Where do triglycerides come from? Part I (Updated)

This post contains sufficient updates from the original (dated 5/11/11) that I consider it more of a fully updated version vs. a bump.   This began with my intent to link to this post in an upcoming discussion of triglycerides, when I noticed that Dr. Ronald Krauss was amongst the authors.   This study originally caught my eye because of   Marc Hellerstein's name, he of DNL not a major pathway in humans fame, that I've blogged on previously.

Now I have excerpted copiously from the discussion because the authors make several points relevant to the discussion of what comprises a healthy lipid profile.  As part of updating, I am breaking those excerpts up a bit more and adding some/more emphasis and additional commentary.

Background



Anyone who has spent any time perusing low carb sites is probably aware of the hallmark of carbohydrate restriction:  Lower fasting triglyceride (TG) levels.   In this study, they assess VLDL-TG, because in the fasting state, the vast majority of triglycerides are contained in the VLDL manufactured in the liver which packages up fatty acids as part of the TAG/FA cycle.

LC diets almost invariably "outperform" the alternatives in reducing fasting triglyceride levels.  Therefore the obvious conclusion is that carbohydrates cause triglyceride formation in the liver - and the obvious source of the fatty acids in these triglycerides is de novo lipogenesis - aka converting excess carbs into fat.  With no attribution, Taubes postulated in GCBC that as much as 30% of carbohydrate in a meal is destined down this path.  Countless others have repeated this or taken this notation from a biochemistry text out of context and stated similar.  I've discussed this misinformation here before, many times.  Yes, hepatic DNL does occur, but even in "normal" overfeeding this usually amounts to no more than a few grams per day.


The Study


So one day I was again looking into these claims and came across this paper:


Increases in fasting plasma triglyceride (TG) concentrations are commonly observed during the consumption of low-fat, high-carbohydrate (LF/HC) diets. ...
Endogenous hypertriglyceridemia, which is observed in individuals on higher-fat diets, is associated with increased coronary heart disease risk, but it is not known whether carbohydrate-induced and endogenous TG elevations share underlying kinetic mechanisms and therefore similar atherogenic risk.  
Differences in lipoprotein dynamics between the 2 forms of hypertriglyceridemia MIGHT contribute differently to the risk for cardiovascular disease.
Right there.  This is the question.  And this extends to other discussions such as the LDL debate.  
In the fasting state, most plasma TGs are carried in VLDLs. Overproduction of VLDL particles might generate a higher flux of atherogenic particles to LDLs in the plasma or into the artery wall. By contrast, reduced clearance of VLDL-TG does not increase the flux of cholesterol into the plasma.  Moreover, nonpharmacologic regimens, such as weight loss and exercise training, may increase TG clearance.
The liver secretes VLDL, and as the triglycerides are "stripped out" by lipases for use by various cells in the body or storage in adipocytes, they eventually become LDL (or are removed by receptors).  Thus if a normal liver pumps 100 VLDL particles, you can theoretically end up with 100 LDL (not quite this simple, but OK for this analogy).  If your liver pumps out 200 VLDL you could get 200 LDL, but if your liver pumps out 100 VLDL and clearance is slowed, the trigs may be artificially elevated but you still can only get 100 LDL in the end.  
TG secretion rates are elevated in endogenous hypertriglyceridemia, and some, but not all, workers have reported high TG flux rates in subjects fed LF/HC diets. Nestel et al. and Boberg et al. suggested, on the basis of indirect evidence, that the source of fatty acids supporting elevated TG synthesis on LF/HC diets may have been from de novo lipogenesis. If increases in dietary carbohydrate elevate de novo lipogenesis, this increase in fatty acid availability to the liver could drive hepatic TG overproduction.
It is interesting to note that increases in triglycerides are not routinely seen with studies of an HCLF diet.   This should be a bit of a red flag ... when there is a "hard and fast" relationship, then repetition, repetition, repetition should result in the same result over, and over, and over again.
Accordingly, the present study was designed to determine the metabolic and kinetic mechanisms underlying carbohydrate-induced elevations in TG in subjects consuming an isoenergetic diet that was composed of whole foods, rich in fiber and restricted in monosaccharides. Normolipidemic subjects and those with mildly elevated fasting TG levels were compared to determine whether these groups differed in their response to the LF/HC diet.... 
This was a nice study because various radiolabeled substrates were used so that the researchers could distinguish the source of the fatty acids as well as the rates of appearance and disappearance of these labeled triglycerides.

