While I was looking through some of the references cited by Krauss in his discussion paper about the roll of omega 3 fatty acids in the lowering of triglycerides, I came across this paper from 1999 on the effects of high and low fat diets on LDL subclasses. I read the abstract, re read it and read it again, scratching my head. Then I slogged through the whole paper. I just couldn't make out whether Krauss considered saturated fats to produce atherogenic changes in plasma lipids or not. After reading the paper I still couldn't decide what his conclusions were! Here are some quotes from the discussion:
"Change in dietary saturated fat was associated positively with mass of larger LDL particles and with peak LDL particle diameter and LDL flotation rate. These results suggest, therefore, that feeding saturated fat is associated with increased mass of larger LDL"
I think this means that saturated fats make "good" bad cholesterol. What do you think? What do you think Krauss thinks? Can you tell?
"In the present study, correlation analyses revealed significant positive relations of change in intake of the long-chain saturated fatty acids myristic and palmitic acids with change in plasma concentrations of large LDL particles. These findings are consistent with studies showing that, of the long-chain saturated fatty acids, myristic and palmitic are the most hypercholesterolemic"
Gasp, saturated fats are hypercholesterolaemic. Good or bad? Surely hypercholesterolaemis is bad? Keys said so. But large LDLs... atherogenic or benign????
"In summary, the present study showed that changes in dietary saturated fat are associated with changes in LDL subclasses in healthy men. An increase in saturated fat, and in particular, myristic acid, was associated with increases in larger LDL particles (and decreases in smaller LDL particles [did you spot this inset???? It matters]). LDL particle diameter and peak flotation rate were also positively associated with saturated fat, indicating shifts in LDL-particle distribution toward larger, cholesterol-enriched LDL"
Now, would you eat something which produced "cholesterol enriched" particles?
"Although there is a possibility that a subset of large LDL particles may be atherogenic (46), earlier results (37) point to a differential benefit of low-saturated-fat diets on LDL concentrations in individuals who have an atherogenic lipoprotein profile denoted by a predominance of small LDL particles"
What on earth does "a differential benefit of low-saturated-fat diets" mean? Your guess is as good as mine on that one. Answers on a postcard to Hyperlipid.
From the results, from whence it didn't seem to make it to the discussion, except in that little inset in brackets I highlighted:
"During the high-fat diet, saturated fat was correlated negatively with mass of smaller LDL particles"
What does this mean? Isn't that good? Does this mean the small dense LDLs are less dense?
As far as I can see, saturated fats improve every lipid marker they looked at BUT just look at this comment, again from the discussion:
"An increase in dietary saturated fat has been associated with the progression of CAD independent of LDL-cholesterol concentrations (56), and in cross-cultural studies, higher intakes of dietary saturated fat are associated with higher prevalence rates of CAD (57)"
Translates as: Never mind the lipid paramenetrs, saturated fats are BAD. Full stop. I think it's called nailing your colours to the mast.
BTW Ref 57 is Keys. Scraping the barrel a bit there!
So what did I do? I was beginning to doubt anything I'd read about lipid particle sub types, so I googled something along the lines of "LDL particle size CAD" or the like.
This was the first hit. Really, no kidding.
It's clear cut, cutting edge science from 1994:
"The lowest quintile--those with the smallest and most dense LDL particles--had more than three times the risk of heart attack as the quintile with the largest LDL particles"
Classic quote from: Ronald M Krauss, circa 1994.
Hmmmmmmmmm
Peter
Subscribe to:
Post Comments (Atom)
35 comments:
Interesting study:
"Impact of plasma oxidized low-density lipoprotein removal on atherosclerosis"
http://www.ncbi.nlm.nih.gov/pubmed/18559699
That fits neatly with recent articles of Chris Masterjohn.
http://www.cholesterol-and-health.com/Does-Cholesterol-Cause-Heart-Disease-Myth.html
Wow...it is confusing. But the research is consistent, isn't it? The dogma of cholesterol aside, isn't there a growing body of literature showing increased risk for smaller LDL particles?
And to answer your question, yes, I would eat something that was hypercholesterolaemic if it increased LDL particle size.
As an aside, this study was published in 1999. How much older are the data? Did they have the ability to discern (or were the even concerned with) trans fatty acids?
And since I'm lazy and don't feel like putting this to yesterday's post, I answer a big, saturated fat NO to your question. I'm betting their numbers would be better given your parameters.
