De Novo Lipogenesis Again (and Science and Science Reporting)
In my last post, I just posted a graphic taken from an "FYI" inset by a scientist, Marc Hellerstein, whose work I've blogged on quite a bit here. (Anyone wanting to refresh can click on the image and enlarge.) One good post to begin with if you are not familiar with that name is Where do Triglycerides Come From? I would say that between Marc Hellerstein's No common energy currency: de novo lipogenesis as the road less traveled and the works of Eric Jequier such as Nutrient effects: post-absorptive interactions, the case that hepatic DNL is quantitatively insignificant in human energy balance/fat accumulation is strong and has not been refuted by some newer scientific findings.
I do hope to expand on the role of adipose DNL at some future point -- teaser: I have a paper showing that reducing WAT lipolysis upregulates adipose DNL and improves glucose homeostasis in the IGT irrespective of body weight -- but adipose DNL had not been central to the arguments of folks from Taubes to Lustig to Eades to Naughton to a list far too long to go through here. The claim has been that excess carbs are converted in the liver to fat and sent to the fat tissue for storage.
Does this occur? Yes. Is it significant? Only if you consider a couple grams fat synthesis per day significant at the metabolic cost "wasting" one gram carb for every four grams converted to fat ... and only if you consider a few grams of fat significant in the context of consuming 100, 200 or more grams of fat that is already fat.
So on the last post, Beth of Weight Maven blog made the following comment.
I must admit to being pretty skeptical of extrapolating the results of overfeeding carbs to normal weight folk. A quick Google search turned up a more recent piece on 60 Minutes from 2012 with one of the researchers in the above piece (Jean Marc Shwarz) telling Sanja Gupta re fructose that "we know that if you look at specific types of fat in the blood there is huge proportion maybe 30% to 40% of that fat in some condition can come from that sugar." Gupta: "So if people say are you sure sugar gets turned into fat, you say yes?" Shwarz: "Here it is ... it becomes very obvious that fructose was a powerful way to stimulate the conversion of sugar to fat."
In a frustrating way to start the day, I lost the rather lengthy response to an inadvertent browser refresh, but it's all the same because I decided this would make a better blog post after all. Unfortunately the video is not in a format I can embed here but here's a screenshot to keep the visual fresh.
There's the powerpoint on the computer, the GC/MS in the background behind Jean-Marc Schwarz in the lab coat and Sanjay Gupta .... the setup is classic staging for science reporting nowadays. The caption is somewhat laughable, as mass spectrometry was hardly described as cutting edge roughly 25 years ago when I used it in a former life, and Schwarz (with Hellerstein) has been using this method for roughly 20 years. Caveats aside, I picked this screenshot for what's on the computer screen: fructose on the lower left, and palmitate on the lower right.
Schwarz has been involved in studies designed to elucidate a hypothesized role for fructose in metabolic disorders for a while now, and I'm afraid he's gotten a little caught up in the hyping of it all. As someone who worked with gas chromatographs in a past life (though usually with an electrochemical detector, not a mass spec) I kinda cringed a little at the description of how they separated the fatty acids in gas in the oven, but ... In any case, Gupta sets this up as that there was some mystery as to whether fructose could be converted to fat ... as if that were some novel discovery. The conversion of carbs to fat is discussed on p. 201 of Newsholm & Start's Regulation in Metabolism (1973, thus at least 40 year old biochemistry though I'm sure it's been around much longer I just don't have the time to track down the actual discovery) so it should hardly be a surprise and it's really just scientific journalistic sensationalism that Gupta is engaged in here. I want to discuss two parts of this discussion. First:
Gupta: So after separating out the fatty acids and then analysing it ... you can definitively say that sugar that was tagged like this ... got turned into ... not only did it get turned into fat but it got turned into a very specific kind of fat.
Gupta: I find it ... I find that amazing still ... Because I ... I don't ... know ... Did you, did you believe that this was true before you started studying this?
So Dr. Sanjay Gupta who seems by all accounts to be a very accomplished doctor (what happens to these guys/ladies when they hit the media???) is acting like he never heard of this concept before, that this is some novel finding! I find that disturbing, but this is what journalism has been reduced to. It's not like he's new to this sort of gig and he sounds -- as a medical doctor -- amazed at the prospect that sugar can be turned to fat in the human body. Sigh.
Nobody claims that carbs cannot be turned into fat. Nobody claims they aren't turned into fat. Where the issue lies is in the degree to which this happens, and more specifically, the degree to which this happens in the liver and subsequently contributes to adipose tissue lipid stores. The overwhelming evidence points to hepatic DNL being quantitatively insignificant in humans. The books and videos out there are incorrect, and the technology discussed in this particular 60 minutes piece is hardly new: Schwarz JM, Neese RA, Basinger A, Hellerstein MK. Effect of oral fructose on lipolysis, fat oxidation, fractional and absolute de novo lipogeneis (DNL) using mass isoptomer distribution analysis (MIDA).FASEB J 1993;7:867(abstr). Unfortunately I cannot track down that reference cited in this 2001 article: De novo lipogenesis during controlled overfeeding with sucrose or glucose in lean and obese women . I'll get to this in a moment, but first I want to address the highlighted part of Gupta's comment above. Because of (a) what that fat is, and (b) how the shocking findings cited by Schwarz are not really all that shocking after all. So fructose is bad ... really bad ...