One of my major updates here is to look a bit more at the study design and subjects.  

Subjects:    n = 6 normolipidemic  ,  n = 5 hypertriglyceridemic
It is important to note the definitions used here are not conventional.  The "normolipidemic" individuals had TG under 100 mg/dL (designated NL) while "hypertriglyceridemic" individuals had TG between 100 and 200 mg/dL (designated HTG).  
In my opinion, it's somewhat irresponsible to attach the latter label to a range of values that is usually associated with normal to slightly elevated. While mild hypertriglceridemia is still technically hypertriglyceridemia, the suffix "hyper" is generally reserved for levels well above normal.   In this case, half of the range of their HTG designation belongs in the NL.   In any case, there's a huge gap in TG recommendations where genetic variants result in, and issues occur at, over 1000 mg/dL.  Further, TG become the "first course of action" when over 500 mg/dL.   This gap will be part of a future discussion.  

Choice of Triglyceride Levels to Study:

So, why did they look at high-NL-to-mild-HTG subjects compared with lower NL?   They tell us:
We chose to study subjects with moderately elevated TG, on the basis of previous work (E.J. Parks, unpublished data; and ref. 29) showing that carbohydrate-induced elevations in TG tend to be more variable in individuals with higher TG concentrations (> 250 mg/dL), whereas those with TG in this moderate range have a more reproducible response to LF/HC diets.  
In other words, in people with true HTG, carbohydrates may or may not contribute to the triglyceride levels predictably.  The carbs only have a predictable effect in ranges at or near normal.  This dickering about in Normalville will also be the topic of a future post.  But the low carbers seize on this sort of thing and attribute true HTG (over 250 or 500 or whatever ... not in the 100's that's for sure) to consumption of evil carbohydrates.  

Other Subject Characteristics:

Sigh.  I think these researchers fell down at the plate here in this aspect of this study.  There were only 11 subjects total.  We've established that they weren't looking for outlier HTG's, so I consider the following differences between groups to be unacceptable.  

Not sure how difficult it would really have been to find a dozen participants that differed only in triglycerides without the confounding weight/body fat differences.  Considering study costs and all that, it seems a minor imposition.  Further, since these were men, it is likely that the increased fat mass was abdominal in nature.  They could just as easily have cast this study as a comparison of normal to overweight men and noted there was a difference in triglycerides!

Study Diets:  

The diets studied were weight-maintaining calorie level diets where control was  F/C/P  35/50/16 and the "high carb low fat" study diet was 15/68/17.   

Now, study after study, after clinical nutritional study, where "usual intake" is assessed will put the SAD at around 50% carb.  I know, that's not what you are led to believe but it's true.  This control diet was about what averages tend to be while in general the fat tends to be a little higher and carbs a little lower (so we see fat in the high 30's and carbs in the high 40's most of the time).

This is important because the "control" is what is happening with normal and overweight men on the SAD ... not the HCLF diet.  The HCLF diet, at 15% fat is not even representative of any dietary prescription commonly designated and tested as "low fat", which usually means 30% down to 20% in most cases.    Also, given the caloric intake (about 2870 calories) the baseline carbohydrate consumption was around 350 g/day and was increased to around 500 g/day.   We'll return to this in the analysis of results.  


Major Findings


It is quite interesting to see what happened within and between these two groups:  
Fasting TG concentrations increased by 60% in subjects consuming the solid-food, high-fiber, LF/HC diet. These results underscore the ability of LF/HC diets to elevate fasting plasma TG even when the diet is composed of whole foods and is low in simple sugars.

This is the sort of statement that low carb advocates will pluck out of context and cite to demonstrate the insidious nature of all carbs.  Before moving on, let's look at that 60% figure.  Here is the Table of parameters measured (please see paper for P-values and notations)

While the NL group saw an increase in TG of 64%, the final TG of 100 is still well within normal levels.  The HTG group (aka the overweight men) saw their triglycerides jump only 53%, but jump from an average of just high end normal to elevated at around 230.   It is of note that only TG, LDL, HDL and VLDL changes were significantly different for the treatment.   Even in overweight men, add 150 g carb on top of 350 g they are already consuming and insulin and glucose remain the same.  Imagine that.


