Blood Results

[Craig Loizides]

Dr G, the second study you listed is interesting because it shows a U-shaped curve between sleep and HDL/triglycerides for women. 6-7 hours sleep was best with more than 8 hours just as bad as less than 5 hours. The full text states that other studies (but not all studies) have shown U-shaped curves between sleep and mortality, diabetes, CVD, etc.

[Gaspard Winckler]

Hi all, I just got some results after moving towards a Paleo approach:

Triglycerides: 56 down from 63
HDL: 72 up from 60
LDL: 150 up from 90

After doing some searches it appears to be a common profile among people on the Crossfit boards: http://board.crossfit.com/showthread.php?t=42143 wfs

One explanation is problems with the calculation at low TG levels, another is that the 'fluffy LDL' weighs more and comes out with a high total cholesterol but is benign.

Interested if any other paleo eaters have similar experience.

[Gaspard Winckler]

So I found the study that indicates TG/HDL as being the key ratio and highly (r = 0.803) inversely correlated with lipoprotein particle size:

http://dx.doi.org/10.1016/S0009-9120(01)00263-6

That's from 2001. In 2004 we have http://dx.doi.org/10.1373/clinchem.2004.031757

"AIP [atherogenic index of plasma, which is log TG/HDL] was inversely and significantly correlated with measures of insulin sensitivity, such as the homeostasis model assessment and quantitative insulin sensitivity check index."

So knowing your TG and HDL can give you a marker for your insulin sensitivity, i.e., the lower the ratio the better.

The last piece of the puzzle is why people eating relatively clean paleo diets should see (apparently benign?) LDL increase.

[Garrett Smith]

I've seen two people (myself being one of them, an avid CFer the other) with LDL numbers over 200, while eating a pretty darn clean Paleo regimen and obviously exercising.

There is something to the LDL numbers among this crowd. As Gaspard pointed out in his post, both myself and the other CFer both had great HDL (high) and TG (low) numbers, which are what I'm more concerned about.

I will be testing again soon.

[Gaspard Winckler]

A few bits of research:

People's cholesterol levels react differently to diet depending on their genes, so some crossfitters can eat vast amounts of bacon and their LDL doesn't budge, but not all can.

Elevated LDL with good TG and HDL is common with all low carbers, both in the scientific papers and on low carb forums.

So there are at least 3 factors:

-something about low insulin and reliance on fat metabolism may increase LDL (by weight) in some people, due to increase in particle size; however this would go against the fact that LDL is well established as an independent risk factor (i.e. a high LDL still represents a risk even if the other 2 numbers are good, and treating LDL by itself reduces cardiac events)

-some insulin sensitive people may also be more sensitive to dietary cholesterol (eggs) http://atvb.ahajournals.org/cgi/cont...ract/23/8/1437

-some paleo dieters may be avoiding some of the more useful sources of soluble fibre like apples (and potatoes of course) that they used to eat and eating somewhat more saturated fat in getting their protein requirements

The main problem is that most research is not done on people who exercise and have no other risk factors, and who eat a low carb diet even though they are not overweight or 'dieting'.

However, the fact that this LDL (and TC) increase appears to occur across a proportion of low carbers in general, regardless of the activity level, would seem to imply that it is a phenomenon independent of exercise, and of insulin sensitivity (i.e. there will be some fit people who get heart attacks, although they are relatively rare). The fact that insulin resistance correlates only similarly to LDL in heart attack prediction (The Copenhagen Study) (http://atvb.ahajournals.org/cgi/cont...baha;17/6/1114) implies LDL should be monitored independently.

I've concluded from reading all this that I may not be in the genetic group of those who can eat saturated fat and cholesterol and keep a low LDL. I'm going to see what happens if I reduce the sat fat somewhat and make a point of getting plenty of soluble fibre when I do eat it.

[Arien Malec]

Here's a take on similar paleo lab results:

http://www.freetheanimal.com/root/20...poprofile.html

My take is that this is all new science -- epidemiologic studies on correlations of LDL to CV risk may not mean that high LDL causes CV risk, particularly if high LDL is typically found in a pattern with high TG and low HDL, large waist size, insulin resistance, etc. (metabolic syndrome).

What does high LDL mean in the presence of high HDL, low TG, good insulin sensitivity, and a trim waist, particularly when the LDL is of the large particle type, and apo B is low? Risk models based on the western diet don't really cover that situation, but I'd note that those profiles look a lot like Inuit blood lipid profiles (which is probably a good thing).

How you view this probably depends on how you view the lipid heart hypothesis. If you think that low HDL correlates with systemic inflammation, and that's the real heart risk, you'd see the lipid profile above as benign. If you think that high LDL and low HDL are the causes of heart risk, you'd scratch your head a bit about how to treat the lipid profile above...

[Gaspard Winckler]

Thanks, of course we are talking about risk *within* the already low risk population of fit people who exercise. This article http://www.theheart.org/article/527905.do shows the debate within the medical community where there is an example of basic LDL having the same predictive value as particle number.

The main use of the particle number is identifying risk in people who have an apparently low LDL number. So in your link above there is a link to a testing company website that shows someone with LDL of 98 (apparent low risk) who is in fact high risk because of her number of particles.

It doesn't follow from this (and I detect a bit of denial going on, especially on the atkins/ low carb discussion boards) that this means a high LDL ceases to be a risk factor in what is a very complex process. LDL and risk vary independently of HDL, TG and insulin sensitivity. If insulin insensitivity and inflammation were the whole story, HDL and TG would be totally predictive and LDL a dependent variable or irrelevant.

So if you can keep your high HDL and low TG but also lower your LDL by small diet adaptations (which are probably also beneficial in other ways), why not do it?

[Arien Malec]

Originally Posted by Gaspard Winckler

Thanks, of course we are talking about risk *within* the already low risk population of fit people who exercise. This article http://www.theheart.org/article/527905.do shows the debate within the medical community where there is an example of basic LDL having the same predictive value as particle number.

Great reference -- note that the parameter in question is non-HDL, not LDL (LDL is a subset of non-HDL). So high TG counts for non-HDL.

The best predictors of CV risk in that study are TC/HDL ratio and CRP. So someone who adopts a dietary approach that lowers TG and raises HDL dramatically, and raises LDL moderately, will show an overall risk reduction on those terms (particularly if CRP is also reduced).

I do agree with the notion that there's some wishful thinking in the LC community -- my only point is that the science points to overall risk reduction, but this is a fairly unstudied area, because most of the risk epidemiological studies correlate against is associated with metabolic syndrome. So what does low TG, high HDL, borderline high LDL mean for overall risk? Who the hell knows?