Any discussion on the culpability of poor nutrition as the cause of our health woes begins with a discussion on sugar.
One of the world’s experts on this topic is Dr. Robert Lustig, a pediatric endocrinologist at UCSF. Dr. Lustig has great experience treating children with obesity and is really on the front lines of what is becoming an epidemic of childhood obesity. About two-and-a-half years ago, he gave a lecture on the perils of fructose (fruit sugar, which also makes up half of table sugar and high fructose corn syrup). It’s about 90 minutes in length, but the time goes by pretty quickly as Dr. Lustig is an engaging speaker. In addition, Gary Taubes wrote a great piece on sugar toxicity in the NY Times Magazine last year, which references the work of Dr. Lustig. You can find it here. Gary’s article on this topic was the fourth most read feature of 2011 on NYTimes.com (and the most read of all health-related topics). Here’s the video of Dr. Lustig’s lecture:
I’ve highlighted the key points (with corresponding time in video), for those who may not want to watch the video in its entirety:
- 0:09:40 – Helpful summary showing the reduction in fat consumption in the U.S. from 1960 to 2000 (about 45% to 30%) and the concomitant rise in obesity (about 12% to 31%) [Which, of course, doesn’t “prove” anything, it’s just another correlation.]
- 0:23:00 – Change in fructose consumption over time: Prior to WWII (16-24 gm/day); 1977-78 USDA survey (37 gm/day); 1994 NHANES III (54.7 gm/day); Adolescents today (73 gm/day).
- 0:24:00– Perfect political storm of 3 events:
- Nixon and USDA secretary (1973) – insistence to stabilize/reduce food prices.
- Invention of High Fructose Corn Syrup (HFCS), which was half the price of cane sugar and enabled cheap substitution.
- USDA, ADA, AHA, AMA – all call for reduction in fat intake. Why? (For a quick primer on “cholesterol,” you may want to check my previous post on, What is cholesterol?)
- Early 1970’s – LDL-C (The so-called “bad” cholesterol) is discovered (more specifically, a test to measure LDL-C is discovered)
- Mid 1970’s – Observation that dietary fat is correlated with rising LDL-C (“A implies B”) in a subset of people.
- Late 1970’s – Observation that elevated LDL-C is correlated with heart disease and cerebrovascular disease (“B implies C”) [Note: It’s not actually clear this correlation has causation attached to it, in fact most evidence today would tell us that elevated LDL-C does not lead to heart disease and stroke.]
- Early 1980’s – The following connection is (erroneously) made: A implies B, and B implies C, hence A implies C, so no-A means no-C. [I guess it’s easy to see how an untrained person could make this mistake, but anyone who has taken even an intro course in logic knows that if A implies C, it is not the case no-A implies no-C, it is only the case that no-C implies no-A. It’s hard to believe such poor logic was, and is, used to drive health policy. Last editorial point on this – any card-carrying lipidologist today will tell you that the so-called “bad” LDL-C is as relevant to your getting heart disease as your eye color. Heart disease is caused by lipoprotein particles carrying oxysterols into your artery walls. This is not measured or reasonably predicted by LDL-C.]
- 0:33:00 – Overview of Ancel Keys’ flawed “Seven countries study:” Showed the correlation of fat intake and coronary mortality, but failed to explain the cross correlation of sucrose with the fat (that is, sugar consumption rose too, but this was ignored in the analysis).
- 0:58:00 – Fructose metabolism overview (technical, but interesting – feel free to skip if you don’t like biochemistry): Fructose requires more ATP for the first step in its metabolism (fructose to fructose-1-P) than glucose. This requires an AMP scavenger to recycle the ADP and AMP. AMP goes to IMP, which goes to uric acid. This creates the link between fructose consumption and gout and hypertension (uric acid blocks nitric oxide synthase). In addition, the byproducts of fructose-1-P to pyruvate, such as xylose-5-P, upregulate the enzymes that favor the reaction of citrate being turned into acetyl CoA being turned into fatty acids (staying in liver, causing fatty liver) and VLDL (exported out of liver) [i.e., de novo lipogenesis]. In other words, on a per unit basis, much more fructose is hepatically converted into fat than glucose.
- 1:09:00 – Comparison of chronic ethanol exposure and fructose exposure; very high overlap (not surprising, given that ethanol is fermented fructose). As Lustig correctly puts it, fructose is ethanol, but without the buzz…
You could leave Lustig’s lecture thinking that fructose alone is the cause of obesity. There is little doubt that a massive reduction in fructose (e.g., elimination of dietary sucrose and HFCS, and only modest consumption of fruit) does a lot to reduce obesity. But does this mean the rest of carbs get off the hook?
