## Monday, June 29, 2009

### LDL Calculator

Commenter Kiwi Geoff kindly wrote a program that calculates LDL using the Friedewald equation and the equation from this paper, which may be more accurate for people with a total cholesterol over 250 and triglycerides under 100. For people whose triglycerides are over 100, the Friedewald equation should be relatively accurate. You can plug your total cholesterol, HDL and triglycerides into the program (in mg/dL), and it gives you both LDL values side-by side. Here it is:

LDL Cholesterol Calculator

Thanks, Geoff.

Sushil said...

I plugged in my numbers:
TC 178
Trig 59
HDL 57
and got LDL values of 131 and 109. Surprisingly, on my report the LDL (calc) value is 111. I guess this lab (Sonora Quest) is not using the Friedewald eqn.

Kiwi Geoff said...

Hi Sushil,

Is there perhaps a typo in your message? I ran your raw numbers through the calculator and got the following output:

Friedewald = 109.2
Iranian = 90.8

It makes me wonder do some labs use "manual keying" into a spreadsheet or is the LDL number a result of automation from a computer calculation?

So not only do we have formula that are not so good for those of us with a "healthy regime", but perhaps typos can make the cholesterol numbers - a wonderful sophisticated random number generator system!

Later on tonight, I will also do a version of the "LDL Calculator", for those of us who live in a country where we have results given in molecules (mmol/L) rather than by weight (mg/dL) as per the USA system.

Regards, Kiwi Geoff

Sushil said...

I could have sworn ...
You're right, these are the numbers I get now.
Wonder if you were working on the site while I was getting the previous numbers ...

gallier2 said...

I tested the calculator with my numbers and I can tell you that the lab in France where I made the test, uses the Friedwald equation.
The units used in France are a bit different but ot difficult to convert (multiply by 1000/10=100).
TC=1.87 g/l 4.82 mmol/l
HDL=0.43 g/l 1.11 mmol/l
TG=0.81 g/l 0.92 mmol/l
LDL=1.28 g/l 3.30 mmol/l
Friedewald: 127.8 mg/dl
Iranian: 122.7 mg/dl

Kiwi Geoff said...

Bonjour gallier2,

I have just posted a mmol/L version of the calculator. So you don't have to convert to mg/dL.

If you "refresh" the link that Stephan posted, you will see at the top of the page, a link for a mmol/L version of the calculator.

Likewise, here in New Zealand, the calculator proves that "Medlab" use the Friedewald formula, which leads to elevated LDL readings for those of us who are shunning the carbs.

I have also put a "worked example" on the calculator page, so folk can check their computer is giving the right answers.

Regards, Kiwi Geoff.

gallier2 said...

Hi Kiwi thank you for your calculator. My lab gave both numbers, the one in g/l and the one in mmol/l. The conversion I had to do was between g/l in mg/dl which is only a multiplication by 100, so it was not difficult.

Jeff said...

Just plugged mine in

TOTAL: 272
HDL: 76
LDL(calculated): 184
Triglycerides: 62

The calculator drops the 184 to 154. My doctor will be pleased but will still prescribe the statin(which I won't do).

My father and brother have very similar lipid profiles, with higher LDL, than most with the same paleo dietary pattern.

d said...

Recently tested at 23,

Total 264
HDL 94
Trig 51
LDL (Friedewald) 160/159 (Lab gave me 160, equation gives 159)
LDL (Iranian) 125

I eat a diet high in saturated fat. My meat is usually not grass fed, I still drink beer, eat occasional bread, eat potatoes, eat practically no vegetables, and I do not take any supplements or vitamins. The doctor wants to to put me on statins provided I can't change my numbers within 3 months. Even if he doesn't know that my LDL is a calculated number, shouldn't he see that even the American Heart Association cites Trig/HDL and Total/HDL ratios as indicators of heart trouble, and that my ratios are great? Also, my blood pressure is 90/70. Is is reasonable for the doctor to assume I'm at risk for heart trouble?

d said...

Also, I'm more worried at being put at risk for high insurance premiums than heart disease. Can companies discriminate against high LDL numbers? If so, has anyone had luck requesting an actual measurement of their LDL?