Gupta: ... You're saying fructose is getting turned directly into the, some of the worst kinds of fat in the body. (Schwarz can be heard saying yes, yeah in the background.)
This presents a clear inconsistency for those arguing that saturated fats are not harmful. If they are not, then what would be the problem with endogenously produced palmitic acid? And if they are, there's no comparison between a few grams of carb-derived palmitate compared to the fat bombs so many are proudly downing.
Schwarz: We know that if you look at specific fat in the blood, there is huge proportion, maybe thirty percent to forty percent of that fat in some condition, can come from that sugar. And that's, that's a lot in terms of, of a, of fat in the blood that are associated with health risks.
If you like what he's saying about fructose here, you have to accept that Schwarz is on the side of saturated fat being bad fat side. Otherwise that would be called cherry picking. But Schwarz is not saying anything that conflicts with the Hellerstein inset above. Note the words I highlighted here. He is talking about palmitate, which normally comprises around 25% of free fatty acids in the blood so under SOME conditions it would not be totally unfathomable to find 30-40% of the palmitate from DNL.
Schwarz, despite his statements of scientific satisfaction and surprise and whatnot at the ability for fructose to stimulate DNL, is on the record for almost two decades saying that DNL is not quantitatively significant. During those same two decades there's no peer review literature that I've come across that even comes close to supporting the wild claims made by Taubes, Lustig, etc.etc. Furthermore, more reasonable intakes of fructose in energy balance are so inept at producing any sort of reasonable response, almost all DNL experiments involve some sort of overfeeding or large bolus dose.
To Beth's question as to whether this might be different in the obese, in this paper (previously linked) , DNL was measured in lean and obese under energy balance (control) and 50% overfeeding of either glucose or fructose. Here's the results and conclusions:
Results: De novo lipogenesis did not differ significantly between lean and obese subjects, except with the control treatment, for which de novo lipogenesis was greater in the obese subjects. De novo lipogenesis was 2- to 3-fold higher after overfeeding by 50% than after the control treatment in all subjects. The type of carbohydrate overfeeding (sucrose or glucose) had no significant effect on de novo lipogenesis in either subject group. Estimated amounts of absolute VLDL production ranged from a minimum of 2 g/d (control) to a maximum of 10 g/d after overfeeding. This compares with a mean fat balance of ≈275 g after 96 h of overfeeding. Individual subjects showed characteristic amounts of de novo lipogenesis, suggesting constitutive (possibly genetic) differences.
Conclusion: De novo lipogenesis increases after overfeeding with glucose and sucrose to the same extent in lean and obese women but does not contribute greatly to total fat balance.
Might DNL be significant in VLF scenarios? Possibly. But that is not relevant in the context of the obesigenic Western diet and why it is so fattening.
Schwarz and Hellerstein are among the authors on a number of fairly recent papers comparing fructose to glucose sweetened beverages. Whatever these studies show is tainted by the fact that while fructose is present in free form (sometimes quite high concentrations) in nature, it is generally at least partially balanced with free glucose and/or combined with glucose in sucrose in those same foods. In one study, participants were consuming 150 grams of fructose a day for 10 weeks. I found some interesting new stuff while trying to track down (unsuccessfully) the Schwarz references 10-12 in the above paper. Maybe someday when this topic comes up again I can fashion them into a review post of sorts.
In closing, I want to say something about Fun with Statistics. When presenting data to an audience, it is certainly valid for the presenter to choose whichever accurate means they choose to do so. It is, however, disheartening to see a scientist do so in the mass media to feed into anti-sugar hysteria. Let's presume there's this rare deadly disease that I have a 1:50,000 odds of developing, or a 0.002% chance of developing. And let's stipulate that some study demonstrated that eating the equivalent of 1T sugar per day increased my chances of developing this disease to 1:10,000 or 0.01%. My chances of getting the disease are 5X as great, or I have a 400% greater chance of developing the disease, etc. All of which sound nasty compared to the absolute increase in my risk of 0.008 % .
This is a bit more extreme than the way DNL and fructose are portrayed, but not really all that much because we are often talking consumption exceeding even the purported average consumption of your SAD eater. Couple that with the fact that anything more than low single digit grams of fat are synthesized from carbs and what we have here, folks, is a red herring. Why Schwarz chose to describe this phenomenon he himself has described as some-variation of inconsequential as mind-blowing is unclear. It is perhaps an attempt to dumb down metabolism into a visual the masses can latch onto, but I don't think misleading people on the science is ever the way to go.