resulted in a lower urinary nitrogen wasting. As I understand it, the theory is that the body breaks down muscle tissue to free glutamine, which is used to counteract the acidity of the diet. Why not use other mechanisms, I don't know.

Also, this study, found that an acid-forming study increased calcium excretion by 74%.

I have to say that the acid balance theory doesn't make sense, but I cannot explain the results of these studies. Would someone care to comment?

September 24, 2009 | Unregistered CommenterMike


This will definitely be an upcoming blog post. The "acid/base theory" is somewhat plausible in that there are some experimental and theoretical reasons to think that acid base dietary balance might be important, but I do not believe it.

For one thing, it fails the PaNu dogma that diet, as part of a biological sytem, is parametric. That is, biological systems tend to operate within parameters with a fair amount of tolerance. Precise macronutrient levels, daily caloric intake, outside temperature, etc are all variables that the human organism should be adapted to within ranges that are not that precise, as long as foods (and food amounts) that are substantially outside our evolutionary experience (like the three main neolithic agents of disease) are avoided.

It is as implausible that our paleolithic anscestors had to carefully balance their diets with regard to acidity or alkalinity of food as it is that we evolved to weigh and measure our food or analyze its macronutrient ratio to three significant figures.


On 2000 kcal/day that woudl be about 20% of calories from protein. Sounds fine.


If you can relate any abnormal symptom to any novel food, whether it is dairy protein, seafood or even beef, eliminate it! Have you only recently added any of these?

I am highly skeptical that saturated fat per se could be immunogenic or cause any "swelling".

If these symptoms persist, see your local physician

September 24, 2009 | Unregistered CommenterKurt G Harris MD

It's perhaps worth noting that in the (in)famous 1928 Bellevue experiment, both Stefansson and Andersen were observed to be in negative calcium balance. "It was also noted that both the calcium excreted in the urine and the total output were greater than the amounts excreted when they were receiving mixed diets. This may have been due to the acid nature of the diet." (http://www.jbc.org/content/87/3/669.full.pdf+html) A table is included in the text. The authors conjecture, "This may have been due to the acid nature of the diet." As I read it, the Ca deficit was pretty small, but I suppose that over a course of years it could lead to bone loss, if there were no compensatory adaptation to the low Ca intake. At the same time, it suggest that even a very modest intake of Ca from non-meat sources would be enough to make it a non-issue. A few handfuls of a paleo vegetable such as purslane would do the trick.Moreover, as I understand it, vitamin D has an effect on Ca utilization, and Stefansson and Andersen may not have been getting much vitamin D at all.

September 25, 2009 | Unregistered CommenterTodd


That is very interesting. Calcium absorption indeed can increase dramatically in the presence of adequate D and even supplement levels of calcium are often not enough without it.


September 26, 2009 | Registered CommenterKurt G. Harris MD

I have read that the reported bone density for a DEXA scan for a short woman can be lower than what it actually is. Is there any truth to that, and is there an adjustment that should be made for shorter people? I am a 5' 1" female and would love to hear that my hip T-score of --2.7 is artificially low because of my small stature.

October 2, 2009 | Unregistered CommenterLynn_M


Have not heard that one - it would be pretty hard to have a score that low artifactually.

October 2, 2009 | Unregistered CommenterKurt G. Harris MD

Do you basically agree that diets with a net acid load will increase calcium excretion? I suppose carnivorous mammals have a very different system to handle the acid load of diet, like lions and our nearest relative, the Neanderthal?

The study I mentioned previously only had the people on their diets for four days. Perhaps the kidneys need more time to normalize pH levels?

Are there any other explanations? If an acidic diet really increases calcium excretion by 74%, they have a strong argument of a meat-heavy diet contributing to weak bones.

October 8, 2009 | Unregistered CommenterMike
Comments for this entry have been disabled. Additional comments may not be added to this entry at this time.