Anna said...

d,

Yes, insurance companies can heavily weight LDL numbers, so you are right to be worried about that. You probably won't be denied coverage, but the premiums will be much higher, perhaps for less coverage.

When you say "tested at 23", I assume you mean 23 yoa. You can try to avoid cholesterol panels at this age, esp since your more significant HDL and trig are great and better than most peoples, unless you are testing privately, such as through www.mylabs.com or at the local health food market on the day they have a machine set up. If you test through your doctor or insurance, then the results and your doctor's notes on your compliance with his/her recommendations goes into "your permanent record" and into the big medical data base (that sort of works like a credit file but with medical data).

Personally, I'd look for a new doctor, one outside an HMO, and one with a more enlightened understanding on lipid factors and what they mean and don't mean (if at all possible). If you don't want to do that, insist on an NMR or VAP test to directly measure LDL, particle number and size (pattern A, AB, or B), to get this doc off your back about a statin Rx. Or you can do a coronary calcium scan to show that you don't (yet) have plaque building in your arteries. You probably can't get insurance to pay for that, though. But insurance will pay for a lifetime of statin meds. Ironic, huh?

Also, you could periodically spot-check your post meal blood glucose numbers to make sure the results stay under 120 mg/dL, as elevated glucose is a big factor in the long term development of CVD. With your trig I wouldn't expect a problem, but glucose tolerance status is good to monitor and trigs can stay low if carb intake is low enough, even if glucose tolerance is impaired.

Fasting BG isn't a good way to monitor BG health - though it is the most common, cheapest, and most convenient way for docs to monitor glucose regulation. FBG is often the last indicator become abnormal, demonstrating that glucose regulation is blown. Doctors miss glucose regulation problems all the time by relying on FBG. I know this personally.

You can check 1-2 hours post meal containing about 75 g of fast-absorbing CHO. You can do this yourself with a glucose meter and test strips purchases OTC at any drug store. Expensive yes, but worth the info, IMO. Or if you know someone with diabetes, ask for a BG check (use a fresh lancet!). Jenny has great info on self-testing BG at her excellent site, www.bloodsugar101.com.

Kurt G. Harris MD said...

Anna

"Or you can do a coronary calcium scan to show that you don't (yet) have plaque building in your arteries. You probably can't get insurance to pay for that, though. But insurance will pay for a lifetime of statin meds. Ironic, huh?"

I agree with you recommendation to get a doctor who understands NMR LDL measurements.

However, the sensitivity of calcium scoring for angiographically proven coronary disease in women under 60 is only about 50%, and 90% of women in the 40-44 age range have a score less than 6. For men, 75% of men in the 35-39 age range have a score less than 4. So using some Bayesian reasoning, you can see that a calcium score as predictor of coronary disease would be useless for a 23 year old woman, and even for most women under 40, and useless for a male aged 23 as well.

My LDL
Friedewald = 44
Iranian = 33

I have always had a very low total cholesterol :D

JD said...

Friedewald = 201
Iranian = 184

Lab LDL = 201.

Mark said...

First, since this is my first post after lurking here for more than six months, I want to heartily thank Stephan for creating this wonderful resource.

It's coincidental (for me) that the topic of insurance came up in these comments. I've been following a paleo eating pattern for close to a year, and the one "negative" is that my (calculated) LDL has gone up (while my HDL has increased slightly and Trigs have dropped to the 50s). The drawback is that I'm at the point in my life where I need to think about buying some life insurance, which requires a cholesterol test that will make me seem like a higher risk than I really am. (Although if anyone has an incentive to truly understand what really matters in blood lipid data, it's the insurance co's...)

Anybody know whether there is a way to "game" a cholesterol test, to temporarily lower my TC/LDL numbers? Some ideas I've contemplated following for, say, a week or more before the test: laying off the sat fats (perhaps replace with mono fats?), take more EPA/DHA than usual, take a higher dose of niacin, and/or eliminate all carbs.

Any suggestions would be appreciated!

Mark

Kurt G. Harris MD said...

Hi Mark

Everything you suggested (esp. lay off sat fats) plus oatmeal to bind cholesterol in your gut. No benefit to your heart but may reduce your TC.

Stephan Guyenet said...