Very humorous - at many levels!!
Thanks,
Philip Thackray
Hi Sven and Brian,
I've not had time to read my Cholesterol Times in detail but yes, Chris M is pretty consistently driving towards oxLDL and some specific problems with small dense LDL exposing their hydrophobic core as the cause of arteriosclerosis. I was interested to see the Lp(a) and oxLDL may be the same particle, if my skim was correct. These are all pieces of a jigsaw which is becoming fairly self consistent. I'll just keep plodding along and see if I end up in the same place!
My biases drive me towards glycation as the first step in oxidation and questioning what the oxidised lipids are there for... I don't think they are random.
And of course small dense LDL, which oxidise easily, are there in response to a glycating diet, ie sucrose/white flour based.
Peter
Sven,
That's interesting. What a clever study design.
I just read Chris M's article, it's very informative.
Hi Peter,
I have currently switched to a higher fat way of eatting, yet i find it increasingly hard to get some variety in as I appear to be sensitive to dairy.
I haven't really eaten any grains for over 9 years as I was a raw vegan for a long time. I went back to animal products around 2 years ago and feel alot better from it. but when I eat dairy i tend to get very congested.
In trying to keep protein down I am confused about what to eat.
I eat alot of butter, egg yolks and coconut oil and cacao butter but can't say that i feel nourished but eating large quanties and feel sick and tired of eating them all the time.
any thoughts or ideas on the dairy?
i have friends who tell me that it's normal to get congested for a sort time while your body adapts but i'm just not that sure.
I would love to be following the optimal diet but dairy appears to be a problem and fatty cuts of meat are not that available in Hawaii. Only pork fat..
which I remeber i used to love as a child , crackling.. i'm a brit by birth. Anyway then that's high in PUFA so yes any thoughts would be appreciated.
pipxx
Hi Peter,
Do you love fish oil as much as me??! It shuts off NFKb very nicely too -- but you probably already new that!
I've seen some reports that omega-3 does lower insulin and glucoses -- maybe it depends on the person/genetics? How ketotic they are? I dunno...
What I don't get is how bad is the fish oil affect on apoA1 -- seems to lower serum levels of that stuff. But I guess if we consume enough cholesterol and saturated fatty acids... we're fine -- the apo A1 will be flying as high as possible!
THANK YOU FOR YOUR AWESOME POSTS!!
-G
Dr b g- Omega 3 is anti inflammatory, and since inflammation plays a considerable role in insulin resistance (as in a cycle, inflammation->insulin ->inflammation), it must therefore be true that omega 3 will help ameliorate hyperglycemia and hyperinsulinemia.
I think it's definitely good stuff! Omega 6, in contrast, is pro-inflammatory and probably contributes to higher insulin, weight gain, and IR (I haven't done much research into it but this is my intuitive feeling given that is established inflammation and IR are forever linked).
Regarding Krauss, his view is probably "all LDL is potentially bad" therefore sat fats are bad because they increase LDL, end of story (even though sat fats shift cholesterol production away from "bad" LDL and toward "good" LDL, to Krauss a "good" LDL is kind of like comparing misdemeanors to felony criminals; some are worse than others, but all are criminals thus bad).
Hi Pip,
You can convert just about any low fat meat to high fat meat by adding butter. If you issues with dairy, just clarify the butter. Once clarified there is nothing left except the lipids, which are not only good, they taste excellent too. Then you can have high fat prawns, beef, pork, chicken or whatever. For chicken using skinless allows you to cut down on omega 6s and replace the chicken fat with butter.
OK, to clarify butter, put it in a saucepan, heat it. The butter will separate out in to a water based layer underneath a much thicker fat layer. You just boil the water thoroughly until it's almost gone. Then pour off the fat, cool it and you're ready to use it. I often get a layer of watery material under the ghee. When the ghee is solid just scrape or wash way the watery bit. The protein and lactose are all in the watery bit. The lipid should be fine. After that take any recipe book, replace the vegetable oil with ghee, add some extra ghee to your already generous amount and cook...
Err, hate to mention this but liver has even more multivits than egg yolks. You need vitamin A.