The problem with the “fructose-alone-is-the-root-of-all-evil” argument
Much of what Dr. Lustig says may be correct, but I believe he overstates the importance of exercise in controlling weight (though he acknowledges that most “experts” fail to realize the calorie burning component of exercise is meaningless), the role/importance of fiber, and the lack of harm associated with glucose. I am going to write extensively about these topics at a later date. The “fiber story” is another sad example of observational epidemiology causing more harm than good.
Back to fructose…the main flaw in Dr. Lustig’s argument, in my humble opinion, is that he claims glucose is benign and that fructose (alone) is the culprit of obesity and metabolic syndrome. In this sense, he is partially correct. Fructose, in excess, may be “evil,” to be sure. However, Lustig claims that glucose is “good” in any amount. He does this based on the assumption that the metabolic priority for glucose metabolism is, first, direct oxidation (i.e., real-time use by organs that need it), and second, glycogen formation and storage by the liver and skeletal muscles. He overlooks a few problems, though.
- Direct oxidation of glucose at rest (the state we are in >90% of the time) is quite low. At rest, most adults oxidize less than 20-25 grams of glucose (about 60-90 kcal/hour of glucose). In fact, even during exercise, it is difficult for the mitochondria to oxidize more than 1 gram of glucose per minute. In other words, while this “sink” for glucose is a high priority, it is very limited in size and rate. Most of the time we consume carbohydrates (i.e., glucose precursor) we are supplying much more than we can oxidize at that moment in time.
- Conversion of glucose into glycogen is limited to how much “room” is left in the glycogen tank. Even the most highly trained athlete can only store a finite amount of glucose in the form of liver and muscle glycogen: approximately 400 kcal (120 gm of glucose) in the liver and approximately 1200 kcal (~400 gm of glucose) in the sum of all skeletal muscles, assuming one is starting from a completely depleted reservoir (a profoundly rare physiologic state). The body does not have the potential to store excess glucose, beyond this amount, in a form that is recoverable as glucose. Any excess glucose that is not immediately metabolized, or converted to glycogen, is turned (irreversibly) into fatty acid for storage. At about 47:00 min into the video, Dr. Lustig talks about the fact that glycogen is non-toxic (true) and that regardless of how much the liver stores, it doesn’t cause hepatocellular damage, unlike fructose (true). The problem is, he fails to mention the storage capacity issue. When the liver stops storing glycogen, which it does at about 120 grams, it does convert the excess to fat.
- While the most highly-trained, insulin sensitive individual might be able to replace glycogen (assuming liver and/or muscle have capacity) when they ingest carbohydrate, rather than store fat, an insulin resistant individual is less able to import glucose into muscle to form glycogen. Furthermore, when insulin levels are elevated, fat lipolysis is inhibited, and obviously this problem is confounded in the insulin resistant individual.
Even in the absence of fructose, a diet high in glucose, beyond everything I’ve stated above, still stimulates insulin release from the pancreas. Elevated levels of insulin “turn off” our ability to burn fat and increase our capacity to store fat (see figure below – you’re probably getting used to seeing this figure by now).
Clearly fructose is a significant culprit in obesity and metabolic syndrome. I’ve personally seen many patients with fatty liver (usually a tell-tale sign of alcoholic liver disease), who rarely consumed alcohol, but consumed high amounts of sugar. At the time (i.e., when I was operating on them), I couldn’t make sense of this observation. Now I can.
The impact of chronic sugar exposure is probably more significant than that of tobacco. I’m actually not being hyperbolic. However, eliminating fructose alone will not cure metabolic syndrome and its associated pathology. It will go a long way, but not all the way, at least not in all people. We simply eat too much glucose in our current carbivore lifestyle.
Scientists have done studies showing that if one removes all fructose (e.g., all sucrose, high-fructose corn syrup, and fruit) from subjects’ diets, they tolerate glucose better, PROVIDED glucose intake is limited, in keeping with the constraints I outlined above. If, however, in the presence of fructose restriction, we consume massive amounts of glucose, we will still convert glucose into fat and, as importantly, through elevated insulin levels, switch off our ability to oxidize our fat stores.
To quote Dr. Lustig from this talk, “it’s a numbers game.” We consume far too much glucose to simply fill our glycogen tanks. Invariably, we keep them topped off, and continue to pour the extra glucose into fat, all the while we force our bodies to survive in a high insulin environment.