Mark,

Replacing saturated fats with monounsaturated should help. You could also try taking a plant stanol supplement. I wouldn't recommend it in the long term, but a couple of weeks is probably OK. You may want to give your body a couple of weeks to adjust to the new diet before being tested. Good luck.

Dr. B G said...

Mark,

If you obtain a real or NMR LDL it may indeed be a lower LDL compared with that calc LDL.

Also, if you order the density via NMR then showing your insurance plan that Pattern A (22.5 nm and higher) is heart protective and not associated with coronary disase... may help to actually LOWER your rates. Do you have high BP? Try not to have high BP or abnormal kidney tests (creatinine; drink plenty of fluids even for fasting labs so the Creatinine doesn't temporarily kick up esp if you have nice LEAN Paleo hawwtt BMI -- more mass means more Cr) during any insurance evals.

For my insurance... I sorta failed to tell that I do triathlons... they like to kick up the rates I hear.

-G

exitSector1011 said...

First comment here so, as others have said, thank you, Stephan, for all the information you have published here.

I'm 26 and just got some numbers that have me a little worried.

HbA1C = 5.5
Trigs = 44.5
TC = 370.5
LDL = 222.3 (Iranian calc = 169.3)
HDL = 140.4

Has anyone seen numbers as high as these?

Kiwi Geoff said...

"d" said the following:
---------------------
Total 264
HDL 94
Trig 51
LDL (Lab 160, equation gives 159)
LDL (Iranian) 125
---------------------

Just a quick comment on the above numbers "d". The LDL Calculator that I wrote, takes the numbers you give it, substitutes those into the two equations and gives the answer without doing any rounding up or down. I wanted to do it that way, so we could see how the LABS get their numbers.

For example your (Friedewald) LDL comes out as "159.8" from the calculator. If we round that UP, to "160" it is exactly the same number as the LAB gave you.

I could have added "rounding" code, but it is more useful the way it is - so we can see how the LAB makes the LDL estimation.

Climbing now onto a soapbox!

It would be so nice to lay ones hands on some direct LDL cholesterol results of what "we" would identify as "healthy" people. Using linear regression (which these days is a mere High School task), we could generate a better equation to estimate LDL from the three other measurements.

If the population does change from the current Lo-Fat dogma to Lo-carb lifestyle, then the existing Friedewald equation is going to make a fortune for the Statin industry. People's health will be degraded by nothing more than a "wonky equation" .

Regards, Kiwi Geoff.

Robert McLeod said...

Kiwi Geoff:

It should be possible to get a much better fit with a higher order function, cubic or whatever is required, that is still trivial to calculate in Excel or whatever.

I could certainly do it, all I would need is a database that includes the relevant numbers for many patients: total cholesterol, HDL, measured VLDL and LDL, and trigylercides.

What does it say about the diagnostic value of cholesterol tests that this hasn't been done?

Kiwi Geoff said...

Robert Mcleod said:
------------------------
all I would need is a database that includes the relevant numbers for many patients:
------------------------

I agree the mechanics of doing the statistics are easy, but who are to be the "patients" Robert?

It is like rejecting people on a jury - that person ate an apple last year with pesticide residue, that person used margarine 5 years ago, that person has been seen near a fast food outlet, that person eats grain fed meat.

Maybe we can't use anyone in the USA (or even New Zealand for that matter).

Where do we find a human being that is living in perfect harmony with their metabolism - as decided by their genetic sequence?

Maybe a handful of Masai perchance?

Currently we are judged against the simple equation of Friedewald, that was derived from a population clearly representative of the WESTERN DIET, whose Triglycerides were raised by too much farina and sugar!

We can look at the likes of the WHO database, where we see things like Blood Pressure rising with age. Which to me just confirms the horrors of the Western diet, rather than how the human machine should age with a lifestyle matched to its gene sequence.

So it looks like we need to argue for the next fifty years, as to "whose" cholesterol results we will allow in the survey to derive a better equation to "estimate" LDL.

Regards, Kiwi Geoff.

Stephan Guyenet said...

ExitSector1011,

Wow, those are some unusual numbers! Are you on an atypical diet? I wouldn't necessarily be concerned. I've never seen an HDL that high.

exitSector1011 said...

Stephan,
My diet is not atypical of the readers of this blog and Peter's. I've been eating high fat for ~1.5yrs now; before that I followed a conventionally healthy diet --- low fat, whole grains etc --- for about 2yrs; before that it was all out junk.