Peter
Enjoyed your post! Here's a recent Krauss paper: "Protein in optimal health: heart disease and type 2 diabetes." http://www.ncbi.nlm.nih.gov/pubmed/18469290?ordinalpos=6&itool=EntrezSystem2.PEntrez.Pubmed.Pubmed_ResultsPanel.Pubmed_RVDocSum
The abstract seems to indicate that he may be coming around. For instance, it states: "High-protein, low-carbohydrate diets have been found to have positive effects on reducing risk factors for heart disease, including reducing serum triacylglycerol, increasing HDL cholesterol, increasing LDL particle size, and reducing blood pressure. These diets appear particularly attractive for use with individuals exhibiting the atherogenic dyslipidemia of metabolic syndrome."
And then he goes on to discuss the need for more research. But at least he's not still spouting "fat is bad." I see it as a hopeful sign.
This article was co-authored with Clifton from Australia, some of whose papers insinuate that low carb diets are bad, but also published that people on low carb diets had improved cognitive function. Am J Clin Nutr. 2007 Sep;86(3):580. These two are perhaps embracing protein, but not really fat as yet. Maybe a step in the right direction- we'll see.
Peter,
I often read something is pro-inflmmatory and other thing is anti-inflmmatory but I dont understand these terms well.
Since you have the knack of explaining complex science in layman terms, perhaps you could explain what exactly is meant when we say something (such as omega-6) is pro-inflammatory.
Also could you relate omega-3 being anti-inflammtory from the point of view of Bruce or Ray Peat (omega-3 is also harmful)?
A camel-milk drinking community with zero diabetes:
Zero prevalence of diabetes in camel milk consuming Raica community of north-west Rajasthan, Indianext term
R.P. Agrawal et al
Diabetes Research and Clinical Practice, Volume 76, Issue 2, May 2007, Pages 290-296
Dear Drs. Cynthia and David, you said that
"but also published that people on low carb diets had improved cognitive function. Am J Clin Nutr. 2007 Sep;86(3):580."
Yeah low-carbers improved their results, but keep in mind that the HCLF-dieters performed even better.
From the study:
"The results showed that participants consuming the
LCHF had significantly less improvement in the minimum stimulus
time required to make a correct response than did those
consuming the HCLF diet, and analysis indicated a moderate
effect size. Whether this effect was caused by the HCLF diet’s
promotion of greater improvement in processing speed or the
LCHF diet’s moderation of practice effects is difficult to determine"
From the conclusions:
"There was some evidence for a smaller improvement in cognitive functioning with the LCHF diet with respect to speed of processing, but further studies are required to determine the replicability of this finding."
They tested at baseline and week 8. That time should be enough for the body to make the needed adjustments for a high fat intake.
I like my ketones but I don't think LCHF-diet is any better for cognitive function if you have NORMAL GLUCOSE CONTROL. I haven't read any research that has concluded otherwise. The cognitive function of rats decline significantly if they are on a LCHF-diet. But yeah, always take rat-studies with a pinch of salt.
Hi Peter,
Just an update on the "calorie vs. macro counting" argument that I've made my own pet project. So here I am at 65kg, and I wanted to compare some notes with you:
Braesilika, another poster, got to 65kg and now maintains ideal weight on 200-250g fat/70-80g protein/and anywhere from 50-80g carbs per day. Calories must range from 2000-3000 per day. Dunno how much daily exercise is helping here. He also has a daily fast from around 7 a.m. to 4 p.m.
You maintain 63 to 65kg weight on pretty much unlimited fat and calories, around 65g protein and anywhere from 50 to 70g carbs a day, without much regular exercise. You do seem to have some intermittent fasting thrown into your schedule.
After reaching ideal weight, I've bumped up my daily calories to 200-250g of fat a day (yeah, it's great) for the last week. I don't think I've gained weight, and I'm never hungry. But time will tell. I stopped recording foods on Fitday in an effort to achieve some relaxed normalcy (and to see if all this is sustainable long term). I guess if worse comes to worst, I can always cut calories again if I gain weight.
My question is: how do you do all this without counting, tracking, weighing, or fussing? And how often do you monitor your weight to see if you're still on track? You seem to have a tendency toward needing to GAIN weight now, as opposed to the rest of us. Any idea why?