Production vs. Clearance:
Increased secretion of TG was not the primary kinetic mechanism responsible for this carbohydrate-induced TG elevation, ...  
... reduced TG clearance from the plasma was the major metabolic mechanism involved
In other words, while there may have been some increase in output, most of the elevation in VLDL-TG is due to the rest of the body (adipose and other tissues) taking up triglycerides.  More specifically:
...the transport rate of VLDL-TG ... was not increased on the LF/HC diet. Second, the half-life of VLDL-TG was prolonged, and the clearance of VLDL-TG was significantly reduced. Third, fasting apo B-48 concentrations were elevated.  
Also:
In the present study, the TG secretion rate of HTG subjects was 2.7-fold higher than that of the NL subjects on the higher-fat diet.
Given that the TG levels on the control diet were 61 and 149 for the NL and HTG respectively, a factor of 2.4, it would seem that in normal ranges of fasting triglycerides you're pretty much looking at a fairly direct relationship between production rate and circulating levels.  But even in the NL group you had changes within normal ranges of concentrations when carbohydrate content of the diet is increased:
 ... carbohydrate-induced elevations in TG were explained by reduced clearance of VLDL-TG rather than increased production.
....Reduced VLDL-TG clearance may be due to competition between VLDL and chylomicron remnants for TG hydrolysis in the plasma compartment, as fasting apo B-48 concentrations were elevated in the HTG subjects. ... The mechanisms responsible for reduced VLDL-TG clearance on LF/HC diets need further study.
This is in contrast to those with truly elevated triglycerides on the SAD-style high fat diet that do exhibit increased triglyceride production rates. The authors acknowledge differences in this result vs. prior studies but emphasize that those differences are probably explained by the form of the carbohydrate (solid/starch vs. liquid/simple sugar).


Where do the fatty acids in the TG come from? 
Finally, de novo lipogenesis was not increased.
There are clear distinctions between the source of the VLDL triglycerides between the NL and HTG groups as shown at right.  

The normal men (TG and weight) get virtually all of their fatty acids from circulating NEFA and show no difference between the diets.  

Comparing the overweight mild-HTG group to NL on the control diet, we see that about 10% fewer triglycerides are accounted for as coming from NEFA on the control diet but this drops even further on the LF/HC diet.   These "missing" triglycerides are speculated upon later in the paper (and this post).

But look at the white slivers atop the black bars.  Those represent the contribution of DNL which is insignificant for all groups (<5%).  The "worst case" is the comparison for the HTG group on control vs. high carb diet, where the contribution from DNL increased from 3% to 4% of total.   Remember, we are talking "insidious" carbohydrate levels here -- averaging 350 grams on the control diet and increasing to 500 grams on the HC diet (that's over a pound of starch for those of us not on the metric system!!)   In reality, most Americans on their SAD diets do not consume this much carbohydrate so it's not the carbs that are upping the triglycerides through DNL.  It's just not.

Now this paper is dated 1999 and I have yet to uncover any newer research disputing the results.   
Not only is hepatic de novo lipogenesis NOT a significant factor in contributing to body fat, it is also NOT a significant factor in contributing to the amount of triglycerides secreted by the liver.  
(Previous statement is mine.)     It cannot be stressed enough that in the normal weight men, their triglyceride levels on a "whopping" 350 g/day carbohydrate were a "mere" 61 mg/dL.   Putting them on a highER carb diet only nudged their TG levels to 100 mg/dL, which is well within normal range.

The overweight men, averaging just borderline of normal-to-high by usual standards are getting around 10% of the fatty acids in their VLDL-TG from some other source.   And yes, that "other source" is contributing even more to the VLDL-TG when these men are consuming a rather high carb (68%), rather low fat (15%) diet -- to the tune of an additional 17% from "other sources"


Where do the "missing" triglycerides come from?  

A reasonable observation to make comparing the NL vs. HTG is that NEFA supplies a relatively fixed amount of triglycerides up to a point.  When levels exceed a certain point, there's another source contributing.    Irrespective of diet, the HTG group derived a lower percent of their VLDL from NEFA, but this was "aggravated" by switching to a low fat diet, and considerably so (from 84 to 67%).  However this study clearly demonstrates that the other source is not the carbohydrates.