When I switched to high fat I ate a lot of meat, almonds and walnuts, olive oil, fruit and almost no starch for about a year but never felt that good on it. I've never had the tiredness after carbs effect and never felt energetic with carbs <35g even after a year. I found your site and hyperlipid ~6 months ago and upped carbs to 60-70g mainly from starches and get ~70-80% cals from fat, hardly any omega6, sats >40% cals. Might up the carbs a bit more to see how it affects my enery levels. I know there's no requirement and all that but I gave it a year to adapt at very low carb levels and never felt right.

My doctor was, suprisingly, not trying to push statins on me but did warn me that the levels are right up at the top of the scale. He was very open about the uncertainty in all this but I've been referred to a specialist. It's NHS in the UK so I don't have to worry about insurance but I seem to be in uncharted territory at these levels.

Stephan Guyenet said...

exitSector1011,

It's hard for me to comment on your situation intelligently, except to say that if you didn't feel good on low-carb after a year, that's probably a sign.

The studies that noted associations between different cholesterol fractions and heart disease were done on people eating typical Western diets, so I don't think they are directly applicable to your situation. I will say that your HDL and triglycerides suggest that your LDL is about as large and fluffy as it gets (a good thing).

I do think it's interesting to note that hunter-gatherers, even those on high-fat diets, generally have low cholesterol. I haven't really succeeded in wrapping my head around that fact yet. Masai also have low cholesterol despite getting 33% of their calories from saturated fat. When they're put on a vegan processed food diet lower in saturated fat and cholesterol (this has been done), their blood cholesterol shoots up immediately.

Stephan Guyenet said...

By the way, your lipoprotein profile also suggests that you have good insulin sensitivity.

homertobias said...

Mark,
Get some pharmex red yeast rice and take it for a few weeks. (natural statin, not available in the US thanks to Big Pharma, currently available ryr has been stripped of its natural lovastatin).
Stephan,
The kitivans interest me. They smoke, eat alot of starch, have a whole lot of nonobstructive vascular disease on autopsey, yet don't have heart attacks,strokes, or periferal vascular disease in life. I bet if you put one of them in a MDCT they would have a high calcium score. I bet if you measured their carotid IMT it would be bad. My take is that they have structural but not functional disease. Very interesting. Do you know anything about their cholesterol levels or their hscrp's?

Stephan Guyenet said...

Homertobias,

The Kitavans have a lot of vascular disease? I wasn't aware anyone had done autopsies on them. Do you have a reference for that?

I agree that artery disease doesn't necessarily lead to heart attacks. There are several examples of populations with extensive artery disease and a low rate of heart attacks, including the U.S. and U.K. around 1900, and the Masai in the 1950s.

Kitavan men age 40-59 have an average serum cholesterol of 186 mg/dL, women are at 243 mg/dL. HDL for men at 39 mg/dL, LDL at 128 mg/dL, trigs at 46 mg/dL. Hardly an "ideal" profile by mainstream standards, although the trigs look good.

Kurt G. Harris MD said...

Stephan and Homer

I would be interested in that reference as well. I have heard the same about the Masai - "sclerosis" without heart attacks. One problem is terminology used in some of these old studies. Atherosclerosis, calcification of plaque and coronary occlusion or myocardial scarring as evidence of MI are all different things. Calcification of plaque may depend on a variety of variables other than the rate of atheroscloerotic plaque growth, and, for instance, one could have atherosclerosis at a given rate but be less prone to complete occlusion with plaque rupture via platelet aggregation effects (aspirin, fish oil)*

Until I see Gadolinium Cardiac MRIs with late enhancement or autopsies on all of the Kitavans, I will stay low carb. EKGs used in the Kitavan studies are not a sensitive way to survey for previous MI, and definitely not for coronary atherosclerosis.

*As I recall from GCBC, the japanese have higher stroke rates to partly offset their presumed lower rates of MI -this may be the gotcha for the gimme with all that O-3 fish oil - prevent clots in the heart, cause hemorrhagic strokes in the brain. Another reason not to go nuts with fish oil.

homertobias said...