Again, thanks for all your help. G
Hi Siberian,
The paper under discussion is, unfortunately, beyond the reach of my athens account. From the multiple discussions here, it's pretty obvious that people who have normoglycaemia on a high carb diet do not need to eat a low carb diet. Kitava and the Bantu seem to get discussed frequently. So I have no problem with any macronutrient ratio, provided it results in normoglycaemia at physiologically "normal" insulin concentrations (whatever that means). For a remarkably large portion of the western population, especially those who are metabolically "broken", this is most easily achieved with LC eating.
The paper abstract is interesting and reminds me of this paper,with it's remarkable achievement of inducing deterioration of metabolic control in diabetic humans using a LC, high fat diet. I'd love to know how the results were achieved in both papers.
Peter
Hi Gunther,
Go for it. I just really have on meal a day which is usually based on half a pound of meat and the rest is eggs, fat, cream and chocolate. The bulk of carbs/protein comes in the main meal. I've some idea what the protein and carbs in cream and chocolate are and overall I allow myself some sloppiness. Protein at 50 or 80g/d doesn't seem that important any more than carbs from 30-70g/d. I weigh myself at least every other day but I have no idea what the subtle behavioural effects are of knowing my weight, particularly on my eating habits... Have a play with this too if you like...
Peter
Hi Gyan,
Yes, some discussion of anti and pro inflammatory might be a good idea. The camel milk paper was discussed at some length on Dr Bernstein's form back when I had the time to visit there regularly. Ultimately the carbs have to go somewhere, so the milk must either contain an insulin mimetic, and insulin release stimulator or an insulin sensitising agent. I never came to any real conclusion and the paper was not in the clearest of English, if I remember correctly.
Peter
Hi g,
Yes, I think that long term low dose omega 3s may well have beneficial effects on insulin sensitivity compared to the effect of large doses on hepatic lipid stores. It's even more complex as there are papers out there showing in cats, which are notoriously prone to hepatic lipidosis, that omega 3s are protective. But the mechanism of hepatic lipidosis in cats is neither fructose nor alcohol based, so it may be a different ball game. I also suspect the omega 6 background may have a lot to do with variable results...
Peter
Peter,
It is in the paper itself that camel milk contains an insulin-mimetic factor.
Didnt Cordain has a paper where he shows/claims that milk and milk products are hugh insulin-provokers.
Is that the mechanism for low blood sugar in milk-drinkers?. Is high insulin trouble ?
I am esp interested as my own diet is very close to Raica camle milk drinkers (only I consume buffalo milk products).
About 50% of my calories come from dairy
and 50% is carb (mostly wheat and lentils). PUFA is at 4%. Added sugar is 5-10 g/day.
My recent fasting sugar was 83 mg/dl
and HBA1c was 4.8%
(Raica camel milk drinkers were 82.7 mg/ml). I am sedentary though and no much in sun. (They are into heavy physical work in sun).
Cholesterol was borderline at 215
HDL 57, LDL 130 (particle size not
done).
Should I get my insulin measured?
Peter,
Here is one paper that seems to be saying that dietary PUFA downregulates lipogenesis.
He infers that PUFA promotes oxidation of carbs. I suppose he has not taken into consideration that the fatty liver.
PS I didnt understand 95% of the paper. I have the entire paper in case you want it. Has many interesting-looking plots
Obesity – A perspective based on the biochemical interrelationship of lipids and carbohydrates
Medical Hypotheses, Volume 68, Issue 5, 2007, Pages 1159-1171
Roger Jeffcoat
Hawthorn House, Higham Road, Chelveston, Wellingborough, Northants NN9 6AH, UK
Received 30 May 2006;
accepted 2 June 2006.
Available online 25 January 2007.