So where did the "missing" fatty acids come from?  In NL folks (C or LF-HC) only about 7% of the fatty acids were unaccounted for by NEFA+DNL, while this more than doubled to 16% for the HTG-C group and increased to 29% (almost doubling again) when the HTG group consumed a LF-HC diet.   
Hepatic stores:  Obese subjects have increased liver TG content.  Havel et al. observed that about two thirds of the NEFA taken up by the HTG liver escaped oxidation and were apparently stored, with only one third of the NEFA being secreted as TG.  This process could lead to the slow turnover of hepatic TG droplets in HTG subjects.  In the present study, a contribution from unlabeled hepatic lipid stores to TG synthesis may be less likely, because the subjects had been fasted for 24 hours by the end of the infusion test, which should substantially reduce hepatic lipid stores.
Seems the authors don't consider hepatic fatty acid stores as the likely source.
Visceral adipose stores are also a potential source of fatty acids for TG assembly.  This adipose depot is metabolically active and is an important predictor of abnormal plasma lipoprotein concentrations. Furthermore, individuals with upper body obesity have elevated VLDL-TG production rates. Visceral fatty acid sources could have contributed to the unlabeled VLDL-TG observed in the HTG subjects on the control diet, but an increase in their contribution on the LF/HC diet is less likely, as we and others found no change in body composition on the LF/HC diet.
Nor does it seem that the authors consider visceral stores as likely.  The increased contribution over the course of the study from this depot would have resulted in reductions in visceral fat.
Chylomicron Remnants:  Finally, an interesting potential source of unlabeled lipid for TG assembly on the LF/HC diet is chylomicron remnants cleared to the liver.  In rats, chylomicron remnants contain approximately 50% of their original TG when cleared to the liver.  Very little is known about this process in humans. The 2-fold elevation in fasting apo B-48 concentrations observed in the HTG subjects after the LF/HC diet suggests that chylomicron remnants might provide a significant source of VLDL-TG fatty acids in the fasting state. The increased presence of B-48–containing TG-rich lipoproteins was surprising, given that the blood draw for this measurement occurred after the subjects had fasted 15 hours and that the last meal they consumed contained very little fat (15% of energy).  However, Karpe et al. have shown that the chylomicron remnant pool contains lipoprotein populations that are heterogeneous. One population may have a very short half-life (<15 minutes), whereas others may circulate for much longer periods. The elevation in fasting apo B-48 associated with LF/HC feeding could have resulted from alterations in the intestinal secretion pattern of chylomicrons, such that more particles were secreted in the fasting state or that a subpopulation of the chylomicrons secreted postprandially circulated for a longer time. This result requires further investigation.
A long excerpt but couldn't find a way to chop it down.  It sounds to me that of the authors are leaning towards chylo remnants as the source.


Summary


Bottom line:  Not all hypertriglyceridemia is the same!  

The HTG group had higher VLDL-TG to begin with because they produced more, while the NL/LF-HC group had higher VLDL-TG vs. baseline because they cleared what they produced more slowly.

What is going on in the low carber?  I believe Volek has done some studies on more rapid clearance.   So you make a bunch of VLDL but they degrade rapidly to LDL and/or are cleared completely by receptors.  What of those with low VLDL-TG but high LDL?  To me the answer seems obvious but I'm open to other ideas.  To me it means you are producing a lot of VLDL that rapidly lose TG but are not removed by the receptors leaving a lot of LDL -- both particles and cholesterol in those cute and fluffy lipoproteins.  There is another possibility of direct LDL manufacture/secretion ... a topic for another time (if ever, limited time and all that).

What to make of the effect of carbohydrates on triglycerides?  This will be a topic of an upcoming post looking at the relationship between triglyceride levels and CVD.  Stay tuned ... (and if you are reading this some months from now and I haven't come back and edited in a link, shout out in comments!)   For now, there's two more parts to this series that I hope you'll check out!



Comments

Stephan Guyenet said…
I came across this study a while back, also found it informative. DNL wasn't very important on this particular HC diet. My suspicion is that DNL is dependent on the speed of digestion of the carb. If the carb is hitting the blood so fast that it can't be oxidized in a timely manner and glycogen is full, it turns into lipid. If it's hitting the blood slowly, the body just uses it directly for fuel and there's no need to use DNL. That, plus the fructose element, probably explains the difference between simple sugars and whole food starches IMO.