Stephan and Kurt

I think the mistake is mine here. I will double check later. I may have been remembering Dr. George Mann's work on the Masai, partially modernized diet, normal BP and Chol, normal ekg's, and severe atherosclerosis on autopsey. I wonder what their hscrp was, how they would scan, etc. I also would love to dig back into the Jupiter study. They had to screen a whole lot of people to find the relatively rare low ldlc high hscrp group whom they studied. I would want to study the opposite group - high ldlc (preferably large and fluffy) and abysmally low hscrp. Of course I imagine no drug company would sponsor that one.

Kurt G. Harris MD said...

Homertobias

Large artery atherosclerosis as indicated by, say, aortic calcification, does not correlate as well as one might expect with coronary atherosclerosis. I intend to-re-read the Masai autopsy study when I get time.

Stephan Guyenet said...

Homertobias and Kurt,

There have been autopsy studies of high-carbohydrate traditional cultures. The South African Bantu didn't have much atherosclerosis 50 years ago, compared to European caucasians. This is one of the big justifications for the Ornish crowd. That being said, apparently they had a fair bit of strokes (probably hemmorhagic). Here's a reference (free full text):

http://www.circ.ahajournals.org/cgi/content/abstract/29/3/415

I wrote a post about the two Masai autopsy studies, including Dr. George Mann's, a while back. I posted a couple of the figures:

http://wholehealthsource.blogspot.com/2008/06/masai-and-atherosclerosis.html

Kurt G. Harris MD said...

Thank You Stephan

homertobias said...

Stephan, Kurt:

I can't find any data on HSCRP and the Kitivan. (I do mean kitivan this time.) So they have a diet high in carbs from tubers, smoke, the women have cholesterol profiles in their 40's with TC 240's, low hdlc, increased triglycerides which would worry a traditional internist. But they are thin, don't exercise much, and are free of CV disease. Yes, I hear you Stephan, no refined grains,sugars,or omega 6 PUFA's. I'm more interested in markers of disease than you are. Do you know of any data on HSCRP in the Kitiva? For that matter, do you know of any data on HSCRP and any paleolithic tribe free of disease?
Kurt,
I absolutely agree with you on MDCT's in women under 60 unless they are diabetic. Also the data that says that the percentage of plaque which is calcified is constant across the atherosclerotic disease spectrum to me is suspect. CIMT may be a better modality but it is so tech dependent. I would love to know how you would ideally use imaging techniques to help stratify disease at reasonable cost.

Kurt G. Harris MD said...

Homer

I am going to post some soon on Coronary calcium. The idea that there is a constant percentage of coronary plaque that is calcified is false. I have not found much so far directly proving this but I believe it is easy to prove inductively. My own reading and experience interpreting both coronary calcium and coronary CTA does not support the "constant at 20% of plaque" argument nor the utility to calcium scoring in asymptomatic young persons.

Calcium score is a good tool for epidemiology and evaluattion of symptomatic patients.

CIMT is noninvasive, easy and cheap but is operator dependent and insensitive unless you have gross metabolic abnormalities.

Stephan Guyenet said...

Homertobias,

I'm pretty sure Dr. Lindeberg's group didn't measure CRP. They didn't measure HbA1c either, which would have been awesome to see (personal communication from Dr. Lindeberg). However, there's an interesting new marker that came out of the study:

http://www.ncbi.nlm.nih.gov/pubmed/17374168

Kurt,

I'll be interested to see your post. One of the things I learned when I was reading autopsy studies about CHD throughout the 20th century in the UK is that the prevalence of severe atherosclerosis with calcification stayed the same or even decreased from 1908-13 to the 1944-49, while the prevalence of milder forms such as fatty streaks increased substantially. The milder forms actually paralleled the increase in heart attacks, while the more severe forms didn't. Any thoughts on that?

Kurt G. Harris MD said...

Stephan

Any chance of providing links to those autopsy studies? I am a non-academic (former acedemic with no university access) and am handicapped by having to pay quite a bit for full articles when they are old. It is nice to buy only the good ones if i can.

Briefly, I think there is a degenerative type of atherosclerosis (DA) that is a necessary but not sufficient condition for coronary events (acute inflammation and thrombois, IT). DG alone, calcified plaques on autopsy, DG + IT american style heart attacks (MI)

Masai on blood and milk, Inuit, plains indians - no DA, no IT

Old Masai - DA but no IT

SAD DA + IT

DA caused by high carbs or fructose?