Summary
Many factors affect the onset of obesity including satiety control, reduced levels of physical exercise as well as hormonal and genetic parameters which influence the metabolic pathways leading to the net accumulation of triacylglycerol (TAG). The predominant fatty acid of human adipose tissue TAGs is oleic acid, reflecting primarily the composition of the diet but also the product of de novo lipogenesis. Consequently, both carbohydrates and lipids are potential sources of these stored fats. Many studies have been carried out using a variety of differing experimental protocols on healthy, obese or diabetic humans and animals in positive or neutral energy balance to establish the underlying molecular basis for obesity particularly in humans. This short review discusses the interdependence and control of the metabolism of lipids and carbohydrates as it relates to lipogenesis and proposes a unified hypothesis for obesity which brings together a number of different approaches focusing on (i) the interaction of dietary fat and carbohydrate, which typically represent not, vert, similar80% of the daily caloric intake, and their role in the synthesis of TAGs, (ii) the biochemical pathways which control the amount of TAG produced by controlling the composition of their fatty acids via the action of stearoyl-CoA desaturase (SCD), (iii) the control of lipogenesis and SCD by dietary polyunsaturated fatty acid (PUFA)next term and (iv) the interaction of previous termPUFAsnext term with the transcription factors, peroxisome proliferator activated receptors (PPAR) α and γ, which maintain the balance between oxidation and storage of lipids. The hypothesis focuses on the central role of stearoyl-CoA desaturase (SCD) and its inhibition by polyunsaturated fatty acids (previous termPUFAs)next term acting via transcription factors based upon data obtained from both animal and human studies. From these observations it should be possible to determine the relevance of the hypothesis to humans and to speculate how these aspects of metabolism may impact the risk of developing related diseases such as coronary heart disease, Type 2 previous termdiabetesnext term and certain forms of cancer.
Hi Peter and Siberian,
I didn't have access to the full text either... The paper you cite is interesting Peter. I've seen papers that show that fasting actually increases insulin resistance, e.g., J Clin Endocrinol Metab. 2008 Jul;93(7):2900. And of course there is the correlation of free fatty acids with insulin resistance. Maybe in diabetics at 27% carbs and high fat the blood glucose levels don't show improvement because they are shifted into fat burning mode (PPAR delta activation?) and effectively not disposing of the carbs they do eat.
Also searching the literature for "ketogenic diet" shows that the neurobiologist types are fascinated with the anti-epileptic mechanisms of high fat diets (strict ketogenic, but also Atkins induction now). Many mechanistic studies are being published (enhanced K channel conductance and increased GABA responses have been proposed, if I remember rightly). Clearly something is going on to affect the activity of some circuits in the brain. Whether the presence of ketones or lack of carbs (and better glucose control presumably) has an affect on cognitive function (in addition to anti-epilepsy or other benefits) I don't know, though many people say that it does for them.
Cynthia
I saw this at fightaging.org:
Dr. Nir Barzilai, who has been studying centenarians for genetic and biochemical clues to longevity. At the top of the list so far: HDL and lipoprotein size ... Small lipoproteins imply a shorter, less healthy life.
Gyan,
For what it's worth, I discovered years ago that milk is a potent hypoglycemic trigger, far greater than what carbs would imply. Skimmed types are the worst. Eliminating it completely from my diet made a huge step in controlling my blood sugar (I no longer have hypoglycemia anymore, now that I have gained weight and fixed my metabolism, but at the time, I had hypoglycemia constantly and my blood sugar numbers were pretty horrible).
A lot of diabetics and hypoglycemics report milk is way worse than what the carbs alone would indicate.
I suspect for normal people, drinking milk will only help lower blood sugar and prevent hyperglycemia by sharpening and quickening the insulin response. This might explain all those associations between dairy products/milk and weight loss.
Sometimes more insulin is not worse. Sometimes a little bit of insulin early can help prevent a lot more being needed later. Sometimes a little bit of insulin can normalize the metabolism so that we feel energetic and actually are losing weight.
You will find also this paper interesting:
Ttile: "A very-low-fat diet is not associated with improved lipoprotein profiles in men with a predominance of large, low-density lipoproteins"
Darlene M Dreon, Harriett A Fernstrom, Paul T Williams and Ronald M Krauss
http://www.ajcn.org/cgi/content/full/69/3/411
Especially Figure 1 and 4:
http://www.ajcn.org/cgi/content-nw/full/69/3/411/F1
http://www.ajcn.org/cgi/content-nw/full/69/3/411/F4
The last sentence of the paper says it all:
"A consuming average American diets, lipoprotein changes induced by further restriction of dietary fat and isoenergetic substitution of carbohydrates are not indicative of reduced risk of coronary artery disease; in a subset of men who convert to phenotype B, the changes are suggestive of an INCREASE IN CORONARY DISEASE RISK." (capitalized by me)
That means your friend Mr. Krauss must know exactly that carbs are the culprits but simply became more cautious about talking about it. People do strange things just as their retirement approaches.
It was the same story with the Framingham directors Caselli and Mann who begun writing books and telling it all only after they stepped down.
Aren't we glad to be self employed? 8-:)
Brilliant as always, Peter.