However, if you look at Kitavan serum lipids, they're high in palmitate even though their diet is low in palmitate, probably reflecting increased DNL. Their diet is 69% carb, with a significant fructose load from fruit. So I do think a long-term whole food high-carb diet can increase DNL, although I suspect the lipid produced in this way is still not a large contribution to total lipid stores (and in any case, does not alter the energy balance equation much). But DNL in that context is not necessarily a bad thing IMO, as long as it's physiological and not due to insulin resistance.
Melchior Meijer said…
Stephan, what's your take on physiological insulin resistance on very low carb? I just cannot wrap my hands around it. I've seen low carbers with horrible, lingering, blood glucose excursions after eating two potatoes. Low carb proponents argue it's no problem as long as you just don't eat many carbs. My question is if a physiological insulin resistance strong enough to dramatically impair glucose tolerance is really harmless in the long run. Could it mean that insulin signaling eventually goes too low, or is this just a naive fantasy? It could explain the high prevalence of osteoporosis in the traditional Inuit (lack of insulin signaling at the level of bone cells induces severe osteoporosis).

Regards,

Melchior.
CarbSane said…
Stephan: What's your take on whether trigs are "atherogenic" - e.g. there's some mechanism that these lipoproteins directly stimulate plaque formation - or that elevated trigs in the context of the SAD are merely a marker? It seems to me that all the high carb lean cultures with low CVD would argue against the former.
Stephan Guyenet said…
Hi CarbSane,

I don't know enough about it have an opinion, but I will agree with you that Kitavans don't have particularly low triglycerides. I've also noticed that most of the studies where people were fed a high-carb, high fiber whole foods diet for a period of weeks didn't see the expected rise in trigs. I think the form of the carbohydrate makes a big difference. I think Ornish and those guys often make that point.
CarbSane said…
Reinstating lost comment for:
Melchior Meijer
Stephan, what's your take on physiological insulin resistance on very low carb? I just cannot wrap my hands around it. I've seen low carbers with horrible, lingering, blood glucose excursions after eating two potatoes. Low carb proponents argue it's no problem as long as you just don't eat many carbs. My question is if a physiological insulin resistance strong enough to dramatically impair glucose tolerance is really harmless in the long run. Could it mean that insulin signaling eventually goes too low, or is this just a naive fantasy? It could explain the high prevalence of osteoporosis in the traditional Inuit (lack of insulin signaling at the level of bone cells induces severe osteoporosis).

Regards,

Melchior.
CarbSane said…
Stephan: What's your take on whether trigs are "atherogenic" - e.g. there's some mechanism that these lipoproteins directly stimulate plaque formation - or that elevated trigs in the context of the SAD are merely a marker? It seems to me that all the high carb lean cultures with low CVD would argue against the former.
CarbSane said…
Reinstating lost comment for:

Stephan

Hi CarbSane,

I don't know enough about it have an opinion, but I will agree with you that Kitavans don't have particularly low triglycerides. I've also noticed that most of the studies where people were fed a high-carb, high fiber whole foods diet for a period of weeks didn't see the expected rise in trigs. I think the form of the carbohydrate makes a big difference. I think Ornish and those guys often make that point.
So are you suggesting that liquid sugars (like HFCS) are more likely to lead to DNL and high triglycerides than a solid form of sugar?
Anonymous said…
With havin so much content and articles do you ever run into any issues of plagorism or
copyright infringement? My site has a lot of unique content I've either authored myself or outsourced but it looks like a lot of it is popping it up all over the internet without my authorization. Do you know any ways to help prevent content from being stolen? I'd definitely appreciate it.


Have a look at my webpage; www.culturachianti.it
Bris Vegas said…
The Kitavans have a very extreme eating pattern. They gorge virtually all their food in one quick session each night and fast for about 23 hours per day. They also eat fruit ONLY when doing physical work. This extreme fasting/gorging pattern very likely explains their unusual blood lipids and triglycerides.
newbie said…
RE -
What of those with low VLDL-TG but high LDL? To me the answer seems obvious but I'm open to other ideas. To me it means you are producing a lot of VLDL that rapidly lose TG but are not removed by the receptors leaving a lot of LDL -- both particles and cholesterol in those cute and fluffy lipoproteins. There is another possibility of direct LDL manufacture/secretion ... a topic for another time (if ever, limited time and all that)



Did you ever get around to posting any other thoughts on this scenario of lowTG/highLDL? In a lean person with a normal OGTT, not in chronic ketosis?