IT caused by insulin and linoleic acid?

There also seem to be differences in the distribution and certainly the clinical meaning of atherosclerosis in large central vessels like the aorta and in the 2 mm diameter coronary arteries. The worse the metabolic abnormality, the more peripheral the arterial pathology, to the point where Type II diabetics are famous for having such severe disease at the small vessel and arteriolar level that the pathology is too small to bypass.

This is why I am not as sanguine about the Kitavans. Even though they clearly tolerate high carb consumption and smoking by not having the endpoints of MI and maybe even lung cancer, we have no idea what their vessels look like, AFAIK. You could say DA does not matter, but it might be nice to have compliant, noncalcified arteries even if you don't die prematurely.

I want autopsy studies or coronary CTA on Kitavans, before I go back to actively seeking carbohydrates like gluten-free starchy tubers. I'll stay closer to Optimal Diet ratios in the mean time.

homertobias said...

Kurt
Well put! My feelings exactly. I like your DA and IT. DA I call "the mess in the arterial wall". Dr Peter Libby writes eloquently about the intimate details of this mess, and all of the potential biomarkers
involved. Unfortunately there are way to many intermediaries. Tumor Necrosis Factor probably has more subtypes than SUBWAY.
As regards to IT. HSCRP may be our best marker here. The question becomes, what makes a mess in the arterial wall all of a sudden decide to become Mount Vesuvius and erupt? Is it a characteristic of the mess itself, (most research centers here), the fibrous cap over the mess, (which can be attacked from both sides) or a characteristic of the blood rushing through the lumen. What is the best plaque instability marker? Maybe Dr Paul Ridker is right and it is a good old fashioned acute phase reactant.

homertobias said...

Stephan
Thanks for the reference. But what? Apc Igm is cardioprotective but is positively associated with lp (a) levels and pufa levels in cholesterol esters? And is negative associated with myristic SFA levels? That's wierd.

Stephan Guyenet said...

Kurt,

I've been coming to basically the same conclusion, that atherosclerosis seems to be necessary but not sufficient for MI. You need thrombosis on top of it. The way I see the coronary heart disease epidemic in the US and UK is something like this: we always had atherosclerosis, but we didn't start having frequent heart attacks until we created an inflammatory/thrombotic environment by adding industrial vegetable oils to our diet right after the turn of the century (increased n-6). I don't know what was causing the atherosclerosis. Maybe wheat, maybe sugar, maybe infections, vitamin D deficiency, vitamin C deficiency, weird snake oil elixirs?

Here's the free full text of the Finlayson et al. paper on MI in the UK in the 19th and 20th centuries. I think you'll get a kick out of it:

http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2557402

And here's a post I wrote on it, as well as another classic autopsy paper by JN Morris:

http://wholehealthsource.blogspot.com/2009/05/coronary-heart-disease-epidemic.html

I would love to have more information on the Kitavans as well. Autopsies would be fantastic.

Homertobias,

I haven't read that paper in detail so I don't know how good the story is. Changes in cholesterol ester composition don't always respond to diet as you'd expect if I recall correctly, but I don't remember the details off the top of my head.

In any case, the Kitavans eat a lot of saturated fat (coconut) and very little PUFA, particularly linoleic acid.

Tom said...

Stephan,

I've been catching up on your posts from the summer and just discovered this. Many thanks to Kiwi Geoff for taking the initiative on this LDL calculator! I wanted to share my numbers because I have especially low triglycerides, which makes for a massive difference in calculated LDL:

TC: 248 mg/dL
HDL: 70 mg/dL
TG: 28 mg/dL

Friedewald: 172.4 mg/dL
Iranian: 121.5 mg/dL

cobb529 said...

worried about my 48yr old husband labs. Total cholesterol 280. LDL cholesterol 197. Triglycerides 147. LDL density patern A/B (abnormal) also remmant Lipo is 30. some other numbers were high but I don't know what they mean. apoB100-calc 141 and LDL-R-C178. He exercises every day, does not smoke, is not overweight. Lately his BP is also getting higher 148/87. Dr wants him to take a weird ultrasound test called an IMT scan. Anyone heard of that?