I visited my local farmers' market and bought some goat cheese. 55 % fat, and the goats are raised on rich pasture. And it tastes quite good as well.
Cheers,
Emil.
Hi Gyan,
Probably much easier and more useful would be a post prandial glucose curve after a typical meal. If you go above 7mmol/l (a bit arbitrary) you are probably doing damage with glucose, never mind insulin. A fasting insulin would tell you how much insulin you are needing to maintain that 83mg/dl (Bernstein's ideal value) but testing post prandial insulin (would be very interesting, but what is normal?) gets you in to areas where there is not a huge amount of data. The fact you run and HbA1c below 5% is a reasonable indicator that your glucose control is not bad. Of course if the milk has an insulin mimetic you wouldn't pick this up on an insulin immunoassay. Cordain on cows milk just said it provokes a big insulin spike, which probably relates to the specific amino acid composition of cows casein. Of course gluten is gluten......
Peter
PS the stearoyl-CoA desaturase paper is interesting. I believe insulin controls stearoyl-CoA desaturase... Have to go through that paper and also that HUFA ref you put up a few posts ago. Never enough time and 4 days surfing coming up starting Friday. They're not forgotten.
Hi Cynthia,
I've had a read at the full text of the cognitive function paper and it's interesting but doesn't put me off LC just yet. The authors do discuss reasons for their findings and they may well be explicable but I don't want to be just saying sour grapes because LC didn't win hands down the way it usually does. The p value was p=0.04 and there may have been issues with the learning process involved. The LC diet looked pretty good. There are a number of papers showing acute carb load improves memory, brain function and even alertness, but they tend to be single meal studies. I think one of them got a mention in the Low Fat Moods post. For the time being it's filed under "interesting".
Thanks to Napero for the paper.
Peter
Hi some commenter,
Would that have been on Jenny's blog? I think her conclusion is that longevity is a few genes in lucky people, so the rest of us can all forget it and go for the quick heart attack... Each to there own.
Jenny's post here.
Better get me some Prozac!
Peter
Hi ItsTheWoo,
You said
"Sometimes a little bit of insulin early can help prevent a lot more being needed later"
Loss of first phase insulin secretion is a hallmark of type 2 diabetes. Causal?
Peter
Stan,
I don't think there is any doubt that Krauss know exactly which lipoproteins are "associated" with heart disease. I don't think there is any doubt that he knows exactly which dietary interventions alter them in good or bad directions.
Does being a big noise in the AHA mean he has to bury this information somewhere deep in his brain, to let it come out in a few years time, when being a big noise is less important and isn't happening anyway? How many must die in the meantime?
Peter
Ta Emil,
OK, that's most of the comments up to date. Time to go use the industrial hired carpet cleaned to carpet clean babysick up off of several carpets. Ah, nursery tummybugs!
Peter
I think it is both causal and symptomatic. If you've got a weak first phase response (either because of genetics or early diabetes) your blood sugar rises unnaturally after eating. This increase in sugar makes your pancreas bolus out lots of insulin which then causes mild relative hypoglycemia. This tells your liver to bolus out lots of glucose... and the cycle begins again.
So in this way it is causal.
It is also symptomatic of course because an inability to regulate blood glucose and insulin is a result of the disease as well.
Hi Peter,
I wish I could read the paper in detail to see what I missed.
Why should acute carb loading improve "memory, brain function and even alertness" in the absence of merely alleviating hypoglycemia? I would suspect there may be other factors at work besides blood glucose per se. Neurophysiology is way more complex than that. I know my son used to be decidedly grumpy when hungry and would go into giggles and good nature when fed (didn't seem to matter much what the food was either). Now he's an avid reader of your blog and low carb practitioner!
I guess "sugar high" is an accepted concept, I just don't understand it. Sounds like it could be related to dopaminergic systems like other addictions and have nothing to do with carbs directly? Maybe a topic for full blown blog discussion someday...
Here is a National Public Radio (US) broadcast with Gary Taubes and Ronald Krauss as "guests."
35 min long.
http://www.npr.org/templates/story/story.php?storyId=15886898
Hi Petro,
Isn't there a relation with CAD and particules number, instead of size ? as suggested by this study pointed by Peter Attia here : http://eatingacademy.com/nutrition/the-straight-dope-on-cholesterol-part-v
Thanks!
Post a Comment